Drought, Flood, Fire, How Climate Change Contributes to Catastrophes

Drought, Flood, Fire
Every year, droughts, floods, and fires impact hundreds of millions

of people and cause massive economic losses. Climate change is
making these catastrophes more dangerous. Now. Not in the
future: NOW. This book describes how and why climate change
is already fomenting dire consequences, and will certainly make
climate disasters worse in the near future. Chris Funk combines the
latest science with compelling stories, providing a timely, access-
ible, and beautifully written synopsis of this critical topic. The book
describes our unique and fragile Earth system, and the negative
impacts humans are having on our support systems. It then exam-
ines recent disasters, including heat waves, extreme precipitation,
hurricanes, fires, El Niños and La Niñas, and their human conse-
quences. By clearly describing the dangerous impacts that are
already occurring, Funk provides a clarion call for social change,
yet also conveys the beauty and wonder of our planet, and hope for
our collective future.
Chris Funk is an internationally renowned climate hazard scientist

who develops data sets and forecasts that routinely help safeguard
lives and livelihoods around the world. His publications focus on
climate and climate change, highlighting how climate science can
provide opportunities for predicting natural disasters, thus helping
alleviate their dire consequences. His research has been featured in
Science magazine, on TV, in mainstream print media, and on the
radio. He currently directs the Climate Hazards Center at the
University of California, Santa Barbara.
Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 15 Oct 2021 at 10:59:38, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108885348
How Climate Change Contributes
to Catastrophes
Chris Funk
University of California, Santa Barbara
University Printing House, Cambridge cb2 8bs, United Kingdom
One Liberty Plaza, 20th Floor, New York, ny 10006, USA
477 Williamstown Road, Port Melbourne, vic 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,

New Delhi – 110025, India
79 Anson Road, #06–04/06, Singapore 079906
Cambridge University Press is part of the University of Cambridge.
It furthers the University’s mission by disseminating knowledge in the pursuit of

education, learning, and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781108839877
doi: 10.1017/9781108885348
© Chris Funk 2021
This publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2021
Printed in the United Kingdom by TJ Books Limited, Padstow Cornwall
A catalogue record for this publication is available from the British Library.
Library of Congress Cataloging-in-Publication Data

names: Funk, Chris, author.
title: Drought, flood, fire : how climate change contributes to catastrophes / Chris Funk,
description: Cambridge, UK ; New York, NY : Cambridge University Press, 2021. |
Includes bibliographical references and index.
identifiers: lccn 2020056635 (print) | lccn 2020056636 (ebook) |
isbn 9781108839877 (hardback) | isbn 9781108813792 (paperback) |
isbn 9781108885348 (epub)
subjects: lcsh: Climatic extremes. | Climatic changes.
classification: lcc qc981.8.c53 f86 2021 (print) | lcc qc981.8.c53 (ebook) |
ddc 363.34/1–dc23
LC record available at https://lccn.loc.gov/2020056635
LC ebook record available at https://lccn.loc.gov/2020056636
isbn 978-1-108-83987-7 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy

of URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.

to Catastrophes
Chris Funk


www.cambridge.org
doi: 10.1017/9781108885348
© Chris Funk 2021


isbn 9781108885348 (epub)
ddc 363.34/1–dc23
isbn 978-1-108-83987-7 Hardback


to Catastrophes
Chris Funk


www.cambridge.org
doi: 10.1017/9781108885348
© Chris Funk 2021


isbn 9781108885348 (epub)
ddc 363.34/1–dc23
isbn 978-1-108-83987-7 Hardback

CONTENTS
Acknowledgments vii
1 Climate Extremes, Climate Attribution, Extreme

Event Attribution 1
2 Welcome to an Awesome Planet: A Series of Delicate

Balances Support Earth’s Fragile Flame 21
3 The Earth Is a Negentropic System, or “the Bright Side

of Empty” 40
4 Do-It-Yourself Climate Change Science 59
5 Temperature Extremes – Impacts and Attribution:

Shocks, Exposure, and Vulnerability 92
6 Precipitation Extremes: Observations

and Impacts 122
7 Hurricanes, Cyclones, and Typhoons 140
8 Conceptual Models of Climate Change and Prediction,

and How They Relate to Floods and Fires 164
9 Climate Change Made the 2015–2016 El Niño

More Extreme 186
10 Bigger La Niñas and the East African

Climate Paradox 212
11 Fire and Drought in the Western United States 234
vi / Contents
12 Fire and Australia’s Black Summer 252
13 Driving toward +4C on a Dixie® Cup Planet 268
14 We Can Afford to Wear a White Hat 287
Appendix A Few Resources for Further Reading

and Research 305
Index 321
ACKNOWLEDGMENTS
First and foremost, I would like to thank the “first

responders” from all walks of life and all over the world who
inspired me to write this book. Every day, you put yourselves on
the line to help others, teaching us what it means to be “careful”
human beings. This book grew out of dozens of conversations
with volunteer firefighters here in California (Mike Williams,
Ted Adams, Rocky Siegel, and Yesi Thomas) and humanitarian
relief workers focused on Africa (my colleagues at the Unviserity
of California Climate Hazards Center and the Famine Early
Warning Systems Network, as well as John Magrath, James
Firebrace, and Mathilde Berg Utzon). Day in and out, we were
helping inform real-world responses to real world catastrophes,
yet struggling to keep up as climate change contributed to more
extreme hazards. These discussions grew into Drought, Flood,
Fire. Thank you, friends, for your inspiration.
Next, I would like to thank my lovely wife, Sabina, my
adorable children, Amelie and Theo, and my wonderful in-laws,
Maurizio and Rochelle Barattucci. This project has required
years of long weekends and many late nights locked in the
garage. Only the patient support of my family made this work
possible – so thank you! I also would like to thank my many
coworkers and collaborators, both here in the United States, and
in Africa and Europe. I have been incredibly lucky to collaborate
viii / Acknowledgments
with so many brilliant and dedicated people. Our work together

has filled my life with purpose and meaning, and sometimes you
even laugh at my jokes, which is kind. My friends Joe Peterson,
Tana Kincaid, Jim Semick, and Ramzi Hajj helped me believe in
this book. Without their support and guidance, Drought Flood
Fire would have never seen the light of day. I am very grateful
for their time, effort, and insights.
Finally, the assistance of Sarah Lambert, my editorial
assistant at the Press, provided critical guidance as I navigated
the complex process of modern publishing. My remarkable
editor at the Press, Matt Lloyd, provided excellent suggestions
for modifying the text and structure of this work. I have been
very lucky to have had his support and suggestions. I would also
like to thank Aleksandr Kats for his thoughtful and thorough
copyediting.
CLIMATE EXTREMES, CLIMATE
1 ATTRIBUTION, EXTREME
EVENT ATTRIBUTION
Introduction
Drought. Fire. Flood. Words that echo through our
calamitous past and collective future. Words that threaten to
rip our everyday day away, to tear our normal lives apart.
Anticipating, managing, and preventing such disasters
has always been a critical test of civilizations. A test of our
shared humanity.
Consider, for example, the story of Eregae Lokeno
Nakali, shared with me in 2017 by Mathilde Berg
Utzon.1 Mathilde worked as an Africa correspondent for
DanChurchAid, a Danish humanitarian nongovernmental
organization (NGO) dedicated to supporting the world’s
poorest people. Driven by compassion, DanChurchAid works
to create a more equitable and sustainable world. They provide
emergency relief in disaster-stricken areas and long-term devel-
opment assistance for poverty-stricken communities. Together
with Action by Churches Together Alliance, DanChurchAid
1
www.danchurchaid.org/stories/the-global-climate-threat.
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108885348.001
2 / Drought, Flood, Fire
works in 130 countries to provide about $1.5 billion in humani-

tarian assistance each year.
Mathilde and I connected via e-mail in 2017. She was
spending a lot of time in East Africa. I spend a lot of time
looking at East Africa on my computer screen, working with
the US Agency for International Development’s Famine Early
Warning Systems Network (FEWS NET, www.fews.net). East
Africa is one of the most food-insecure places in the world.
In 2017, I had recently started blogging about East
African droughts and their relation to climate change.2
Mathilde had recently traveled to Turkana County in
northwestern Kenya, to document the severe drought that
region was experiencing. There she met Eregae Lokeno Nakali
(Figure 1.1). Eregage is eighty-one years old. Mathilde writes,
“his look is intense and friendly. Between his chin and his lower
lip he has a round, silver coloured piece of jewelry. His cheeks
are sunken and the wrinkles are deep. It bears witness to a long
life. He has experienced how a wealthy life with three wives and
enough food has been replaced by extreme poverty and daily
uncertainty.”
The seminomadic pastoralists of Turkana have evolved
a way of life that is, in good times, well suited to their barren
surroundings. They raise herds of livestock that can capitalize
on sporadic episodes of rain, turning this precious moisture into
milk and meat. But Eregae tells Mathilde that in the last year
and half there has been almost no rain. He tells Mathilde that he
has lost his wife, sister, and daughter due to disease attacking
their starved bodies. Eregae and his neighboring villagers have
seen their livestock herds destroyed. Eregae takes Mathilde to
see the grave of his daughter (Figure 1.2). Mathilde describes the
scene:
Sand coloured, brown and grey stones the size of a fist

lie placed in a row in a light brown sandy landscape.
2
Funk C., Concerns about the Kenya/Somalia short rains,
October 19, 2016, blog.chc.ucsb.edu/?p=10.
3 / Climate Extremes, Climate Attribution, Extreme Event Attribution
Figure 1.1 Eregae Lokeno Nakali. Photograph by Mathilde

Berg Utzon.
Around the stone pattern lie dried branches that the

wind has spread randomly across the landscape. This
is where Eregae’s daughter Napua Eregae lies buried.
Only a local person from the area could point out her
grave, which is almost invisible in the surroundings.
It was in May last year that her body gave up to
hunger. Eregae had no food to give her, and he couldn’t
afford to take her to the hospital that was too far away.
He explains that he couldn’t do anything while his
daughter died before his eyes.
“It was terrible. I found it very difficult,” Eregaye
says.
And as though it was not enough for Eregae Lokeno
Nakali to lose three of his closest family members, he
has also lost his wealth, his identity and employment.
All because of the drought and almost no help
from outside.
Unfortunately, East African droughts like the ones that afflicted
Eregae and Aita are becoming more common. What used to be a
one-in-five-years drought now often happens every third year or

so. And now warmer air temperatures accompany poor rainy
seasons in regions that are already very hot. This heat can
exacerbate the atmosphere’s ability to pull moisture from the
soil and vegetation, desiccating plants and weakening life-
sustaining herds of livestock. These more frequent hotter
droughts prevent families like Eregae’s from accumulating
wealth. More frequent disasters sap resilience, enforcing a
crushing cycle of poverty. In a poor year these households’ herds
of goats, sheep, cows, or camels may be destroyed, a huge
financial loss for people often living on close to $1 a day.
Eregae and Aita live in the arid Turkana region near the
northwestern tip of Kenya. Since Mathilde visited in 2017, the
region has been wracked by weather extremes. In the main
March–May rainy season of 2018, severe floods in this region
and other parts of Kenya displaced more than 300,000 people and
led directly to some 132 deaths.3 Then, in the next October–
December 2018 “short” rainy season, the rains came late and
were well below normal. Crop production was dismal, about
30 percent of average. Milk production from livestock (a key
source of nutrition) dropped by 50 percent. The March–May rains
of 2019 were once again below normal, and the region once again
faced extreme hunger, conditions right on the border of famine.4
As a scientist supporting FEWS NET, these are the poor
people I try to protect. Every year, NGOs such as
DanChurchAid and government agencies such as FEWS NET
and the World Food Programme (WFP) provide billions of
dollars in humanitarian assistance to people like Eregae and
Aita. These efforts are largely successful at minimizing loss of
life. But the number of severely food-insecure people in the
world is growing rapidly, and our ability to limit economic
devastation remains very limited.
3
www.reliefweb.int/report/kenya/ocha-flash-update-5-floods-kenya-10-may-2018
4
Kenya Food Security and Nutrition Working Group, “Short Rains Food And
Nutrition Security Assessment Findings,” Kenya Food Security Meeting, March
8, 2019.
Mathilde and I work at opposite ends of a spectrum of

international experts dedicated to preventing, or at least respond-
ing to, international disasters in the developing world. Mathilde
works with a faith-based NGO helping motivate charitable
interventions. I am a hard-core academic at the University of
California, Santa Barbara, specializing in using satellites and
computers to identify and predict climate hazards. Mathilde
works with people. I look at data. We both try to help people
like Eregae’s young daughter Aita Eregae Nakali (Figure 1.2).
I share Mathilde’s story with you because it conveys, in ways that
my numbers, charts, and maps cannot, the human cost of climate
extremes. Climate change is making climate extremes more fre-
quent and intense, not just in Kenya, or in Africa in general, but
on every continent. Not in the future, but right now.
Every day I work on predicting and monitoring climate
hazards. I follow the weather and climate closely, designing and
developing information systems that help humanitarian relief
agencies save lives and livelihoods. Like a doctor, or a drought
detective, I diagnose the factors driving extreme events, so I can
improve our capacity to monitor and predict them. I never
wanted to study climate change, but it crept into my office like
an unwelcome guest, like the coal dust lurking behind a miner’s
black lungs. Twenty years of data-driven diagnoses brings me to
this book. Climate change is hurting people like Aita Eregae.
Over the past few years (2015–2020), the “fingerprints”
of climate change have seemed more like a slap. Extreme heat
waves, floods, droughts, and wildfires have exacted a terrible
toll on developed and developing nations alike. These extremes
have impacted hundreds of millions of people and resulted in
hundreds of billions of dollars in losses, all across the globe.
Fire-afflicted movie stars in California; conflagration ravaged
ranchers in Australia; drought-stricken South Africans; poor
flooded fisherfolk in Bangladesh; Houston’s middle-class fam-
ilies riven by flood – these are just some of the people who have
felt the crushing blow of more extreme climate.
But early warning systems can anticipate these disasters,
helping to save lives and livelihoods. Such foresight depends on
Figure 1.2 Eregae’s young daughter Aita Eregae Nakali.

Photograph by Mathilde Berg Utzon
attention and understanding. Viewing the world in an informed

way can open the door to a more meaningful and careful life.
Perception is a function of both sensory input and our internal
conceptions about how the world works. Perceiving the
contribution that climate change is making to real-world disas-

ters can clarify our moral, rational, economic, and existential
imperatives. My goal here is to provide a data-driven “climate
change toolkit” that you can use to see for yourself the influence
of climate change on real-world weather and climate extremes.
Along the way we will also learn a lot of neat stuff about our
beautiful planet. Understanding how our miraculous mother-
ship works can only enrich all our lives.
Exploring Recent Extremes

Since global temperatures jumped up the 2015/16 El
Niño, the world has experienced a dramatic increase in the
number of extreme events. Climate risks arise through the inter-
action of climate shocks, exposure, and vulnerability. Climate
shocks in turn are comprised of natural and human-induced
(anthropogenic) components. So untangling the specific contri-
bution of climate change to multiple extreme events is extremely
difficult, and beyond the scope of this book. But we can learn
how, in general, climate change is influencing these events.
There is little doubt that climate risk is rapidly increasing.
Figure 1.3 shows the number of extreme events between
Figure 1.3 The number of extreme events for each year, based on
the Munich Re reinsurance company’s Natural Catastrophe
database.
1980 and 2018, based on the Munich Re reinsurance company’s

Natural Catastrophe database.5
Reinsurance companies insure insurance companies,
helping regular insurance companies cope with the huge poten-
tial losses associated with severe events like floods and fires. So
companies like Munich Re track the number of different
weather-related disasters closely. For these companies, such
information is critical to their survival. What should be concern-
ing to all of us is that the frequency of such events appears to
have more than tripled since the early 1980s. Each vertical bar
in Figure 1.3 shows the number of big weather-related catas-
trophes in each year, beginning in 1980 and ending in 2018.
While there are year-to-year variations, we also see a large
upward trend. According to this data, the number of extremes
has risen from about 200 a year in 1980 to about 800 in 2018.
These catastrophes are not caused by weather alone.
A humanitarian disaster – of course - always involves humans,
and these weather-related catastrophes arise through the com-
plex interplay of human exposure, vulnerability, natural vari-
ability, and human-induced climate change. Still, this time series
should really get your attention, especially if you think of any
children with affection. These children face an increasingly dan-
gerous adulthood. According to Munich Re, these extremes are
accompanied by billions in losses (about $755 billion [€680
billion]) between 2015 and 2018.6
While both news reporters and climate scientists have
been very interested in these recent extremes, there has been
little in the popular literature addressing these events in an
accessible way. That is what we are doing here. This book
examines extremes over the 2015–2019 time period, cataloging
their severity, and explaining how climate change may have
contributed to their intensity or magnitude. The 2015–2019
time period was chosen because this period has been
5
www.munichre.com/en/reinsurance/business/non-life/natcatservice/index.html.
6
Throughout this book loss estimates will be provided in inflation-adjusted
“real” values.
exceptionally “disastrous,” and is also recent enough to still

resonate with our personal experience. This work focuses on
categories of extremes with substantial immediate human
impacts (droughts, floods, fires). My goal is to help you both
understand and experience how climate change is contributing
to more frequent and intense extreme events.
As we will explore in detail, climate change is already
impacting us. Relevant recent examples, which we will explore
in this book, include the unprecedented 2017 hurricane season;
the 2017 California wildfire season; extreme global tempera-
tures in 2015–2019; the gigantic 2015/16 El Niño event; the
sequence of repetitive droughts striking East Africa in 2016,
2017, and 2018; severe flooding in Houston and Bangladesh;
coral bleaching in the Great Barrier Reef; and the increase
in wildfire frequency and extent in the western United States
and Australia. These events have had catastrophic impacts.
In 2017, ten Atlantic hurricanes arose in rapid succession:
Franklin, Gert, Harvey, Irma, Jose, Katia, Lee, Maria, Nate,
and Ophelia.7 In August 2017, Hurricane Harvey dropped
heavier and more widespread rainfall than any other U.S. trop-
ical cyclone on record - more than 60 inches or 152 cm in at
least two locations (p. 6 of https://www.nhc.noaa.gov/data/tcr/
AL092017_Harvey.pdf). resulting in approximately $85 billion
in estimated damages according to Munich Re. In September
2017, Hurricane Irma reached Category 5 on the Saffir–
Simpson scale with sustained winds of over 185 miles per hour –
for more than 37 hours – longer and stronger than any hurri-
cane on record.8 Then Maria, another Category 5 storm, dev-
astated Puerto Rico, causing almost 3,000 deaths and leaving
millions without power for months on end.
In California, where I live, wildfires ravaged 4.5 million
acres between 2015 and 2018, an area almost as big as the state of
New Jersey. In 2018, the official cost figures for the Camp and
Woolsey fires were $9 billion and $2 billion dollars, respectively,
7
www.nytimes.com/2017/10/11/climate/hurricane-ophelia.html.
8
www.bbc.com/news/world-us-canada-42251921.
with estimates for all wildfire losses ranging from $15 billion to
$19 billion.9 Estimates for 2017 hover around $18 billion.10
According to 201711 and 201812 reports from the risk manage-
ment/reinsurance company Aon, the global economic cost of nat-
ural disasters in 2017 and 2018 totaled $653 billion, the costliest
back-to-back years for weather disasters on record. Public and
private insurers paid out over $237 billion in 2017 and 2018. In
2019–2020, a staggering and globally unprecedented 21 percent
of Australia’s forested area burned,13 as climate change contrib-
uted to extreme temperatures and dry conditions.14 Ecologists
estimate that a billion or more animals (mammals, birds, and
reptiles) may have perished in these conflagrations.
In 2017, hurricanes Harvey, Maria, and Irma produced
some $240 billion in damages. Flooding and Typhoon Hato in
China resulted in $15.5 billion in losses. Extreme precipitation
and a landslide in Sierra Leone led to the catastrophic death of
1,441 people. Droughts in East Africa pushed some 13 million
people into severe food insecurity: these millions of people faced
a very real threat of famine.15 The 2017 drought in southern
9
www.insurancejournal.com/news/west/2018/11/27/510160.htm.
10
www.artemis.bm/news/california-wildfire-industry-losses-put-at-13-2bn-by-aon-
benfield/.
11
thoughtleadership.aonbenfield.com/Documents/20180124-ab-if-annual-report-
weather-climate-2017.pdf.
12
thoughtleadership.aonbenfield.com/Documents/20180124-ab-if-annual-report-
weather-climate-2018.pdf.
13
Boer, Matthias M., Víctor Resco de Dios, and Ross A. Bradstock.
“Unprecedented burn area of Australian mega forest fires.” Nature Climate
Change (2020): 1–2.www.nature.com/articles/s41558-020-0716-1?proof=
trueMay.
14
van Oldenborgh, G. J., Krikken, F., Lewis, S., Leach, N. J., Lehner, F., Saunders,
K. R., van Weele, M., Haustein, K., Li, S., Wallom, D., Sparrow, S., Arrighi, J.,
Singh, R. P., van Aalst, M. K., Philip, S. Y., Vautard, R., and Otto, F. E. L.:
Attribution of the Australian bushfire risk to anthropogenic climate change,
Natural Hazards and Earth System Sciences Discussion, doi.org/10.5194/nhess-
2020-69, in review, 2020.
www.worldweatherattribution.org/bushfires-in-australia-2019-2020/.
www.nat-hazards-earth-syst-sci-discuss.net/nhess-2020-69/.
15
Funk, Chris, et al. “Examining the role of unusually warm Indo-Pacific sea-
surface temperatures in recent African droughts.” Quarterly Journal of the Royal
Meteorological Society 144 (2018): 360–383.rmets.onlinelibrary.wiley.com/doi/
full/10.1002/qj.3266.
Europe led to some $7 billion in economic losses. In 2018,

Hurricanes Michael and Florence battered the United States,
while typhoons Jebi and Mangkhut and Tropical Storm
Rumbia pummeled China and Japan, with total losses estimated
at around $66.4 billion. During July 2018, torrential rains in
Japan resulted in almost $10 billion in damages. These natural
disasters, combined with the impact of an earthquake,
threatened to push Japan into recession. In August 2018, in
the Indian state of Kerala, a multibillion-dollar flood led to
more than 480 deaths.16 In Europe, in the summer of 2018,
unprecedented wildfires spread as far north as Sweden, and
drought-related economic losses were around $9 billion in
northern and central Europe. Government and industry are
beginning to pay attention.
While humans have always faced the perils of natural
disasters, these data suggest that the human and economic cost
of climate and weather extremes is increasing quickly as our
population and economies expand and our planet rapidly
warms. Understanding the link between extremes and warming
is both a moral and an existential imperative. If global warming
is increasing the intensity of extremes, then we are all harming
people now – drought-affected people like Eregae Lokeno
Nakali in Africa, or the nearly 3,000 Puerto Ricans who per-
ished due to Hurricane Maria, the 88 people who died in the
Camp Fire in California or the billion animals who perished
during Australia’s Black Summer.
These increasing perils also pose an existential threat.
Already the magnitude of these losses is already sufficient to
rattle some of the largest global economies. In poor and
moderate-income countries, the toll of the climate crises helps
retard economic progress and enforce vicious cycles of poverty.
Extreme events can and do contribute to human migration and
refugee crises. Understanding the role of climate change can
improve our crisis preparation and powers of prediction. If we
16
www.indianexpress.com/article/india/483-dead-in-kerala-floods-and-landslides-
losses-more-than-annual-plan-outlay-pinarayi-vijayan-5332306/.
think global warming is increasing the intensity or frequency of

a certain type of disaster, then we can better prepare for these
events. As we will see later, this preparation often involves a
“follow the energy” principle – in other words, we should look
for where climate change is creating predictable disturbances
to the climate system by essentially ‘turning up the volume’ of
natural variations. Good examples of this are recent improve-
ments to the United States’ hurricane forecasting capabilities, or
the successful drought forecasts made for East Africa in 2016 and
2017.17 By learning more about the interaction of climate change
and climate extremes, we can face the future better prepared and
better informed. One goal of this book is to describe how energy
moves through the Earth’s energy system, so you can both better
appreciate the beauty of our life-sustaining complex planet, and
how human-induced warming is altering this system in dangerous
and alarming ways.
Extreme Event Attribution and Prediction

This book draws in an informal way from the new
science of extreme event attribution. This new science seeks to
rigorously identify the “fingerprint” of climate change within
weather and climate extremes.18 Like Sherlock Holmes or an
epidemiologist, attribution scientists examine the causes of dan-
gerous events, asking questions like “when Eregae and Aita
suffered severe droughts in 2016 and 2017, were these droughts
made more intense or more probable due to climate change? Or
when the region flooded the following year (in the spring of
2018), was that event made more probable due to climate
change?” “Climate attribution” is the accessible process of
assigning causal explanations. I work in this research field
because it improves my ability to provide effective support for
humanitarian relief agencies. I started as a drought detective,
17
www.sciencedaily.com/releases/2018/03/180314144932.htm.
18
National Academies of Sciences, Engineering, and Medicine. “Attribution of
extreme weather events in the context of climate change.” National Academies
Press, 2016. www.nap.edu/catalog/21852/attribution-of-extreme-weather-
events-in-the-context-of-climate-change.
Usual Suspects Extreme Cyclones

Rain
Extreme Droughts Wildfires

Heat
Underestimated
El Nino
Hazards
La Nina
Low High
Event Complexity
Figure 1.4 Extreme event attribution schema. The top row lists
the “usual suspects” discussed in the National Academy of
Sciences report on climate extreme attribution (footnote 18). The
bottom row lists underestimated hazards examined by the author:
El Niño and La Niña, which will be discussed in Chapters 8–10.
trying to understand why rains sometimes fail in Africa. This led

to my involvement with climate attribution research. My motiv-
ation is humanitarian. If global warming is making an extreme
event more common, then we should prepare.
Experts describe (footnote 18) that there is a range in
our ability to understand and attribute the role played by cli-
mate change in climate extremes (Figure 1.4). At one end of the
spectrum, the role that warming plays in heat waves is pretty
obvious. As the Earth gets warmer, heat waves get more fre-
quent, warmer, and more extensive in space and time.
Ironically, perhaps, both very dry and very wet condi-
tions (droughts and extreme rainfall events) are the next easiest
to understand and attribute. We can be almost certain that
global warming is going to make both droughts and extreme
rainfall events more common and intense. Why the certainty?

Because the same basic physical relationship underlies both
tendencies. We can be sure the atmosphere will get warmer.
Warmer air can hold more water. The fact that warmer air
can hold more water will make droughts drier and extreme
rainfall events wetter and more frequent.
Think of the atmosphere as a sponge that grows bigger
as the atmosphere warms. When the atmosphere warms, the
individual molecules bounce around more, moving farther
apart. This produces more room for gaseous water vapor mol-
ecules to squeeze in between. This makes it easier for water
molecules to move from plant leaves into the atmosphere – a
process called transpiration. It also makes it easier for water to
evaporate from bare soil. So when conditions are dry and the
atmosphere warms, this bigger atmospheric sponge can draw
more water from the land. This increases the intensity of
droughts. Conversely, when the atmosphere is humid, this
bigger sponge will hold more water. Even though we can’t see
this invisible water vapor, it can be rapidly caught up in storms,
lifted, cooled, and sent hurtling down to earth as precipitation.
So warmer air leads to more intense rainfall. Another important
aspect of these precipitation processes is that atmospheric water
vapor moves around with the winds. An individual water mol-
ecule may be swept into the atmosphere from a distant ocean,
travel thousands of miles, and then converge with H20 mol-
ecules from a totally different part of the ocean, helping fuel
extreme rainfall events like hurricanes.
Climate extremes like El Niño and La Niña fall within
the intermediate range in our ability to understand and attribute
(Figure 1.4). This is where much of my research has been
focused, because these “climate events” evolve much more
slowly than “weather events” – which means that we can some-
times predict their impacts many months in advance, helping
prevent famines.
When El Niño events occur, sea surface temperatures
get exceptionally warm in the eastern tropical Pacific. This
exceptional persistent warmth can trigger droughts in dozens
of countries, some of which (like Ethiopia and Zimbabwe) are

very poor and very food insecure. Very recent research, some of
it mine, has shown that climate change is making the eastern
Pacific sea surface temperature warmer during El Niños. These
warmer ocean conditions can trigger more intense droughts in
Africa, Asia, Indonesia, and Central/South America. El Niño’s
sister, La Niña, often follows a strong El Niño event. Ocean
conditions in the eastern Pacific cool while temperatures in the
western Pacific become very warm. These exceptionally warm
west Pacific conditions can trigger severe droughts in places like
Kenya and the western United States. Understanding these rela-
tionships can lead to important forecasts. For example, in late
2016 we predicted the spring 2017 drought that struck Kenya,
Somalia, and southern Ethiopia. This prediction helped motiv-
ate early and effective humanitarian responses, helping save the
lives and livelihoods of thousands of people. These types of
climate extremes begin very simply; we know that the oceans
are getting substantially warmer. It then gets more complicated
when we analyze how these warming ocean conditions can
trigger extreme droughts and precipitation over land.
At the most difficult end of the spectrum, it can be very
hard to attribute the role that climate change may play in
individual wildfire or hurricane events. Humans frequently start
wildfires, and once started, wildfires are driven by very compli-
cated interactions between vegetation and highly local weather
and winds – which can themselves be driven by wildfires.
Hurricanes and cyclones are also rapidly evolving, complicated,
nonlinear, chaotic disturbances. Climate scientists have more
difficulty modeling these events and their changes. But as we
will see, there does appear to be solid observational data linking
global warming to more dangerous fires and more devastating
hurricanes, cyclones, and typhoons.
Book Structure and Intent

Between 2015 and 2019, climate and weather extremes
placed millions of people in harm’s way while causing hundreds
of billions of dollars in damages. Climate change is making

weather extremes more frequent and intense. This book pro-
vides an accessible entry point for nonspecialists who want to
know how and why. Recognizing how climate change is exacer-
bating expensive and dangerous weather and climate extremes,
now, should help us understand why we need to immediately
curb our rampant greenhouse gas emissions. But understanding
why climate change is happening so rapidly may be just as
important. Our atmosphere is very thin. Yet our rapid growth
in global prosperity is unnecessarily predicated on burning fossil
fuels, which threatens the very basis of our fragile life support
system. To understand how greenhouse gasses are messing with
our planet, it helps to appreciate how awesome the Earth–Sun
system is. We absorb solar radiation and turn it into growing
complexity. This is our Earth’s “fragile flame,” a miraculous
negentropic order-inducing life support system that we are
really messing up.
Addressing the question of how, Chapters 2–4 provide a
sometimes lighthearted introduction to climate change and cli-
mate science. These chapters seek to explain how our Earth–Sun
system works, how unique and beautiful it is, and why it is so
fragile. We live on a “Goldilocks Planet” where a cascade of
energetic balances create excellent conditions for the evolution
and sustenance of life. But this life support system depends on a
very thin atmosphere and the maintenance of temperatures
within a narrow range in which water can take on liquid, frozen,
and gaseous states.
Chapters 5–12 then focus on 2015–2019 extreme heat
waves (Chapter 5); precipitation extremes (Chapter 6); hurri-
canes, typhoons, and cyclones (Chapter 7); El Niños and La
Niñas (Chapters 8–10); and droughts and wildfires
(Chapters 11 and 12). While we can’t examine every event, what
we find, and summarize in Table 1.1, is that the aggregate
impact of these extreme events is massive, from both a humani-
tarian and an economic perspective. Climate change is already
increasing the intensity and frequency of extremes, contributing
to dangerous, expensive, and disastrous climate-related crises.
Table 1.1. Notable extremes and impacts examined in this book.

Category Highlights:
Ch. 5 Temperature Extremes Between 2015 and 2019, 59 extreme-temperature disasters, related to 8,800 deaths, 65,592 injuries,
affecting 4.4 million people, and resulting in $1.8 billion in losses.
Exceptional warmth, over more than 20% of the Earth’s surface, has become the new norm.
These exceptional temperatures threaten the Earth’s basic ecosystem services: fisheries, coral reefs, and
carbon dioxide-absorbing rainforests.
Without reductions in emissions, global temperature extremes may increase by more than +5C (9F).
An analysis of daily global temperature data indicates massive (~15 billion people-days) increases in
observed extreme heat exposure events.
The 2050 climate change projections indicate further increases of about 60 to 75 billion people-days.
Ch. 6 Precipitation Extremes As air temperatures increase, the atmosphere can hold more water, leading to increases in the intensity of
the most extreme precipitation events.
Rainfall observations indicate that global precipitation extremes have already increased by more than 8%.
Between 1998 and 2017, floods, storms, and hurricanes affected more people than any other type of
disaster, impacting 2.7 billion people overall and resulting in $1.99 trillion of recorded economic losses.a
The 2015–2019 disaster data suggests that the most dangerous non-cyclone storms affected 223 million
people, led to more than 9,000 deaths, and resulted in $80 billion in damages.
Ch. 7 Hurricanes, Cyclones, According to NOAA data, in 2015–2019, extreme hurricanes, cyclones, and typhoons caused $315 billion
and Typhoons in damages in the United States.
Between 2000 and 2019, these extremes caused $746 billion in damages in the U.S.
In 2017, Hurricane Harvey dropped more than 50 inches of rainfall in a few days on the Houston area
causing $125 billion dollars in damage.
Table 1.1. cont’d

Category Highlights:
In 2017, Hurricane Maria devastated Puerto Rico. The NOAA-estimated price tag was approximately $90
billion dollars. Deaths: estimated at about 3,000 people.
Ch. 8 Conceptual Models of The amount of energy in the upper ocean is increasing very rapidly. Between 1960 and 1990, the total
Climate Change and energy increased by about 31022 Joules. Between 1990 and 2019, heat content increased by more than
Prediction five times this amount.
Between 2014 and 2019, the global upper ocean heat energy increase was equivalent to the energy released
by about 12 million one-megaton nuclear bombs.
This heat moves around in the oceans, interacting with natural variability to produce potentially
catastrophic climate disruptions – and opportunities for prediction.
In October–November 2019, extremely warm western Indian Ocean sea surface temperatures contributed
to extreme flooding and locusts in East Africa and drought in southern Africa and Australia.
Ch. 9 Climate Change Made The 2015–2016 El Niño was associated with extreme drought and air temperatures in Ethiopia, Southern
the 2015–2016 El Niño More Africa, India, Southeast Asia, Oceania, and Brazil.
Extreme These dry arid conditions triggered widespread crop failures, pushing more than 35 million people into
food insecurity.

Climate change made the 2015–2016 El Niño about +0.8 C warmer, making terrestrial droughts
more intense.
Climate change models predict that more extreme El Niños are likely over the next 20 years.
Ch. 10 Bigger La Niñas and The 2015–2016 El Niño was followed by La Niña conditions, characterized by cool East Pacific sea surface
the East African Climate temperatures; but climate change is intensifying the severity of La Niña impacts in some areas by
Paradox dramatically warming the Western Pacific.
In late 2016 and early 2017, East Africa suffered from back-to-back droughts that pushed millions of
people into near-famine conditions.
Climate change enhanced the severity of these droughts by increasing Western Pacific sea
surface temperatures.
This long-term warming of the Western Pacific may explain the East African Climate Paradox.
Ch. 11 Fire and Drought in In the United States, annual wildfire extent observations exhibit a strong upward trend, with average fire
the Western US sizes tripling between the early 1980s and late 2010s.
These increases in wildfire extent are tightly coupled with increases in aridity, which are related to both
increases in air temperatures and upper-level atmospheric ridging.
The 2017 and 2018 US wildfires have caused more than $40 billion in damages, and more than a
hundred fatalities.
Ch. 12 Fire and Drought in In late 2019, half of Australia’s Kangaroo Island burned, killing more than 17,000 koalas, and more than a
Australia third of the island’s kangaroos.
In 2019–2020, a staggering and globally unprecedented 21% of Australia’s forested area burned.
During this “Black Summer,” fires stretched over 186,000 square kilometers (72,000 square miles),
destroying over 5,900 buildings (including 2,779 homes) and killing at least 34 people.
Exceptionally warm conditions and climate change enhanced the intensity of the Australian drought and
associated fires.
Expert assessments indicate that a billion or more animals (mammals, birds, and reptiles) perished in
these conflagrations.
a
United Nations report, Economic losses, poverty & disasters: 1998-2017, www.unisdr.org/files/61119_credeconomiclosses.pdf.
While we hear about these crises in the news, treating them in

aggregate in this book underscores the grave collective nature of
this growing peril.
Unfortunately, as we will see in Chapter 13, our current
emissions put us on track for calamitous warming. But rising
greenhouse gas emissions can also be seen as symptomatic of
beneficial growth. Chapter 14 highlights our growing capacity
to discover, communicate, and create. Education, technology,
and rapid economic growth have lifted billions from poverty.
Our ability to image and observe the world has expanded
tremendously, and we can afford to do the right thing.
Between 1961 and 2050, we will carry out humanity’s greatest
experiment in parallel processing, as billions of individuals
grow, think, discover, and consume. We are living in the midst
of a potentially positive time bomb. The years between
1961 and 2050 will contain as many person-years as
8000–1500 BCE, 1501 BCE–1000 AD, or 1001–1959 AD.
Never have so many seen so much, known so much, or done
so much – or had such a profound capacity to affect the world
for good or ill. We can avoid a global climate catastrophe.19 But
we need to believe in science, believe in each other, and do what
is right. This book serves these goals by contributing to a
broader understanding of climate extremes in a rapidly
warming world.
19
Hoegh-Guldberg, Ove, et al. “Impacts of 1.5 C global warming on natural and
human systems.” Global Warming of 1.5 C.: An IPCC Special Report. IPCC
Secretariat, 2018. 175–311. www.ipcc.ch/sr15.
WELCOME TO AN AWESOME PLANET
2 A Series of Delicate Balances Support
Earth’s Fragile Flame
To appreciate how deeply concerning climate change

is, and will be, we need to start by understanding that we are
living on a truly amazing life support mechanism that only
arose through a series of incredible enabling occurrences. To
our knowledge, there is but a single planet supporting life, the
Earth. We hang isolated in space, with only an incredibly thin
layer of atmosphere standing between us and oblivion. If we
could drive our car straight up at highway speeds, we would
approach the edge of the atmosphere in a matter of minutes.
And into this thin membrane we are dumping about 28 mil-
lion gigatons of carbon dioxide every day.1 The rational
decisions of nearly 8 billion people are resulting in collective
insanity as we choose to destroy the delicate balances that
support Earth’s fragile flame.
Scientists tell us that life on Earth has only been possible
due to a series of delicate balances that arise at all scales of the
universe (Figure 2.1). Physicists often refer to this as the “anthro-
pic principle.” There is a lot debate on this topic. Some scientists
believe that the structure of the universe is inherently conducive
1
www.globalcarbonproject.org/index.htm.
Scale Factor 1 Factor 2

Universe “Dark Energy” Gravity
rapid expansion slow expansion contraction
Inter-galactic Black Hole Expulsions Gravity

Blue Galaxies Green Galaxies Red Galaxies
Andromeda Milky Way M32
Galactic PressureWaves Gravity

Galaxy Edge Mid-point Galaxy Interior
Rapid Formation Modest Formation Slow Formation
Solar Strong Force Gravity

Blue Sun Yellow Sun Red Sun
Planetary Heating Gradient Gravity

Thick Atmosphere Modest Atmosphere Thin Atmosphere
Venus Earth Mars
Hot Temperate Cold
Hadley circulation Heating Gradient Gravity + Angular Momentum

Drier sub-tropics Wetter sub-tropics
Walker circulation Heating Gradient Gravity + Angular Momentum

Stronger trade wind circulation Weaker trade wind circulation
Plant circulation Insolation Transpiration

More vigorous plant growth Less vigorous plant growth
Figure 2.1 Reality as a series of fortunate compromises. At all

scales, the evolving structure of our universe is always composed
of a set of balancing forces.
to the evolution of complexity, life, and intelligence. Others

believe the “miraculous” life-supporting structure of the universe
is simply due to “selection bias,” i.e., a universe has to be pretty
similar to this one to support life in order to be observed.
While this issue is keenly debated, and extremely hard to prove
one way or the other (since we only have one universe as our
scientific “sample”), there is no debating the fabulous series of
fortunate balances that have supported the development of intelli-
gent life on Earth. A brief review of these balances will provide
context for our later discussions of climate change.
Enabling Occurrence 1: Balanced Dark Energy

and Gravity
Throughout this book we will encounter complex
dynamic systems. While tremendously varied, these systems
23 / Welcome to an Awesome Planet
Figure 2.2 A figure from NASA depicting the temporal evolution of the
universe, as seen by the Wilkinson Microwave Anisotropy Probe (WMAP).
At the far left, the Big Bang. To the far right now. map.gsfc.nasa.gov/
media/060915/index.html
always involve multiple balancing forces that evolve across

space and time (Figure 2.1). At universal scales, dark energy
and gravity play a cosmic tug-of-war. The still poorly under-
stood “dark energy” drives the universe toward rapid expan-
sion. In contrast, gravity pulls on the fabric of space-time,
tugging toward contraction. At different stages of the universe,
dark energy seems to have had the upper hand, while at others,
gravity ruled. Informed by satellites, computer models, and
decades of research, astronomers have assembled an exciting
view of the temporal evolution of the universe (Figure 2.2).
Some 13.7 billion years ago, at the time of the Big Bang,
some “quantum fluctuations” are thought to have occurred in a
universe that may have started out as a single dimensionless
point. Astrophysicists posit an incredibly small dense hot place
under incredible pressure. Then the quanta got queasy, and
blammo, in less than one-quintillionth of a second (1 32) the
universe underwent a cosmic inflation. Ripples in space-time
formed the seeds of latent intergalactic structures. At just
one-millionth of a second we began to see the formation of

protons and neutrons. Mind you, average temperatures were
still about 1 billion degrees Celsius. This Big Bang caused the
size of the universe to increase incredibly, leaving behind the
background radiation afterglow that satellites see today in all
directions of the cosmos.
From school you might remember the universal law of
gravitation: F = M (m1 + m2)/d2, where M is the universal
gravity constant, m1 and m2 are the masses of attractive bodies,
and d is the interbody distance. What you may not know is that
a small variation in the magnitude of M would radically alter
the history and shape of our universe. Increase M by 10%, and
the Big Bang that birthed our universe may have already con-
cluded with a big crunch of coalescing and colliding stars.
Decrease M by 10%, and energy and matter may have become
so widely dispersed that no coherent structures could have
formed – no stars, galaxies, light, or life.
But without the repellent force of dark energy, the Big
Bang would have never happened. According to the current
“standard model of Big Bang cosmology,” the force of gravity
is offset by “dark energy” – an elusive, extremely sparse type of
vacuum energy pervading every place in space. While great
uncertainty surrounds this new area of cosmology, most
astronomers now believe that something (dark energy) exists
everywhere in space, and is everywhere exerting a very weak
negative pressure on the space-time manifold. This negative
pressure counteracts gravity, pushing the universe apart.
So at the most massive cosmic scale, scientists have found a
delicate balance between dark energy, with a tendency toward
expansion, and gravity, which produces contraction and
compression.
Just as space-based infrared satellite observations of
distant stars in the 1990s helped fuel a revolution in cosmology,
leading to an ever deeper view of cosmological time, so twenty-
first century space-based microwave observations helped define
and explore the evolving shape of our Universe. By measuring
the radiation emissions associated with the Big Bang, the newest
NASA satellites are tracking the expansion of the universe.

Everything with a temperature emits radiation with a character-
istic frequency. Super-warm things emit very high-frequency
radiation (gamma rays and x-rays). Medium warm things like
our sun emit mostly medium-frequency radiation – visible light.
Very cool things emit radio and microwaves.
The universe is very cool, but not as cool as it could be.
Observed microwave radiation emissions from distant empty
locations are about –269.43C (2.725 K), close to absolute zero,
but not absolute zero. When satellites peered around the uni-
verse, looking for this faint background radiation, the spatial
distribution of these background emissions temperatures was
isothermal – pretty much the same in all directions. This isother-
mal background has been interpreted as the leftover heat from
the Big Bang.
As matter coalesced, the relative impact of dark energy
decreased, and the rate of the universe’s expansion slowed to a
near constant rate. It is estimated that only very small changes in
either the strength of the universal gravitational constant (M) or
the negative pressure force associated with dark energy would
have led to the early destruction of our universe. A little dash
more M, and the universe would crush in upon itself. A little
more dark energy, and all celestial objects would end up hurt-
ling away from each other at speeds beyond the speed of light,
cut off forever from even communication and observation. Each
object would be ultimately alone in a universe that appeared
utterly black.
This, thankfully, was not to be. After about 400 million
years, some pockets of matter became dense enough to support
nuclear fusion. Stars began to form, and our longest Dark Age
came to a close.
Enabling Occurrence 2: A Green Galaxy

As we go from intergalactic to galactic, to solar, to
planetary, and then to biotic scales, life (defined by our sample
of one-planet Earth) appears to spring up where a delicate
balance between forces exists (Figure 2.1). At the grandest

scales, gravity contracts, while dark energy expands, our uni-
verse. Then within and between galaxies, gravity is offset by
black hole expulsions and the pressure waves of rotating matter.
Then within a star, the competition between gravitational attrac-
tion and the quantum-scale strong force determines solar density
and magnitude. The strong force refers to the repulsion between
protons at very small distances. Then with our planet’s oceans
and atmosphere, vertical and horizontal temperature gradients
balance insolation, producing our terrestrial climate. Within
plants, the offsetting effects of solar insolation and transpiration
produce the energy supporting all terrestrial creatures.
Just as contrasts of dark and light create art, or contrasts
in pitch and tone create music, most physical systems need con-
trasting forces in some dynamic quasi-balance. Scientists find
these balances beautiful, and our inquiries into them fueled the
seventeenth century Scientific Revolution. In 1609, Kepler plotted
the orderly sweep of the planetary spheres. In 1638, Galileo
measured the attraction of gravity. In 1668, Isaac Newton con-
nected these scales in one of the most breathtaking leaps of
intellectual thought. He realized that the same gravitational force
that describes the motions of objects on the Earth also controls
the planets in our solar system, resulting in one universal law of
gravitation. The same force drives the dance of planets and my
own body’s descent as I watch the bicycle handlebars pass below
me as I tumble past. But as we shall see, gravity is a creative force.
Understanding why can make our lives more meaningful, careful
and wonderful. The universe is neither cold nor barren, nor is she
likely to coddle and forgive. While the Earth seems special to us
(and it does seem fantastically unique and complex), it is also just
another chunk of rock in space. Like nebulae and galaxies, we all
have to follow the same rules, the laws of physics. The arc of the
cosmos from the Big Bang to our final end resembles in some
ways the arc of a human life. So often what we perceive as stasis is
in fact the result of a delicate balance of forces that can forge
complex and evolving patterns. But so far the push and pull of
galactic gravity has been “just right.”
At intergalactic scales one sees this interplay in blue,

green, and red galaxies. Astronomers believe that many galaxies
have black holes at the center. In blue galaxies, like Andromeda,
a massive black hole at the center pulls in the surrounding
matter, creating rapid rotation speeds while emitting large quan-
tities of high-energy gamma radiation. At the other end of the
spectrum, the center of “red” galaxies such as M32 are thought
to contain small black holes. They rotate slowly, and exhibit less
radiation extremes. Within the middle of the galactic scale, the
Milky Way seems to occupy a “just so” happy medium, with a
black hole at the center big enough to keep things mixing but
not large enough to send gene-smacking gamma radiation all
over the place.
Enabling Occurrence 3: Good Galactic

Real Estate
Within galaxies we see an opposition between gravity
and pressure waves. Pressure waves are generated by galaxies’
rotation and the energy emissions of stars. Near the edge of
galaxies, large rotational speeds tend to support intense pressure
waves and rapid levels of star generation. Near the center of
galaxies, rotational speeds are slow, and weak pressure wave
activity leads to slow levels of star generation. Near the center of
galaxies, where we find our sun, conditions appear to be com-
mendable for life formation. Gravity and pressure wave activity
is strong enough to support star formation, but the pressure
wave intensity is not so severe as to create frequent collisions
and disruptions.
Within our galaxy we see the opposition of gravity
(pulling in) and effects of rotation and radiation (pushing out).
Once again, it seems that Sol’s position is close to “just so”: just
close enough to the Milky Way’s rim to get a decent dose of
high-molecular-weight elements (like carbon, phosphorous, and
oxygen, which are formed by fusion in large stars) but just far
enough from the center that tidal gravitation variations haven’t
pulled us apart.
Enabling Occurrence 4: A Very Nice Star

Within stars we see the opposition of gravity (pulling in)
and effects of the strong force (pushing out). The strong force is
the repulsive force released when neutrons and protons are
forced together during the process of fusion. Some new or
massive stars have very high rates of fusion, and appear to us
as blue in telescopes. These stars typically have a much greater
solar mass, sustain fusion at much higher temperatures, and
release much more gamma radiation. Some old or small stars,
on the other hand, appear to us as red and burn at lower
temperatures. The temperature of a star may have trade-offs
when it comes to supporting life. Cool red stars release less
gamma radiation, which can cause mutations, but they also emit
less potentially life-supporting visible light. Blue stars may
occupy the other extreme. Yellow stars, like ours, may provide
a happy medium, at least for life similar to Earth’s.
Enabling Occurrence 5: A Very Nice Atmosphere

And our Earth itself seems pretty special. At its core,
molten nickel revolves, creating a giant magnet, which in turn
creates a belt of radio waves (the Van Allen belts), which keeps
our precious genetic heritage safe from descending gamma radi-
ation death rays. Without the Van Allen belts, genetic mutations
would likely have been so common as to make life impossible.
The Earth also has a fine assemblage of elements, a reasonably
stable atmosphere and climate, and lots and lots of water, in all
three forms (liquid, gas, and frozen). As far as we know, such
planets are very rare.
Our atmosphere may also be very conducive to life.
Way up high in the stratosphere, a layer of ozone helps protect
us from ultraviolet radiation. At the same time, the nitrogen and
oxygen that comprise the bulk of our atmosphere are transpar-
ent to the solar radiation beaming down on us from the Sun.
Evolving life loves to get all that high-quality incoming energy.
Our atmosphere, ocean, and distribution of land also play an
important role in making things interesting by producing com-

plicated patterns of climate. These complicated climate crenula-
tions support numerous niches in which diverse life forms can
develop to consume the Sun’s bounty.
The Sun’s steady input of energy allows for complex
systems, motions, and emotions to evolve on the Earth. We get
weather and ocean currents because we are differentially heating
the bottom of the atmosphere and the top of the oceans. There
are also important north-south differences in heating. The equa-
tor and summer hemispheres receive more energy, and this
gradient of heating drives atmospheric and oceanic circulations.
It turns out that the amount of atmosphere a planet
maintains has quite a bit to do with a planet’s mass and distance
from its sun. Too small and too close, and the combination of
weak gravity and strong solar wind can cause the atmosphere to
slowly blow away. This has the effect of cooling the planet,
because the atmosphere tends to absorb and re-emit energy.
This process of absorption and re-emission, called “radiative
transfer” by atmospheric scientists, is also called the “green-
house effect.” While there have been some crazy variations in
carbon dioxide levels over the Earth’s nearly 4.5-billion-year
history, they have been pretty stable for about the last
800,000 years (Figure 2.3), ranging for the most part between
approximately 170 and 280 parts per billion. These carbon
dioxide levels, together with the complex influence of clouds
and the redistribution of energy by our oceans and atmosphere,
have kept temperatures in a magical range that embraces a
miraculous substance – water – in all its phases: liquid, gas,
and solid. Gaseous water transports moisture from the oceans
onto land, sustaining terrestrial life. Liquid water teams with
aquatic flora and fauna. Ice caps help reflect some of the sun’s
rays, keeping us reasonably cool.
The Earth occupies a fortuitous middle ground in the
atmospheric arena. We have enough atmosphere of the right
kind to protect us from the worst impacts of high-energy radi-
ation and enough of a greenhouse gas effect to bring us up to a
comfortable average temperature. Temperatures conducive to
Figure 2.3 Atmospheric carbon dioxide levels from ice cores and
the Mauna Loa Observatory. Data between 800,000 years ago
and 1957 are based on ice cores. Values from 1958 to the present
are based on observations at the Mauna Loa observatory. Image
courtesy of Scripps, accessed September 28, 2019. scripps.ucsd.
edu/programs/keelingcurve/wp-content/plugins/sio-bluemoon/
graphs/co2_800k.png
rainfall, rivers and reflective ice caps. This can be contrasted

with planets like frigid Mars, with almost no atmosphere, and
Venus, where we find an intense greenhouse gas effect and
extremely high temperatures.
The current manmade increase in CO2 is also shown in
Figure 2.3. This figure combines very long estimates of carbon
dioxide levels derived from polar ice core estimates with the
1958–present “Keeling Curve,” based on continuous measure-
ments taken at the Mauna Loa observatory in Hawaii. The
Keeling Curve is named after Charles David Keeling, who began
monitoring CO2 at Mauna Loa in Hawaii in the late 1950s. The
current jump beyond 400 parts per million is way beyond the
range of natural variations observed over the last 800,000 years.
We are all on a rock-and-roll ride of unprecedented proportions.
Pause for a moment to ponder Figure 2.3. Before the
end of this book you should understand very well that we are on
a speck of rock in a void in space with a wafer-thin atmosphere,
chain-smoking unfiltered cigarettes and singing along to the

radio as we slam down the gas pedal hurtling our bus into
Dead Man’s Curve.
Enabling Occurrence 6: The Hadley

and Walker Circulations
If all the energy arriving in the tropics stayed there, the
Earth would be much more boring, and most places would be a lot
cooler. Luckily, we have a set of interconnected global circulation
patterns, the Hadley and Walker Circulations, which redistribute
heat from the tropics. Climatologists refer to the north–south
average climate conditions as the Hadley Circulation. The origin
of the Hadley Circulation is the differential heating of the tropics.
This creates a band of low pressure near the equator (Figure 2.4).
Winds are drawn toward this low pressure. As they move towards
the equator, however, the Earth rotates beneath them, from west
to east, causing the apparent westward motion of the trade winds.
Near the equator the northern and southern trade winds converge,
creating vertically ascending motions, often accompanied by rain-
fall. In the upper troposphere (more than 10 km above the Earth’s
surface), these winds move towards the poles and curve east,
forming the jet streams, and feeding into bands of upper-level
convergence near 30 south in the southern hemisphere and 30
Figure 2.4 The Hadley Circulation.
Figure 2.5 Generalized Walker Circulation (December–

February) during ENSO-neutral conditions. Rainfall associated
with rising branches of the Walker Circulation is found over the
area surrounding Indonesia, northern South America, and
central Africa. Credit: NOAA Climate.gov drawing by Fiona
Martin. Downloaded from Tom Di Liberto’s blog, www.climate
.gov/news-features/blogs/enso/walker-circulation-ensos-
atmospheric-buddy, on September 28, 2019
north in the northern hemisphere; where at the surface we often

find subtropical high pressure cells and arid regions like in the
western United States.
We can build up a more detailed picture of how the
global average circulation works by including east–west circula-
tions along the equator, typically referred to as the Walker
Circulation (Figure 2.5). The Walker Circulation was named
after one of my personal heroes, Sir Gilbert Walker, the British
Royal Meteorologist in India during the early 1900s. Motivated
by the terrible Indian droughts and famines at the end of the
nineteenth century, his research into Indo-Pacific climate vari-
ability led to his discovery of the “Southern Oscillation,” the
atmospheric component of the El Niño–Southern Oscillation
(ENSO). In 1896–1897 and 1899–1900, India experienced
severe droughts that produced some of the most severe famines
on record. In 1896–1897, some 5 million people are thought to
have died in British-controlled districts, and 1 million are esti-
mated to have perished in 1899–1900. In 1904, Walker, a British
mathematician familiar with fancy new empirical analysis
techniques such as correlation, entered the British Colonial

Service as Director of the Indian Meteorological Observatory,
with the goal of predicting Indian Monsoon variations.
Combing through the worldwide weather records, Sir
Gilbert discovered the Southern Oscillation. The Southern
Oscillation captured the “covariability” of the sea level pressure
variations near Darwin Australia and Tahiti. Under normal con-
ditions (Figure 2.5) higher pressure over Tahiti and lower pressure
over Australia helped drive an east–west wind pattern associated
with heavy rainfall and ascending air in the area surrounding
Indonesia. This pressure differential helped drive the east-to-west
surface winds that sweep across the equatorial Pacific. These
winds feed heat and moisture into the area around Indonesia,
fueling intense precipitation and rising air.
During an average year, this region (10S–20N, 70E–

180 E) receives a whopping 82 inches or 2 meters of precipitation.
This much rain across 4 million square kilometers makes a cube of
water 43 km on each side. If all this water were to condense at
once, it would release about 184 billion terrajoules of energy –
roughly equivalent to 184 billion big nuclear bombs going off at
once. We receive a tremendous amount of energy from the Sun,
and quite a bit of it ends up evaporating water from the oceans,
and then much of this energy ends up being released as precipita-
tion over the area surrounding Indonesia, where the eastern Indian
and western Pacific oceans meet. This region of exceptionally
warm ocean waters is referred to as the Warm Pool. This energy
rises up, is carried by eastward-flowing “westerly” winds in the
upper troposphere, and is then lost as radiation to space as the
upper-level winds travel east. This cooling, combined with rela-
tively cool sea surface conditions along the eastern coasts of the
Americas, causes subsiding motions that help produce the dry
subtropical high-pressure cells to the west of Mexico and Peru.
These high-pressure cells, in turn, help drive the westward-flowing
“easterly” trade winds that feed moisture into the Indo-Pacific
Warm Pool. Together, the components of the Walker
Circulation comprise a giant “heat engine” that keeps the general
circulation of the earth circulating, with all of the energy, ultim-
ately, coming from the Sun.
As we have seen with astronomical phenomena, the

strength of the Hadley and Walker Circulations arise, again,
through a balance of forces. The Hadley Circulation is largely
driven by the strength of the gradient in surface warming
between the poles and the equator, along with the speed of the
Earth’s rotation. This rotation feeds the “apparent” Coriolis
force, which turns equatorial-headed winds to the west (in the
Northern hemisphere, as in Figure 2.4). Given a stronger heat
gradient, or stronger rotation effects, we would expect to see
stronger trade winds near the equator, and stronger westerly
winds in the mid-latitudes. The influence of heating and spin-
ning are not identical. Spinning, in general, tends to produce
“banding”; when we look at images of Jupiter or Saturn, for
example, the bands of these large planets are formed of counter-
rotating longitudinal cells (bigger planets tend to have higher
surface rotation speeds).
If the Earth’s rotation slowed sufficiently, we would
expect to have just one boring vertically overturning circulation.
At all longitudes, air would meander down toward the equator,
warm, and rise. Then, near the top of the tropopause, it would
move to the poles, sinking somewhere like Manitoba or the
Central Siberian Plateau. Such a unicellular circulation would
produce much greater north–south climate extremes. At present,
mid-latitude winds and ocean currents transport a great deal of
heat to Western Europe and America. This mixing of warm and
cool waters keeps much more of our planet in a nice temperate
regime. Conversely, increasing our rotation speed would create
more counterrotating bands, like we see on Jupiter or Saturn,
potentially disrupting ecosystems, habitats, and migration pat-
terns. Fortunately, weather patterns on Earth have been, for the
most part, conducive to the rapid development of life.
Enabling Occurrence 7: Amazing Life

To the long list of “just so” life-supporting features
associated with our world, we might also add expeditious loca-
tion and timing (Figure 2.6). We find ourselves in a nice part of
Figure 2.6 Universal timeline showing the background radiation

released by the Big Bang, a nebulae, a galaxy, the Earth, and
stromalites in Australia.
the Milky Way Galaxy: not too close to the center where we are
likely to get fried by high-energy radiation being emitted by the
galactic core, but not so far out on the fringes that the density of
energy and matter is terribly sparse. The Earth formed as a
chunk of rock some 4.5 billion years ago, and then slowly
cooled. The first evidence of life (that we know of ) appeared
about 3.8 billion years ago, in the form of small strands of
organic carbon now embedded in sedimentary rocks. These
compounds may indicate early life forms living near thermal
vents. They could have used the vents as a source of energy.
There does seem to be a good case to be made for a very early
evolution of life, perhaps very early indeed in Earth’s history,
but since the oldest rocks on Earth have since been melted, the
evidence has been erased.
Over time, the atmosphere of the planet has changed a
lot. At first, the atmosphere had a high concentration of carbon
dioxide because of all the intense volcanic activity during the
Earth’s first billion or so years. Then CO2 levels dropped, and
oxygen increased, because of the rise of plant life and
photosynthesis. Atmospheric CO2 was also locked up in sedi-
mentary rocks, like limestone and coal, which are themselves
produced by life. In other words, life on Earth dramatically
changed the atmosphere on Earth, making the Earth cooler
and more hospitable, at least to animals like ourselves.
Each day, plants perform a subtle dance with desicca-
tion and photosynthesis. If we zoomed in to look at a leaf’s
surface, we would find tiny holes, called stomata, through which
the plants exchange water vapor and CO2 with the atmosphere.
Generally, plants get water from the ground and CO2 from the
air, and then use the magic of photosynthesis to make carbohy-
drates and sugars. The problem plants face is that stomata are
either open or closed, and water and CO2 flow through the same
channels. So when they open the doors to draw in CO2 to fuel
photosynthesis, the plant’s precious vital fluids (water) are also
escaping through the same door. So again, we have a balance.
Plants definitely want enough nice yummy heat energy (incom-
ing solar radiation) to drive photosynthesis, but (especially in
semi-arid regions) they want conditions cool enough to limit
water loss. Our current temperature conditions provide wide
regions of the Earth where that balance can be satisfied.
Welcome to an Awesome Planet

What does the very early evolution of life on Earth
mean? It is very hard to be certain, given the small size of our
sample (1), but maybe it means that life, as an organizing
principle that produces self-replicating, evolving, intelligence-
auto-augmenting organisms with a genetic attraction to micro-
wave popcorn, may not be as rare as we might think, given the
“right” conditions.
In the European fairy tale of Goldilocks,2 a young girl
wanders into the forest and stumbles into a house owned by
bears who are out for the afternoon. She examines the chairs/
breakfast/beds, always finding a middle version that is “just
right.” As our understanding of our Earth and universe has
evolved over the past fifty years, what seems remarkable to
many scientists is how “just right” the universe and Earth seem
to be. At many scales, the details of the universe seem
2
Credit for this discussion goes to Jan Zalasiewicz and Mark Williams in their
excellent book, The Goldilocks Planet: The 4 Billion Year Story of Earth’s Climate
(Oxford University Press). Fred Adam’s Origins of Existence: How Life Emerged
in the Universe (The Free Press) also informed this discussion, as did several
popular articles by Caleb Scharf on the important role black holes play in
maintaining galactic structure.
surprisingly conducive to life, with the complex structures sup-

porting life arising through a delicate interplay of offsetting
forces. In a myriad of ways we seem to live in/on a universe,
galaxy, solar system, and planet that are extremely well suited
for habitation. One so inclined could see this as evidence for a
divine plan or creator. Another interpretation has to do with
“selection bias.” While other configurations for our universe/
galaxy/solar system/planet are very likely, such configurations
are unlikely to support life, and hence be observed. So the
uniquely life-supporting qualities we observe may just be due
to the fact that we are around to observe them. Yet if you
wanted to believe that a benign Creator, or a divine evolution-
ary process, acted to produce a universe conducive to the cre-
ation of life, that would also seem to fit the facts just fine.
Seventeenth-century scientists, such as Galileo, believed that
the Earth was like another Bible, where God had written his
beautiful and mysterious messages for us to discover. I don’t
think exposure to the past four hundred years of tremendous,
astounding, and beautiful scientific discoveries would cause
Galileo to lose faith, but rather to double down, stunned
by wonder.
We only have one planet. Life appears to have formed
on this planet just about as quick as possible, and then exploded
in awesome and amazing ways to fill a crazy number of niches
with fantastic and beautiful creatures. Scientists estimate the
number of species currently populating Earth at around 8 mil-
lion, with perhaps about 7 million more still waiting to be
discovered.3
So welcome to an awesome planet. Going back over our
checklist (Figure 2.1), we find that a nice balance of gravity’s
pull and dark energy’s push resulted in a rich quilt of galaxies
and stars, with us ending up in a comfortable green galaxy, in a
prime mid-spiral section of the Milky Way, rotating around a
nice sun (not too close, not too far) gently emitting in the yellow
3
Most of these are simple eukaryotic creatures; www.nature.com/news/2011/
110823/full/news.2011.498.html.
part of the energy spectrum. A healthy quantity of atmosphere

and greenhouse gasses, along with the Van Allen belts, keep out
most high-energy particles and maintain a reasonable tempera-
ture. The Hadley and Walker Circulations pitch in, creating
clockwise-rotating circulation cells in and over the Pacific and
Atlantic Oceans, transporting heat away from the equator and
depositing it in poleward latitudes. These circulations bring life-
giving moisture to the continents, supporting abundant life.
Over time – and this is very important – life has acted in
such a way as to sequester (reduce) the amount of atmospheric
carbon dioxide, which was very high in the hot heady days
of Terra’s volcanic adolescence. Plants and marine organisms
extracted tons and tons of carbon dioxide through photosynthesis
and the creation of shells and skeletons. These extractions had a
major influence on our atmosphere; an effect that is being undone
at a rapid pace as we burn and release tons and tons of seques-
tered coal, gas, and methane into our atmosphere (Figure 2.2).
Pulling back, what we see at all levels of our world’s
structure, written backward and forward across the face of time
and space in a thousand different ways, is the complex interplay
of forces trading matter and energy in delicate systems that can,
sometimes, evolve and innovate. During the sixteenth and
seventeenth centuries, visionary scientific revolutionaries read
the word of God written in the bones of the world. Today, this
grand adventure bears noble fruit, uncovering a rich tapestry of
complex inventive processes at every scale.
Note, however, that the manual for Spaceship Earth
does not contain a warranty. Right there on the cover, next to
the red “Don’t Panic” logo, is emblazoned “No Returns If
Opened.” The universe provides no guarantee that we can’t
seriously fudge up this miraculous lifeboat.
For religious communities that trace their roots back to
Genesis, I would suggest that humanity’s original sin and con-
comitant free will provide little assurance of divine intervention.
God promised to never send another flood to destroy the world,
but I doubt She will intervene if our greed, neglect, and willful
ignorance leads to Her nice planet’s destruction.
Life has already dramatically changed carbon dioxide

levels and the level of greenhouse gas warming during out
planet’s history. And we are certainly doing so again at a tre-
mendous rate (Figure 2.3). If a divine entity really did go to all
the trouble of helping produce a benign universe for human life,
we are trying very hard to reach out of the cradle and stick our
finger in the power socket. But a key concept here is balance
amid a delicate array of fluxes and forces. This is a challenging,
humbling, but beautiful insight provided by science. One of our
first scientists, Heraclitus, wrote the words ‘Panta Rhei’ (every-
thing flows) in about 500 BC. The more we learn about nature,
the more miraculous it is that everything doesn’t just fall apart.
I hold out my hand. Flex my fingers. Ponder the immediate
complexity of the neurons firing, the coordinated movements
of the muscles and tendons. Think about the release of energy by
adenosine tri-phosphate (ATP) that fuels my muscles. Each day,
my cellular mitochondria, which probably exist due to a eukary-
otic host cell’s happy capture of bacterium in my far distant
past, produce the equivalent of my body weight in ATP, the
principal energy currency of my cells.4 At longer time scales, a
molecule of water in my body might reside for 12 days. Most of
the cells in my body, except for my brain, die and are replaced.
Each of us is more “fluxed” than “fixed,” and it is wonderful
that our bodies can often work so well. I am very glad to be a
small part of this amazing planet. Like a standing wave in a
river, water, energy, matter, and life flows through me, through
us all, binding us together on the bright side of empty.
4
Törnroth-Horsefield, Susanna, and Richard Neutze. “Opening and closing the
metabolite gate.” Proceedings of the National Academy of Sciences 105.50
(2008): 19565–19566.www.pnas.org/content/105/50/19565.long.
THE EARTH IS A NEGENTROPIC
3 SYSTEM, OR “THE BRIGHT SIDE
OF EMPTY”
It’s funny, but of all the profound thoughts, sights,
experiences, and insights from my fifty-four years, I seem to
remember most clearly those times I almost died through my
own acts of stupidity. Damn self-serving, embarrassing, self-
preserving biology. When I design a replacement species, it will
remember my good jokes, which unfortunately arise less fre-
quently than episodes of attempted self-de-existification.
One of the most notable of such episodes arose decades
ago, while I was still in serious bachelor mode. Together with a
bunch of close friends I went canoe camping in Michigan.
There was afternoon beer drinking involved, and afternoon
whiskey drinking as well. And occasional canoeing. So when
we stopped to make camp, I was not at my “woodsman” best
as I went to collect firewood. Not when I broke the fallen rotten
log apart. Not when the nest of wasps winged forth seeking
vengeance. Not when my alcohol-addled brain failed to instruct
me to run. Only when my spinal cord took over control did the
faintest signs of intelligent life emerge, pumping my legs to run
my body through the quickly darkening woods as many wasp
stings came in quick succession.
41 / The Earth Is a Negentropic System, or “the Bright Side of Empty”
Back at the campfire everything was just fine. The tinder

caught, growing flames flickered. Yet venom coursed within my
veins, and suddenly without warning I bent at the waist, pitched
forward, and planted my face in the woodland duff.
Anaphylactic shock then came quickly, my lymph nodes swell-
ing, cutting off the blood flow to my brain. John and Ramzi
went for help. Jack and Steve waited. I gave Steve my pocket
knife and told him he might need to give me a tracheotomy. He
was really not into that idea. More swelling. I figured I was
pretty much done for. I felt pretty okay with that. My mind was
at peace, ready to move on, but some beautiful parting thoughts
blossomed in my mind.
I could see history stretching around us – backward and
forward in time – like a black rubber mat – flexible, resilient,
and yet markable – memorable. Where we placed our feet
mattered, creating channels of propensity and ordered progres-
sion toward a (hopefully) better tomorrow. Like synaptic path-
ways in our brains, our actions left a residue across the
manifolds of space-time, making it more likely that others
followed in our footsteps, wearing in paths toward a future
good or ill depending on the cumulative effect of our individual
decisions. I thought this would be a nice truth to describe in
soliloquy as I exited stage left. Feverish, faint, flushed, and
feeling drawn down a gentle river that flowed in a pink marble
cave toward a glowing light growing closer and closer,
I mouthed my final words of wisdom – which came out as an
eloquent “mrrrmmppphh,” since my tongue and glands were so
engorged. I tried to convey how our individual steps matter,
having mass, helping mold the path of future human time . . . but
all that came forth were guttural groans. Ignominious indeed.
Poor Steve had to sit there with the knife trying to decide
whether or not to – quite literally – cut my throat.
I was saved by a Rainbow Gathering just around the
river’s bend. Or more precisely by some local high school heroes
at the Rainbow Gathering who rushed me to the hospital and
probably saved my life. Rainbow Gatherings are “temporary
loosely knit communities of people who congregate annually in

remote forests around the world for one or more weeks at a time
to enact a supposedly shared ideology of ‘peace, harmony,
freedom, and respect.’”1 My friend John rushed to the
Rainbow Gathering, desperately seeking assistance. Most of
the attendees were no help – too inebriated or high to provide
assistance. But some teenagers from the town next door gener-
ously came to my assistance. They put me in the back of their
car and plowed through the woods, getting me to the emergency
room just in time. I was sliding down that pink marble tunnel
closer and closer and closer to the light.
This chapter flows from that incident, that failed oppor-
tunity to tell a profound truth, the new lease on life, the ensuing
opportunity to go to graduate school and study how this beau-
tiful planet works. Given a few hours left to my life, a candle
guttering and sputtering, a pencil and paper, this I would write.
Principle 1: Emptiness 6¼ Bad

Most of us, deep down, or even not so deep down, feel a
profound sense of panic when confronting our astronomical,
molecular, and psychological isolation. Behind so many of our
frantic efforts to produce, consume, attract notice, and funda-
mentally matter, lies fear of the abyss, the void, and the end of
our existence. We look at the stars at night. We ponder the big
empty. We think about the incredibly insane distances between
the stars. We feel very alone, very small, and very insignificant.
We do the math, deducing perhaps nihilistic despair. Take the
sum of all the stuff (stars, planets, quasars, pulsars, gas, clouds,
nebulae, etc.) and divide it by universal volume, and we end up
with a number very very close to zero [ε/~∞ ¼ bupkis]. Such is
the logic of insignificance. If science tells us that we are not
significant, then perhaps science itself is suspect.
But maybe dividing by the volume of the universe or a
galaxy is dumb math. Who cares about a bunch of nothing?
1
en.wikipedia.org/wiki/Rainbow_Gathering.
Naïve geometry can invoke stupidity, injecting human scales

where they don’t belong. Here, on this planet, we have the feel
of wind, the scent of air; the pigeon/wren/thrush/(insert bird
name here) taking flight in the morning. Not (ever) to forget
the smell and taste of coffee. This can be written in the language
of science.
Thrush taking flight in early morning
Volume of nearly infinite universe
¼ Thrush taking flight in early morning
What if lots of nothing was needed to produce singular
life, a quietly sleeping child? If you can share with me a few
moments pondering the creative necessity of the void, the Big
Empty that surrounds us, you might look out upon the night sky
with greater wonder, and rest assured that the universe is more
benign than you believed.
The Big Empty helps overcome entropy and the univer-
sal tendency towards heat death. Entropy is a measure of the
universe’s organized thermal structure and ability to do work.
The early discovery of entropy and its implications has been
associated with insanity and despair. The idea was first
advanced by Nicolas Léonard Sadi Carnot (1796–1832) in his
Reflections on the Motive Power of Fire (1890). Carnot intro-
duced the idea that the efficiency of a steam engine will be
proportional to the difference in temperatures between the hot
and cold reservoirs, an idea that was later formalized into an
expression for the thermodynamic efficiency of a system:
ðT hot T cold Þ
efficiency ¼
T cold
While Carnot’s focus was on steam engines, it turns

out that his ideas had profound universal significance.
Understanding his description of the energy cycle driving steam
engines also provides us with a key to understanding the
Cosmos, the Earth’s climate, and many types of extreme
weather events. Figure 3.1 shows a diagram of a Carnot Cycle.
On the left and right we find reservoirs filled with hot and cold
'Heathot 'Heatcold
Thot Tcold
Work
Figure 3.1 A schematic diagram of Carnot’s Heat Engine.
fluids. Changes in heat energy (ΔHeat) will pass through a

mechanical device (the circle), which will convert this exchange
of energy into work. According to the first law of thermodynam-
ics, in a closed system energy must be conserved. So the
exchanges of heat and work must sum to zero. According to
Carnot, however, the efficiency of this engine, its ability to do
work, will be proportional to the difference between Thot and
Tcold. So stronger temperature gradients are more efficient.
Unfortunately, Carnot’s work received little attention,
and after serving briefly in the Napoleonic Army, he developed
dementia and entered a private asylum at the age of thirty-six.
His ideas, however, were elaborated in 1865 by Rudolf Julius
Emanuel Clausius, a professor at the Swiss Federal Institute of
Technology in Zurich. Clausius coined the term “entropy,”
seeking a term similar to energy but based on the Greek term
τροπη, which relates to transformation, turning, and change.
Clausius and the British physicist Lord Kelvin placed entropy in
a thermodynamic framework, thereby adding an important new
law of physics. While Isaac Newton was right, energy had to be
conserved, entropy in a closed system also had to increase.
Because heat energy will never spontaneously flow from a cold
to a warm body, energy gradients in a closed system tend to
diminish over time. Entropy therefore increases, and this process
is irreversible, giving rise to the “arrow of time,” a term coined
by Arthur Eddington.
Entropy can be defined as an inverse measure of a
system’s available energy over time; alternatively, it is a measure
of a system’s disorder. Often, available energy is associated with

the gradients within a systems’ molecules. Pour a very strong acid
and base into water and pow: you will get a strong exothermic
(heat-releasing) reaction but (bubble-bubble-stir) will likely end
up with something similar to salty water. Entropy increases. The
acid-base gradient disappears. Pour hot and cold water into a
reservoir: motions will arise as dense areas produce flows toward
less dense regions, leading to an equilibration of temperature. We
can use these gradients to do work, like drive a steam engine. But
when we drive the turbine, there will be mixing. The very hot will
get a little cooler, the very cold a little warmer. The gradient
diminished, our heat engine will eventually become less efficient,
entropy ever increasing as our ability to do work diminishes.
Think of pouring cream in your coffee. Swirl, look,
ooh – pretty pattern. But wait long enough, and everything is
perfectly mixed. The party must wind down.
Think of dropping a very hot slug of molten lead in your
coffee cup and then wrapping it in an infinitely large Styrofoam
wrapper (no energy goes in, no energy comes out). At first, when
there is a very strong temperature gradient, lots of coffee move-
ment will be going on. Over time, though, the temperature
gradients will diminish. Since the temperature gradients drive
motions of the liquid, these will cease. No energy escapes, yet
the increase of entropy brings the system to “heat death.” All
energetically closed systems face this inevitable destiny.
The good thing about these thought experiments is that
they are just that: experimental. In other words: constructed and
hypothetical; and in other words: often wrong, at least locally.
To understand the bright side of empty, to better appre-
ciate the life-shattering beauty emerging from the twist of wave,
wind, branch, and wing, and maybe preserve this planet for our
children, we need to understand entropy and its antonym:
negentropy.
And that we are fundamentally living in an energetically
open system.
And that this life-supporting, energetically open system
has arisen precisely because gravity has produced the Big Empty.
Gravity creates gradients of energy and matter. It pulls

in matter, and these local increases in mass density inculcate
positive feedbacks, pulling in more matter. Soon enough, no
matter what, we get frighteningly high densities and potentially
life-sustaining nuclear fusion. Gravity pulls in matter, forming
and fueling stars, which can support life. To get to this point,
however, requires the Big Empty – the vast swaths of empty
space. Gravity is a Big Empty generator: you can’t have pockets
of intense density without much larger areas of relative emptiness.
The universe’s “empty prairies” are inextricably linked to the
bustle of her dense “big cities.” Over time, entropy can decrease
in a local place (like our solar system). This means that order and
complexity, over time, can increase locally. Life might evolve.
But, the local coalescence of matter must be offset by an increase
in entropy (and a decrease in matter and gravity) somewhere else.
Decreasing entropy means increasing negentropy, increasing
complexity, and an opportunity for the evolution of life. The
term ‘negentropy’ refers to a tendency for a localized system to
become more ordered and complex over time.
So to understand the “value” of the universe we don’t
divide stuff by volume, or the number of kind or brilliant
communications by kilometer squared. Focus on the bright side
of empty, and the fact that stars can support life – and the
universe is a wonderful source of innovation and attraction.
Understanding energetically open and closed systems
can help resolve the central question: If entropy must increase,
then how could life be possible, much less possibly probable? In
the question resides the answer: embedded in the Big Empty.
A lot of empty may be the best thing that ever happened
to us. If it takes a village to raise a child, it takes a mega-load of
nothing to offset the entropy-negating existence of a star. Think
of it this way: What are the chances that a bunch of hydrogen,
helium, and other elementary molecules would just self-
assemble to become a star?2 Nada. But the natural tendency to
2
Fred Adam, Origins of Existence: How Life Emerged in the Universe (Simon and
Schuster, 2010).
reestablish a randomized spatial distribution is offset by the

force of gravity. I am totally pro-gravity. Gravity rocks hard.
Gravity is our greatest friend in the fight against boring. Gravity
makes stuff happen.
At a universal scale, the “empty” and “cold” of the
universe are requisite components of the islands of heat and
complexity we love so much. The universe is not “hostile” to
humans, but we are still terribly fragile, and the scale of the
cosmos grand almost beyond our imaginings. The emptiness of
the universe, from this perspective, is like the silence between the
notes that brings forth music.
Outside our windows sunlight streams down. Climate
scientists study where that energy goes, what it does, how it
moves through the Earth system. As Chapter 2 explored, the
differential heating of the poles and equatorial regions drives
the Hadley Circulation, forcing winds to converge near the
equator, and then returning poleward aloft in the upper
troposphere. These upper-level flows radiate energy and cool,
sinking in the subtropics, giving birth to the deserts and semi-
arid areas near the Tropics of Cancer and Capricorn. This
persistent pattern avoids the curse of entropy by relying on
imports of cheap energy, from the sun, which are ultimately
based on imports of matter from the distant, now empty, parts
of our galaxy that ended up in our star.
So ultimately, history, consciousness, and evolution
appear to depend on the quite specific characteristic of an inside
and an out. Energetically closed systems, in which there is no
outside, will inevitably experience increases in entropy and suc-
cumb to heat death. Gradients will decay, available energy will
eventually disappear. Energetically open systems – those with a
within and a without (like Earth) can keep entropy at bay while
complex systems continue to evolve. Without the vast emptiness
of space, randomly intermixed matter would produce a sterile
form of heat death. Thanks to our Earth’s ability to maintain a
delicate radiative balance, explored more fully in the next chap-
ter, our complex climate supports all the wonders of history and
ecology. There are common aspects to entropy-eating
negentropic miracle machines. Systems like our galaxies, nebu-

lae, our planet, ancient stromatolites (Figure 2.6), and our
bodies. These complex systems all share a common means of
“informing” – producing a within and a without – and evolve
through time within ranges of energy gradients that support
the evolution of semi-ordered behavior. Climate change is
threatening to push the Earth outside our current life-supporting
‘Goldilocks’ zone. We all need to understand the basic physics
of miracle machines. So . . .
Principle 2: Energy Gradients Make Stuff

Happen – or Not
. . . With my last gasps of air, with a flickering candle
illuminating this my last will and manuscript, I would describe
to you how our Earth functions as a fragile flame” The process
is beautiful. We begin, with a lot of void, and a very little
something (the Sun). The sun imports disorder (entropy) from
the empty reaches of space. Extreme pressure gradients support
fusion, which emits high-energy “shortwave” radiation. Visible
and ultraviolet light are examples of shortwave radiation. This
radiation is absorbed by the Earth, and re-emitted as thermal
infrared “longwave” radiation. This is the energy you can feel
when you put your hand over a very hot parking lot. Longwave
radiation has (surprise) longer wavelengths and lower frequen-
cies than does shortwave radiation. Very hot objects (like the
Sun) emit mostly shortwave radiation. Not so hot objects (like
the Earth) emit mostly longwave radiation. Greenhouse gasses
trap the emitted longwave radiation, potentially creating an
energy imbalance.
Luckily for us, the sustained supply of energy from the
Sun has supported a continuous, fairly stable flow of energy for
billions of years. Because the Earth system is energetically open,
constantly receiving energy from the Sun, life and complexity
can evolve. During the seventeenth century, Isaac Newton made
one of the most tremendous leaps of human understanding. He
linked terrestrial and celestial gravity. As above, so below. As
without, so within. As fell the apple, so the Moon orbits the
Earth. Great science exposes connections. Modern science

depends more on group effort, but our collective advances are
no less stunning. In lieu of Galileo’s telescope, we have a digital
imaging revolution rapidly revealing wonders at all spatial
scales. As scientists map, model, and sometimes predict the
evolution of these systems, we find common laws that operate
at all scales. To understand climate change we need to under-
stand the basics of these laws. One of the most important sets of
laws is thermodynamics, how gradients of heat (thermos) relate
to changes in direction and velocity (dynamics).
One of the main laws of thermodynamics tells us that
within an enclosed system, entropy must increase over time. As
the organization of the molecules within the system becomes
more disordered and random, the amount of energy available
in the system to do work will decrease. Figure 3.2 shows sche-
matically such an evolution toward greater entropy. Warm
molecules are denoted by white symbols; cold molecules are
denoted by dark gray symbols. At Time 1, there is a clear
temperature gradient – eight warm molecules on the left – and
the resulting pressure differential produces a well-defined
pattern of motion (indicated with arrows). At Times 2 and 3,
the temperature gradient has diminished, with only six and then
five warm molecules on the left-hand side. The corresponding
pressure gradients and associated motions become muted and
disorganized. By Time 4 we are approaching maximum entropy.
The substance is now well mixed, with equal numbers of warm
and cold molecules on the left- and right-hand sides. Small
random motion and mixing will continue, but the spontaneous
organization of coherent large-scale structures is statistically
impossible.
Luckily for us, the Earth is far from energetically closed.
The Sun supplies a constant source of high-quality energy, and
this energy supports and maintains gradients of many kinds:
temperature and pressure gradients that drive our atmospheric
winds and oceanic currents, chemical gradients that help sup-
port the evolving complexity of life.
What might happen in such an energetically open system?
Figure 3.3 shows how a continuous input of solar radiation, over
Figure 3.2 A closed system experiences heat death. White and

grey stars denote warm and cold molecules. Arrows depict
motion. Time 1 describes a system with a strong heat gradient,
with grey stars warmer than white stars. A well defined pressure
gradient produces coherent large-scale motions. In Time 2 and
Time 3 mixing between the warm and cold molecules results in
more chaotic pressure gradients and more turbulent motions. By
Time 4 the system has experienced ‘heat death’; the distribution
of warm and cold molecules is completely random, all
temperature and pressure gradients are very localized, and all
motions are small and incoherent.
time, can support the Spiral of Life.3 Electromagnetic radiation

comes in different flavors, or wavelengths. Cold objects emit
3
Interested readers may like “The Tree, the Spiral and the Web of Life: A Visual
Exploration of Biological Evolution for Public Murals,” www.mitpressjournals
.org/doi/pdf/10.1162/LEON_a_00321. Joana Ricou and John Archie Pollock
discuss their development of the “Spiral of Life” as an accessible symbol for the
evolution of life – a representation that avoids the shortcomings of outdated “Tree
of Life” presentations.
Spiral of Life
Work
'Heathot 'Heatcold
Thot Tcold
L
Ra ong
io e
at av
di w
n
di tw
at av
io e
Ra hor
n
S
Little Tiny Stars
Lots of Void
Figure 3.3 The Earth–Sun Miracle Machine drives the spiral of life. Spiral
of geologic time produced by Joseph Graham, William Newman, and John
Stacy, 2008, “The geologic time spiral – A path to the past.” pubs.usgs
.gov/gip/2008/58/
low-energy longwave radiation. Hot objects emit high-energy

shortwave radiation. The sun emits shortwave radiation, primar-
ily in the visible-light portion of the electromagnetic energy spec-
trum. This energy is absorbed and re-emitted by the colder Earth
as thermal infrared longwave radiation. Because the Earth is
energetically open (i.e., always getting energy for free), the Earth
system avoids a local increase in entropy and heat death
Time 1
Time 2
Time 3
Time 4
Heating From Below

Figure 3.4 An energetically open system is heated from below. White and
grey stars denote warm and cold molecules. Arrows depict motion.
(Figure 3.2). Sustained over billions of years, this negentropic

behavior supports temperature gradients that drive various circu-
lations at scales ranging from global climate to microscopic
and chemical processes.
Understanding the link between differential heating and
circulations provides deep insights into our long-term average
climate, many anomalous climate and weather extremes, and
how climate change is already altering some of these patterns. In
general, the relationship between temperature gradients and cir-
culations in the atmosphere and oceans is relatively easy to under-
stand. Figure 3.4 shows a schematic representation of the
development of a sustained circulation. At Time 1, we begin at

Time 4 in Figure 3.2. The molecules are randomly mixed, the
system has reached maximum entropy, and motions are small and
random. Now we import energy, heating the bottom left section.
This figure is generalizing the way that an external heat source
can maintain temperature gradients and drive atmospheric circu-
lations. In the real world this differential heating occurs in
all directions: equator to poles, east to west, and up and down.
From the equator to poles, the Sun preferentially heats the tropics,
driving the equator-to-pole Hadley Circulation. From east to west,
preferential heating near Indonesia and the Congo and Amazon
Basins drives the Earth’s east-to-west Walker Circulation. And
vertically, preferential heating near the surface destabilizes the
atmosphere, helping fuel thunderstorms and cyclones.
In Chapter 2 we have already discussed circulations
similar to that in Figure 3.4. Figure 3.5 summarizes some of
these patterns. If Figures 3.4 and 3.5A are interpreted as prefer-
ential surface heating near the equator, and the vertical dimen-
sion corresponds to height in the atmosphere, the north–south
differential in solar heating between the tropics and the poles
provides available potential energy that drives the Hadley circu-
lation (Figure 2.4).
If Figures 3.4 and 3.5B are interpreted as an east–west
up–down atmospheric transect across the equatorial Pacific
atmosphere, then these schema describe how preferential
warming in the western Pacific and over equatorial South
America and Africa maintains east–west temperature gradients
that provide the available energy driving the Walker Circulation
(Figure 2.5).
In the vertical dimension, tropical meteorologists often
use the term “Convectively Available Potential Energy” (CAPE)
to describe the amount of energy available to produce severe
storms or hurricanes. To understand CAPE, think about the
relative energy contained within a parcel of air that was near
the surface of a warm tropical ocean but just got shoved
upward. The difference between the parcel’s temperature and
the surrounding atmosphere is directly related to the parcel’s
Circulation
A. Hadley
Circulation
B. Walker
Convection
C. Deep
Figure 3.5 Heat gradients produce various important atmospheric

circulations. North–south heating differences drive the equator-to-pole
Hadley Circulation (A). East–west heating differences along the equator
support the east–west Walker Circulation (B). At smaller scales, intense
vertical and horizontal heat releases fuel tropical storms and cyclones (C).
“buoyancy,” or tendency to keep rising. A hot air balloon rises

because it is that: a hot air balloon. Hot here is only interesting
in a relative sense. It rises because it is hotter (= less dense) than
the surrounding air. CAPE also goes up when air becomes
moist, and the CAPE calculation takes into account the “heat”
associated with evaporated water vapor (a point we will return
too). This can be very important when trying to understand
climate extremes, including hurricanes and floods.
Here is a description of such an event as a (temporarily)

negentropic process. First, there is a fluctuation, say a chaotic
wind squall, like an atmospheric “rogue wave” that flaps the
butterfly’s wing, creating a temporary convergence of moisture
and CAPE. Most waves, be they sound, light, or watery, tend to
add linearly: i.e., when wave “a” encounters wave “b,” the
resulting pattern is just wave “a” plus wave “b.” Sometimes
such natural meanderings of our atmosphere bring together
squalls, which in warm, moist environments produce uplift
and condensation.
Meandering wave fluctuations can trigger the release of
Convectively Available Potential Energy. When a warm parcel
of air is lifted up into a cooler environment, higher up in the
atmosphere, this produces available potential energy, because
the parcel is buoyant relative to the surrounding air. The parcel
will cool as it ascends, and when it cools enough, water vapor
contained in the parcel will start to condense. When condensa-
tion begins, that also increases the available potential energy,
because the “latent” energy contained in the parcel’s atmos-
pheric water vapor is released. This also makes the parcel
warmer than the surrounding air.
Now we may enter into the land of positive feedback,
where things can get interesting. We all know that nature abhors
a vacuum. So when parcels of air start to ascend, they leave
behind a region of low pressure, which helps to pull in air from
the surrounding region. If this air is moist and warm, and
converging (i.e., coming in from multiple directions), we have
several positive feedbacks going on. First, just in terms of con-
servation of mass, when air masses converge near the surface,
this tends to produce vertical motion. Since they can’t go down,
two colliding air masses tend to go up. Furthermore, if we have
converging moist warm air, we have converging CAPE. These
processes can act to reinforce the strength of our initial
convective outburst.
Once warm moist air starts rising, it will often keep
rocketing on up through the colder portions of our atmosphere
until it reaches a “temperature inversion,” which can often be
the “tropopause” about 12–15 kilometers above the Earth’s

surface. At the tropopause, the atmosphere reaches a very
cold –55C. Above the tropopause we find an increase in
atmospheric ozone, which absorbs sunlight. So, above the
tropopause temperatures increase, which ‘inverts’ the typical
tendency for air temperatures to decrease with height. This
inversion can form a lid, capping the vertical ascent of clouds –
like the common “pancakes” we often see topping massive
cumulonimbus clouds.
At the upper levels of the atmosphere air parcels move
away from cloud tops, cool, and sink. So tropical storms often
form mesoscale (medium scale) convective systems, comprised
of relatively small convective cells associated with heavy precipi-
tation and ascent and relatively broad regions associated with
descending air. Sometimes, over the tropical oceans, storms with
enough rotational spin and CAPE can turn into hurricanes,
cyclones, or typhoons. These are just different names for the
same phenomena. Hurricanes appear in the Atlantic and north-
east Pacific. Typhoons occur in the northwest Pacific. The
Indian Ocean and South Pacific host cyclones.
Figure 3.5C4 shows a typical hurricane structure. Just
like a thunderstorm, a hurricane has convergent inflow near
the surface and divergent outflows near the tropopause. The
energy driving this circulation is provided by moist warm air.
This energy ultimately comes from the ocean, primarily via the
latent energy contained in evaporated water vapor and second-
arily through the heating of the lower troposphere. The latent
heat released in the hurricane’s vertical walls is offset by the
radiative cooling of air in the upper portions of the system,
near the tropopause. It is the combination of this massive
energy release and the hurricane’s rotational structure, caused
by a conservation of angular momentum, that supports the
development of a spectacular and potentially devastating
weather system.
4
upload.wikimedia.org/wikipedia/commons/f/fb/Hurricane_profile.svg.
Conclusion
Thunderstorms, hurricanes, roses, Earth, you: these are
all examples of negentropic systems. We can’t cheat the uni-
verse. Matter energy and entropy must all be conserved. But the
bright side of empty is that by extracting energy from one place
and collecting it in another the universe is capable of supporting
islands of complexity and the evolution of life.
The laws governing these systems produce beauty –
beauty that I simply like to look at. Here, on this planet, we
know that these complex systems have learned to do amazing
things, like both persist and evolve. It is doing both that’s the
tricky part. At a molecular level, some time between 4 and 3.5
billion years ago, simple molecules such as carbon dioxide,
water, and phosphate combined abiotically to form complex
but nonliving organic compounds such as amino acids, nucleot-
ides, and proteins. Then, in ways we don’t fully understand,
“life” arose because a complex chemical soup – maybe in a
shallow pool, maybe near a deep sea thermal vent – began to
develop an inside and an out, as well as the ability to harvest
energy from the surrounding environment, just like a star.
Perhaps ribonucleic acid (RNA) played the pivotal role, acting
as both catalyst and information store. Or perhaps the origin
of metabolic pathways resulted in coherent, replicating, self-
organizing membrane-bound structures. Then came some
means of procreation, patterning, and our first chemical “mem-
ories” arose. Somehow we arrived at a situation in which
nucleic acids could both guide the creation of complex proteins
and their own replication. We are uncertain of the how, but we
know for sure that by 3.5 billion years ago, simple prokaryotic
life arose – because we can still hold the evidence in our hands,
in the form of “stromatolite” rock formations. That’s pretty
amazing. Just like most of the cells in our bodies, the bones of
the earth itself succumb to the ravages of time, being submerged,
weathered, and melted over time. The oldest rock, in fact, is only
slightly older than the oldest stromatolite (approximately 4 bil-
lion years) – that is, if 300 million years qualifies as “slight.”
Simple prokaryotic cyanobacteria (blue-green algae)

evolve and begin the process of photosynthesis. Prokaryotes lack
membrane-bound nuclei and mitochondria. Cyanobacteria
played a critical role in the history of the earth by releasing (via
photosynthesis) vast quantities of oxygen into our atmosphere. In
this oxygen-rich environment, more complex eukaryotic cells
formed, with internal organelles and nuclei holding the cell’s
genetic material. This led to really popular life activities such as
breathing and sex. Then, perhaps due to lucky symbiotic com-
bination of a eukaryote and a proto-bacteria,5 mitochondria
formed, providing a vital source of energy for eukaryotic life –
adenine triphosphate (ATP). This process of aerobic respiration
combines oxygen with sugars to produce cell fuel, which is ATP.
Now the energy of the sun can be captured, stored, and trans-
ported by our bodies. This fueled an explosion of life that con-
tinues apace today.
Finishing this chapter, a deep-seated mental sigh
escapes. I have written the piece on negentropy I have been
pondering for years. As you look up at the stars, my hope is
that you no longer see the Big Empty as negative space. Just like
silence, this nothingness provides the structure for the music of
the spheres.
Between my fateful brush with bee-borne death and
today, I have changed a lot, and raised a family as well. I am
grateful to be so different from the whiskey-drinking yahoo who
almost died in the woods twenty-five years ago. I’m a lot happier
now, hopefully worrying more about others than myself. It’s a
lot more rewarding. Valuing what we have is the first step in not
letting it slip away.
5
Margulis, Lynn, and Sagan, Dorion (1986). Origins of Sex. Three Billion Years of
Genetic Recombination. New Haven: Yale University Press. pp. 69–71, 87.
DO-IT-YOURSELF CLIMATE
4 CHANGE SCIENCE
Some aspects of climate change are complex and hard
to fathom. Some are fairly straightforward concepts and facts
that everyone really needs to understand. This book is mostly
about the latter. Some of the most important mechanisms of
climate change can be understood by everyone: Why do green-
house gasses have such a direct warming effect on our planet?
How does this warming intensify the impact of droughts and
fires? How can this same atmospheric warming, paradoxically,
also increase the frequency of extreme precipitation events
and floods?
This book approaches these questions with a Do-It-
Yourself (DIY) attitude. While we will occasionally make use
of computer simulations results produced by some of the most
sophisticated numerical models ever produced by humans, we
will more frequently look at publically available data (from web
sites you can access yourself ) and examine these results using
plain old common sense mingled with a little atmospheric sci-
ence. Many aspects of climate change are as certain as gravity.
These we will emphasize.
Take for example the relationship between California
fires, warmer air temperatures, and drier vegetation.
For personal context, we begin on December 5, 2017.
This e-mail from my friend Libby came in at just after midnight:
Figure 4.1 Firefighters battled the Thomas Fire in Ventura,

California early Tuesday, December 5, 2017. Tens of thousands
of people in Southern California were evacuated as wildfires
raged. This photo was taken at roughly 2 AM on Manzanita
Court in the hills of Ventura. The firefighters were from Los
Angeles Fire Department Station 10. The single engine arrived
and saved all other homes on this cul-de-sac except the two
pictured. Image credit: Ryan Cullom/RLC Photography
“My town is kind of a lot on fire” (Figure 4.1). From her

apartment in Ventura, Libby watched the hills explode into
flames reaching six stories into the sky, fanned by 50 mile/hour
(81 km/hour) Santa Anna winds. Only contained by January 12,
2018, the Thomas Fire briefly became the biggest and most
damaging fire in California history, destroying more than
280,000 acres and a thousand structures. Yet this baleful record
was not to last long, with the 2018 Mendocino Complex Fire
covering more than 459,000 acres, and the deadly Camp and
Woolsey fires in November 2018 claiming more than 80 lives
and together destroying more than 80,000 structures and burn-
ing over 250,000 acres or 101,00 hectares. As I update this
chapter in early November 2019, a series of powerful Santa
61 / Do-It-Yourself Climate Change Science
Anna wind events, fiercely hot and dry, funneled down the
mountains and toward the Pacific Ocean, and have just sub-
sided, leaving behind fire scars up and down the state, from the
fabled hills near Hollywood to Sonoma County north of
San Francisco.
The Santa Anna wind conditions complicit in California
fires are driven by a strong pressure gradient between the terres-
trial interior of the western United States and the Pacific Ocean;
a high-pressure cell located over the western United States drives
winds toward the relatively low-pressure areas to the west – the
Pacific coast.1 Climate change may be making these winds
stronger, but it is hard to be sure. One recent study by Janin
Guzman-Morales and Alexander Gershunov of the Scripps
Institution of Oceanography, University of California, suggests
that climate change may actually be reducing the frequency of
these extreme wind events, especially during the early part of the
rainy season. This change, together with anticipated changes in
the winter rains, is expected to shift the fire season later in the
year. One aspect, however, is quite clear. California, along with
many other regions of the planet, has become a lot warmer.
Warmer weather has reduced relative humidity values, drying
out vegetation and helping extend California’s fire season into
October, November, and December, when a seasonal increase in
winds has helped lead to many of the most deadly recent
conflagrations.
In today’s web-enabled world, you don’t have to just
take this as a matter of faith. My life is all about weather and
climate data. I love it, live it, produce it, and use it to identify
disasters and help people. This book will present a lot of data,
and most of it you can access yourself using online tools. For
example, good climate records for the United States go all the
way back to the late 1890s, and we can access them using
1
Guzman-Morales, Janin, and Alexander Gershunov, “Climate change suppresses
Santa Ana winds of Southern California and sharpens their seasonality.”
Geophysical Research Letters 46.5 (2019): 2772–2780.agupubs.onlinelibrary
.wiley.com/doi/pdf/10.1029/2018GL080261.
Figure 4.2 Observed June–October average air temperatures

from the southern and central California coastal
drainage regions. Data obtained from www.esrl.noaa.gov/psd/
cgi-bin/data/timeseries/timeseries1.pl on November 8, 2019
resources provided by the National Oceanic and Atmospheric

Administration’s Physical Sciences Laboratory.2 For example,
I have selected the June–October air temperature data for the
central and south coast regions of California (where I live). This
is the period of time between one rainy season and the next,
when California vegetation dries out. These dry conditions,
combined with Santa Anna winds, can produce optimal condi-
tions for fires along California’s central and southern coastal
drainages. How much warming do we see in the data? A lot.
Figure 4.2 shows the annual averages in degrees
Fahrenheit and Celsius. Two horizontal bars in Figure 4.2 denote
conditions typical in the early twentieth century (1895–1914),
and over the recent past (2014–2019). We find a large degree
(pun intended) of warming. Between the beginning of the
2
www.esrl.noaa.gov/psd/cgi-bin/data/timeseries/timeseries1.pl.
twentieth century and today we see 4.5F (2.5C) of warming.

Coastal California has already experienced a very large tempera-
ture increase. You can see for yourself that this long-term
warming is very large compared to the typical year-to-year fluc-
tuations in air temperatures. Typical values for these fluctuations
are 1.1F (0.6C). Later chapters will describe how this warming
contributes to drier vegetation and larger wildfires. In this chapter
we will use a multitude of data sets to describe how the Earth is
warming globally, and also describe how and why that warming
is happening.
But before going forward, let’s first ask ourselves whether
the observed warming is likely to continue. To address this
question, we can use the “Climate Explorer” web tool3 to down-
load the average of a large number of climate change simulations
(described more completely later in this book) for our region of
interest, the Californian coastal region. Figure 4.3 shows a time
series of these modeled air temperature values, along with our
observed air temperature data. The observations are the same
ones shown in Figure 4.2. Both of the simulated and observed
temperatures are presented as running 10-year averages to repre-
sent the long-term tendency. The last value in the black line
shown in Figure 4.3 corresponds to the average observed June–
October air temperature over the 2010–2019 time period. The
next-to-last value corresponds to 2009–2018, and so on. The
dashed line in this figure shows the predictions from the climate
change models, which track very closely with the observations.
A statistical comparison (a regression, not shown) tells us that the
model explains 80 percent of the variability of the observed air
temperature variations. So to date, the best estimate provided by
climate scientists has performed very well. There have been some
fluctuations above and below the climate change prediction, but
these are to be expected, given natural climate variability.
Note the big increase in temperatures that occurred over
the 2015–2019 time period. This more than 1C jump has been
3
The KNMI web tool is a tremendous resource: climexp.knmi.nl/selectfield_
cmip5.cgi.
Figure 4.3 Observed and simulated coastal California June–

October air temperatures. Data obtained from climexp.knmi.nl/
selectfield_cmip5.cgi
annotated with a sad face. As discussed further in Chapter 11,

that sad face represents a deadly and costly increase in fire
extent and tree mortality. By looking carefully at both weather
data and disaster statistics, we can see with our own eyes that
a 1C warming is already having dire consequences. We will
see similar results over and over again, both in this chapter
and in this book: a very rapid 2015–2019 warming associated
with the increased frequency and severity of droughts, floods,
and fires.
Beyond 2020, the dashed line in Figure 4.3 represents
where temperatures are likely to trend if we continue our current
emissions. By about 2045, Californians will experience another
1C (1.8F) of warming, and then in just fifteen more years yet
Solar Radiation Solar Radiation
Greenhouse Gasses
11 11 11 11 11 >1 >1 >1 >1 >1

Surface Surface
Figure 4.4 Schematic diagram showing an atmosphere with and
without greenhouse gasses.
another 1C (1.8F) of warming. These increases are also anno-

tated with sad faces in Figure 4.3. The intent here is serious.
California just experienced a large increase in fire extent, due in
part to an exceptional increase in temperatures. We will likely
see this happen again, and again, over the next forty years.
Curbing emissions now will help prevent these danger-
ous increases. But why should you believe me? First, and per-
haps most importantly, because you understand the basic
physics. They are not really hard to understand.
How greenhouse gasses warm the atmosphere is pretty
straightforward. It is really important that we all understand
this. But almost no one does, because it is not something that
we are taught in school. Even most people who believe in
climate change don’t really understand the basic mechanism
of why adding greenhouse gasses to the atmosphere has to
increase the amount of energy reaching the surface of the
Earth.4 This mechanism is all about the contrast between a
“transparent” atmosphere with no greenhouse effect (left side
of Figure 4.4), and an “opaque” atmosphere with a greenhouse
gas effect (right side of Figure 4.4). The terms “transparent”
and “opaque” refer to how infrared radiation (emitted heat
energy) responds.
4
I owe a debt of gratitude to my PhD advisor, Professor Joel Michaelsen, for
teaching this to me in graduate school. Thanks, Joel.
Under a transparent scenario, incoming solar radiation

comes down and the same amount of infrared radiation is
emitted upward. This equality is indicated by the black and gray
1’s on the left of Figure 4.4. The same amount of energy arrives
(downward black arrows) and departs (upward gray arrows).
The “flavor” (frequency) of the radiation, however, shifts.
When any object emits radiation, the frequency of the emitted
energy will be determined by the temperature of the object.
Warmer objects will emit higher-frequency energy. Because the
Sun is very hot, it emits energy with high frequencies. So, on the
left of Figure 4.4 the incoming solar radiation (downward black
arrows) has a high frequency, mostly corresponding to visible
wavelengths (e.g. red green blue), which our eyes have evolved
to detect. The Earth’s surface and everything on it (like you and
me), emits longer-wavelength, lower-frequency infrared radi-
ation. In an atmosphere without greenhouse gasses (left side of
Figure 4.4), there is a simple balance between the downwelling
shortwave radiation (black arrows) and upwelling longwave
radiation (gray arrows).
Now, let’s add carbon dioxide, methane, ozone, and
other greenhouse gasses to the atmosphere (right side,
Figure 4.4). Furthermore, to keep things simple, let us assume
that all the upwelling longwave radiation is absorbed by a
middle layer of greenhouse gasses. Now, something really
important happens. These gasses absorb the upwelling infrared
energy from the Earth and re-emit it in all directions. Ironically,
this symmetric emission pattern produces asymmetric green-
house gas warming effects. On the way down, visible light
passes innocently past greenhouse gas atoms because visible
light lacks the right frequency to excite greenhouse gas elec-
trons. Then this visible light energy is absorbed and re-emitted
by the Earth’s surface. But the energy that is re-emitted is lower
frequency infrared radiation. This lower frequency infrared
radiation is trapped by greenhouse gasses. So (in our cartoon
world) it is all absorbed by the middle greenhouse layer, which
sends half of its absorbed energy back down to the Earth’s
surface. Which gets warmer, and sends more energy back
upward. This re-emission process mandates an increase in the

Earth’s temperature. It is straightforward physics, a process that
we understand far better than gravity, and there is nothing to
“believe” in. I don’t want you to be concerned about climate
change because you believe me. Rather, I want you to be con-
cerned because you understand how it works.
Understanding how humans can so radically alter the
atmosphere so quickly begins with understanding that the
atmosphere is actually extremely thin. This is not how we
perceive the atmosphere. We look outside at the sky and it looks
really big. But in fact, only a relatively thin band of atmosphere
separates us from space. Over our heads looms about 1 million
kilograms (2.2 million pounds) of atmosphere. But the density
of this layer drops exponentially, as the mass of the air below is
required to sustain the mass of the air above. If we somehow
managed to climb to a height of 16 km, about 90 percent of the
atmosphere would lie below us. From the South Pole to the
North Pole the Earth stretches about 20,000 kilometers –
1,242 times 16 km. Drive your car straight up at highway
speeds, and you reach the edge of the atmosphere – about
10 miles (16 km) – in about 10 minutes. Imagine an egg painted
blue. The atmosphere is as thick as the blue paint.
Another key insight revolves around the difference
between the atmosphere’s volume and its mass. Humans natur-
ally tend to perceive the atmosphere as vast, but science is all
about overcoming our limiting misperceptions. Consider, for
example, the pre-Socratic Greek proto-scientist Heraclitus,5 who
lived in Ephesus around 500 BC. Heraclitus famously stated,
“The sun is the width of a human foot.” Now this is actually
an experiment you can try at home. While Heraclitus was basic-
ally correct, as you can check by lying on your back outside, his
conclusion can be quite misleading. By analogy, when we look at
the sky, we perceive it as massive. But there is really not that
much there up there. While the atmosphere is a very large place,
the amount of mass it contains is relatively small.
5
plato.stanford.edu/entries/heraclitus/.
We can try to wrap our minds around the true size of

the atmosphere by taking all the mass of the atmosphere, pre-
scribing a density equal to air’s density at the surface, and
estimating the resulting volume. This cube would be about
800 kilometers on each side.
This may still sound like a vast volume. But there are a
lot human beings too. Let’s put the Earth’s approximately 8 bil-
lion people into the cube as well. They would each be about
400 meters from each other. Now put each person in a car. Start
the engines. Assume that each car emits about five tons of CO2
per year.6 Imagine those five tons of CO2 building up in your
400 m 400 m 400 m cube of air. Year after year the CO2
lingers and accumulates.
The buildup of greenhouse gasses will trap more
upwelling longwave radiation. The greenhouse effect arises
from two simple, very well-understood physical effects: (1) cer-
tain gas molecules absorb and re-emit longwave radiation, and
(2) energy must follow a balance, i.e., on average, what is
received must be given back. The first aspect, absorption, gets
a little tricky in its details. Gas molecules – whirling, shaking
clouds of electrons – like all the rest of matter, are mostly void.
A water molecule is small, 3 10–13 meters across. The
sentence-ending period dot on this page might contain some
333,333,333,333 water molecules. But the nuclei of the hydro-
gen and oxygen atoms are much smaller in diameter, so the
nucleus of a hydrogen atom and that of an oxygen atom,
connected end-to-end, would span a mere 6.5 10–15 meters
across. The cross-sectional area of the hydrogen-oxygen-hydro-
gen nuclei is only 0.0168 percent of the area of the water
molecule, so it’s just like you concluded as a teenager: a point-
less empty void. Given the vanishingly miniscule chance of a
light photon actually striking a nucleus, why do molecules
6
The 2019 Global Carbon Project estimate for global CO2 emissions was
approximately 37 109 and the 2019 global population was approximately 7.8
109, so 2.6 tons per person would be more accurate. But the main point of this
section is to try to visualize our relationship to our atmosphere.
absorb radiation, and why do different molecules preferentially

absorb and emit specific wavelengths?
Figuring this out is actually why Einstein won his Nobel
Prize. Color is quantized because light waves mainly interact
with atoms through absorption and re-emission. Absorption
and re-emission happen preferentially, based on the particular
wavelength of light and the particular electron cloud configur-
ation (or geometric shape) of a particular atom or molecule.
What Einstein described was the link between the different
energy states of electrons and the ability of atoms to absorb
certain types of radiation, i.e., radiation of specific wavelengths.
As you may remember from chemistry class, electrons have
configurations in different “orbitals,” mostly determined by
the number of protons (since the number of protons and elec-
trons almost always match). But every once in a while, an
electron can absorb a specific amount of energy, and the
“excited” electron will jump to the next highest orbital
(energy level).
When the excited electron descends back to its old
orbital level, it emits a photon with exactly the same amount
of energy that it received, and with the same wavelength as the
energy it absorbed. This is where the change in wavelength
between the incoming “shortwave” radiation from the sun and
the outgoing “longwave” radiation from the Earth’s surface
becomes important. Shortwave radiation (mostly visible light)
does not “see” water vapor or carbon dioxide. Green photons
pass through water vapor molecules without interacting with
the water vapor electrons. The Earth absorbs this visible light.
The Earth re-emits this energy as “longwave” infrared
radiation.
Infrared radiation, it turns out, is able to “see” green-
house gasses. These atmospheric gasses tend to absorb and re-
emit the Earth’s longwave radiation. But when they do so, they
have to emit in all directions. And this makes all the difference.
The parable of Stinky, Bif, Dinxie, and Moo will help us
understand why (Figure 4.5). Stinky and Bif are dogs, and
Dinxie and Moo are cats, but their inter-species differences do
Figure 4.5 Cats and dogs in love. A story of passion, greed, and
education in which the greenhouse effect is revealed, and Dinxie’s
weight loss program is derailed.
not affect the underlying physics. Stinky and Dinxie represent

the Earth’s surface. Bif and Moo represent the top of the Earth’s
atmosphere. Stinky and Bif are in love. Theirs is a simple world
(Figure 4.5a). Bif receives a bone from the solar energy god.7 Bif
gives Stinky the bone. Stinky gives Bif the bone. Bif can’t catch
the bone,8 but that’s okay, because the sun god gives Bif another
bone, and the cycle repeats. Stinky’s final temperature will be
matched to the energy he received from Bif, so that he is emitting
exactly the same amount of energy he received.
Now, let’s look at love in a little more complicated
world. This world’s atmosphere has a middle layer, home to a
dubious complicator we’ll call Taxem. Taxem is so chock-a-block
full up with water vapor, carbon dioxide, and methane, he smells
like a Kansas high plains cattle feedlot crossed with the Los
Angeles 405 Freeway. As our story begins (Figure 4.5b), Moo
also receives an energy fish from the god of the sun. Meoww! He
gives his fish to his best friend Dinxie. Because Dinxie is a cool
cat, she receives a “shortfish” but emits a “longfish,” and is
unselfishly giving it back . . . when whoosh – Taxem steps in
7
His name is Larry. But that’s not important. Really.
8
Why can’t Bif catch Stinky’s bone? A little riddle to see if you’ve been
paying attention.
and eats that fishy up. Taxem does this because his methane and
water vapor and carbon dioxide preferentially absorb longfish. In
this simple world (Figure 4.5b), we’ll pretend that Taxem can re-
emit Dinxie’s longfish in a preferential way (up). In this case, we
end up with a final situation very similar to the Tale of Stinky and
Bif. Taxem gives Moo a fish. Moo can’t catch the fish, but that’s
okay, because the sun god gives Moo another fish, and the cycle
repeats. Again, Dinxie’s final temperature will be matched to the
energy she received from Moo.
Let’s look at cat love in an even more complicated
world (Figure 4.5c). In this world, Taxem operates more like
our real world, emitting energy both up and down. Moo begins
by giving a fish to his best friend Dinxie. Who unselfishly is
giving it back . . . when whoosh – Taxem steps in and eats that
fishy up. This time, though, Taxem has to give the same
amount of fish upward as he does downward. How can every-
one be happy?
We can solve this simple energy balance problem using
algebra. It really does come in handy sometimes, and what
follows may be one of the most important pieces of science that
few people know. We are describing something that is abso-
lutely vital to understand. We are going to work our way
through a simple three-layer energy balance model (similar to
Figure 4.5c). We have three atmospheric levels: the surface
(Dinxie), the middle of the atmosphere (Taxem), and the top
of the atmosphere (Moo). M# (Moo) represents the shortwave
energy flowing into our atmosphere. D" represents the upwelling
longwave radiation flux from the surface (from Dinxie). T" and
T# represent the upward and downward fluxes from greenhouse
gas radiation trapped in the middle of the atmosphere (from
Taxem). To conserve energy at the top of the atmosphere, we
know that M# must equal T" (i.e what goes out the top must
equal what comes in the top). Similarly, at the surface we know
that D" must equal T#+ M#, i.e., the amount of energy emitted
by the surface must equal the amount of energy received by the
surface. We also know that Taxem must emit the same amount
of energy up and down. So we have three equations:
M# ¼ T"
D" ¼ T# þ M#
T# ¼ T"
We have three equations and three unknowns. We can

use simple algebra to solve for D" by substitution: D" ¼ T#+ M#.
Since M# ¼ T", then D" ¼ M#+ M# and D" ¼ 2M#.
Wow. We have just derived a simple model to describe
the basic mechanism of the greenhouse effect. In the absence of a
radiatively active middle atmosphere (Figure 4.5a), and ignoring
complicating factors such as evaporation and the effects of
clouds and convection, the surface temperature would be
exactly warm enough to radiate enough energy to match the
incoming radiation. In the presence of a radiatively active
middle atmosphere, however, the fact that the middle atmos-
phere absorbs and re-emits longwave radiation ensures that the
energy emitted by the Earth will be substantially more than the
amount of incoming solar radiation. This may seem paradoxical,
but we have just used a simple numerical model to explore for
ourselves how this process works. The basic idea behind the
greenhouse effect is simply this: energy must be balanced, but the
middle atmosphere must absorb and emit radiation in all direc-
tions, while the surface of the Earth only radiates energy up. Like
death and taxes, these assumptions are a sure thing. When you put
these simple truths together, you get a surprising result; the surface
of the Earth warms substantially. Increase the amount of green-
house gasses in the middle atmosphere, and the surface of the
Earth must get warmer. This is simply physics and logic.
While simple, the toy model we have explored here is
quite similar in principle to the numerical “radiative transfer”
models used by climate scientists. While some of the details can
be very complicated, and very important, the core science
behind radiative transfer is absolutely certain and relatively
simple. The greenhouse effect is not a theory, but rather basic
physics. Like gravity, it is an everyday phenomenon of our
world. Unlike other types of physics, like the laws that describe
the motions of falling apples, most of us have never been taught
how radiative transfer works, and it is difficult for us to experi-

ence it as viscerally as gravity’s ten meters per second per
second acceleration.
So, it is ironic that “global warming” remains such an
intense topic of “debate.” The physics behind the effect of
greenhouse gasses is much better understood than the mysteri-
ous mechanisms creating gravity, and much more direct than the
circuitous interactions giving rise to evolution. While its effects
are easy to quantify, there is still considerable discussion and
debate about what actually causes gravity. At non-quantum
spatial scales, Einstein’s theories of relativity describe gravity
as function of the curvature of space-time; more matter pro-
duces more curvature. But why? At a quantum level, consider-
able debate continues – with some radically differing opinions
about the basic structure of reality – and the processes that
account for gravity. Is gravity produced by tiny quantum gravi-
tons? By a seven-dimensional super string structure? By m-
dimensional manifolds? Despite continued efforts by some of
our most brilliant minds, we really aren’t sure.
Compared to gravity, the greenhouse effect is easy:
physicists really understand at the core level the processes
involved, and have developed effective models to represent it.
What’s different from gravity, however, is that our personal
experience of greenhouse warming is very limited. Isaac
Newton’s great insight – linking gravitational attraction on
Earth to the attraction of planets, the sun, and the moon – was
brilliant. It took a great leap of insight to link the observable
terrestrial reality of weight – the downward pull of mass on
Earth with the invisible force that drew the planets toward the
sun. Brilliant, mathematically elegant, but ultimately something
we still don’t understand very well scientifically. But on the
other hand, gravity is deeply familiar to us all. When we are
young, we fly over the top of our handle bars, or slide on gravel
and lose the skin off our palms. We grow, stretch, and learn,
laughing in the face of gravity’s pull.
The challenge to understanding the greenhouse effect is
to hurtle our mind’s eye beyond our personal experience. To
ponder and accept the grand balances of nature that enable, and
encircle, the wonder of our lives.
We should also recognize that the greenhouse effect is
not inherently “bad.” Without the greenhouse effect, our planet
would be much colder, too cold to make life comfortable for us.
The greenhouse effect is just one factor among many affecting
the Earth’s climate. In many natural systems, “balance” is typic-
ally shorthand for complex sets of competing forces that interact
to produce complex behavior. In living systems, however, this
process is often formalized by monitoring networks and feed-
back loops capable of supporting “homeostasis.” Our auto-
nomic nervous systems, for example, might monitor our body
temperatures and cause us to sweat when we grow too warm.
Dealing with the impact of greenhouse gasses might not be so
simple. While we have been blessed to live on a fantastic
“Goldilocks Planet,” we have broken the chair, we have eaten
the porridge, and we have made a mess of the bed. Now the big
bad bears are coming home. As we discussed in Chapter 2, the
Earth’s atmosphere has been “just right”: unlike Mars, which
has very little atmosphere, no greenhouse effect, and is too cold
to support life, or Venus, which has a very thick atmosphere,
very strong greenhouse effect, and is too warm for life, the
Earth’s temperature range has been relatively stable as humanity
emerged, and the greenhouse effect helped keep our planetary
temperatures in the magic range allowing frozen, liquid, and
gaseous water. But when it comes to greenhouse warming, there
can certainly be too much of a good thing.
Observational Evidence for Climate Change

Now, using data, let’s review the primary evidence for
current concerns about recent global warming.9 Rather than ask
9
Most of the data examined here was obtained from KNML climate explorer
(climexp.knmi.nl), which is an archive of data and climate simulations used to
support the Intergovernmental Panel on Climate Change‘s Fifth
Assessment Report.
you to “believe” in climate change, we look at the observations

and simulations at the center of the Intergovernmental Panel on
Climate Change (IPCC) Fifth Assessment Report,10 and exam-
ine the factors contributing to the rapid acceleration of emis-
sions over the last decade.
There are three primary lines of evidence. First, green-
house gas impacts originate from the basic physics of radiative
transfer (as described above). Believing in global warming is like
“believing” in gravity. The physical process is not that hard to
understand, but it is not something most people have learned
about. Second, a wide variety of different data sources all con-
verge on the same story: we are experiencing a huge disruption
in our climate, and this disruption will manifest itself in the
twenty-first century, just as population growth and increasing
demands for water, food, and resources will place unpreced-
ented strains on our Earth’s ecosystems. Third, at a global scale,
the agreement between the climate model predictions and the
observed changes that have already happened is very close. If
the models’ future predictions are right – and they almost cer-
tainly are – and we continue to emit greenhouse gasses at an
accelerating rate, then twenty-first-century global warming will
be dramatic, dangerous, and disruptive. Again, simple physics.
We next focus on three separate but convergent sources
of observational evidence: (1) terrestrial air temperature records,
(2) observations of sea surface temperatures, and (3) observa-
tions of global sea levels. We show that all these independent
records converge on dramatic and rapid increases in land and
ocean temperatures. We will also examine a large collection of
climate change simulations, which have been produced to
inform the Intergovernmental Panel on Climate Change‘s
(IPCC) Fifth Assessment Report.
We begin our voyage of discovery simply, by plotting
estimates of global land temperatures going back to 1880
(Figure 4.6). This type of data, based on thousands and thou-
sands of monthly air temperature observations, is one of the
10
www.climatechange2013.org/report/, released on September 30, 2013.
Figure 4.6 NASA GISS annual estimates of global land

temperature anomalies.
main sources of evidence for climate change. These values are

shown as anomalies (differences) from the 1951–1980 mean.
One appealing feature of air temperature observations is their
independence. Each day thousands of measurements are taken,
and each location and time sample provides another piece of
evidence. While systematic shifts or sources of error can occur at
a weather station site (say, by moving the weather station to a
lower, hotter location), it is extremely unlikely that such system-
atic errors could affect large numbers of stations. Time series
like Figure 4.6 are based on thousands of thermometer observa-
tions, with each temperature observation providing an inde-
pendent “vote.” As noted by hundreds of researchers,
warming on land has been dramatic, according to the available
information. Furthermore, the rate of this warming is accelerat-
ing, in line with predictions from climate change simulations.
These data indicate two epochs of rapid warming. The
first epoch, from roughly 1900 through 1940, is thought to be
due primarily to changes in land cover and land use, solar
radiation, and volcanic disruptions.11 After the 1970s, and

through 2019, a rapid epoch of warming reasserted itself,
caused by the rapid accumulation of greenhouse gasses.
During the monster El Niño of 2015/16, massive amounts of
energy were released from the Pacific Ocean, and in 2016 and
2017 global temperatures jumped way beyond the historical
maxima, to a 1.2C anomalies. The years 2015, 2016, 2017,
2018, and 2019 were, by far, the warmest years in the historical
record. These years are marked with circles. Lines also denote
the 1990s average (+0.5C) and the 2015–2019 average
(+1.1C). Just in the last twenty years we have seen a massive
0.6C warming. This book examines climate extremes during
these most recent very warm years (2015–2019).
Another valuable source of long-term climate informa-
tion is provided by the logs of oceangoing ships. Dependent on
wind and wave, sailors have always paid close attention to
ocean conditions, and good ocean measurements stretch back
to the late nineteenth century in many places. The temperature
of the ocean can provide valuable information about a ship’s
location and local fishing conditions. Like terrestrial air tem-
perature archives, these thousands of sea surface temperature
observations each provide separate pieces of evidence, helping
us measure and map the rate of change. Again, these estimates
are based on a large number of independent temperature esti-
mates from ships and buoys. In Figure 4.7 the global average
ocean temperatures are shown as anomalies (differences) from
the 1951–1980 mean. It takes an enormous amount of energy
to heat just one cubic meter of seawater, so the ocean has a
tremendous amount of thermal inertia. So, ocean temperatures
tend to change more slowly than the land. During World War II
(the 1940s) we may see some questionable jumps in the ocean
temperatures, potentially related to disruptions in shipborne
temperature observations. Nonetheless, we see a high degree
of similarity with the air temperatures plotted in Figure 4.6,
especially in the twenty-first century. While terrestrial air
11
IPCC 5th assessment report, www.ipcc.ch/report/ar5/wg1/.
temperatures are more variable from year to year, the general

structure of the land and ocean temperature time series are very
similar to each other. While the ocean has warmed a little less
than the land, because of its high thermal capacity and ability to
lose heat through evaporation and vertical mixing, ocean tem-
perature changes have tracked very closely with terrestrial air
temperature variations. Both have exhibited rapid warming.
The oceans have also warmed about 0.5C between the 1990s
and 2015–2019. We also see that 2015, 2016, 2017, 2018, and
2019 (circles) are way warmer than any previous observations,
even when compared to 2014 (diamond).
Observations of ocean levels from tidal gauges12 and
Earth-orbiting satellites13 provide a third independent set of
global temperature observations. As water warms, it expands
in a very predictable way, so slow changes in sea level heights are
seen as a key indicator (and outcome) of anthropogenic climate
change. Global sea levels have risen substantially (Figure 4.8).
Since the deep ocean warms more slowly than the sea surface, or
land, this long-time series of ocean levels lags behind global land
and sea surface temperatures, but still provides further corrobor-
ating evidence that substantial warming is well under way. These
rising sea levels are also associated with increased coastal flood
risks, especially during cyclones and hurricanes. Massive sea
swells, like those that flowed into New Orleans during
Hurricane Katrina in 2005 and into New York City during
Superstorm Sandy in 2012, were tremendously damaging.
Figures 4.6–4.8 show that three sources of totally
independent data – terrestrial weather station observations,
ship-based sea surface temperature estimates, and global sea
levels – identify a large and coherent warming signal.14 These
12
Jevrejeva, S., A. Grinsted, J. C. Moore, and S. Holgate, “Nonlinear trends and
multiyear cycles in sea level records.” Journal of Geophysical Research 111
(2006), C09012, doi: 10.1029/2005JC003229.
13
climate.nasa.gov/vital-signs/sea-level/.
14
Lots and lots of satellite data also provides corroborating evidence over the past
thirty-five years. We know the most about the Earth when most of greenhouse
warming has occurred – over the past thirty-five years.
Figure 4.7 Global annual ocean temperature anomalies, based

on the NOAA Extended Reconstruction dataset, version 5.
data sets are copious, consistent and continuous, stretching back

to the nineteenth century. Historically, humans paid attention to
temperatures and tide levels because they were important for
agriculture and maritime navigation and trade. These three
independent data sets tell us that the Earth is warming rapidly,
and that even within our “new normal” the 2015–2019 time
period has been exceptional. The data presented here do not
imply that the Earth has never been warmer – it probably has
been at some point in the geological past – but it does suggest
that our rapid expansion in population, fuel usage, and emis-
sions has been accompanied by a very large, rapid, and consist-
ent warming of the Earth’s atmosphere and oceans.
While the Earth may have been warmer in the past, the
recent rate of warming is almost certainly unprecedented. And
what we saw in the time series presented here (Figures 4.2, 4.3,
4.6, and 4.7) is that warming in the last ten years has been
particularly rapid. As we will discuss in later chapters, this rapid
warming is already presenting an existential threat to many
organisms, including humans. The speed of these changes rule
out short-term evolutionary solutions, leaving only migration or
Figure 4.8 Global sea level anomalies. The long black line is
based on historic tidal gauge observations. The shorter, more
recent record is based on satellite altimetry estimates provided by
NASA: www.climate.nasa.gov/vital-signs/sea-level/.
adaption. But coral reefs, poor farmers, and forests have a very
limited capacity to move or adapt.
Figures 4.6–4.8, however, provide little information
about the cause of that warming. Could changes in incoming
solar radiation, linked to variations in sun spots, be the cause?
The data do not seem to suggest this. Figure 4.9 shows estimates
of solar insolation based on satellite15 and terrestrial16 observa-
tions. Incoming solar radiation corresponds with the amount of
solar radiation hitting the Earth. These correspond with the
downward-pointing arrows from Figure 4.4.
15
Ohlich, R., “Observations of irradiance variations.” Space Science Review 94
(2000): 15–24.
16
Lean, J., “Evolution of the Sun’s spectral irradiance since the maunder
minimum.” Geophysical Research Letters 27.16 (2000): 2425–2428.
1,367.5
1,367
Estimated Solar Irradiance [Wm-2]
1,366.5
1,366
1,365.5
1,365
1,364.5
1880 1900 1920 1940 1960 1980 2000 2020
Estimated Solar Irradiance Satellite Observed Solar Irradiance
Figure 4.9 Incoming solar irradiation estimated from weather

observations (black) and satellite observations (gray).
Changes in incoming solar radiation are influenced by

semi-regular cycles associated with sunspots, but these vari-
ations bear little resemblance to recent variations global tem-
peratures, suggesting that this has not been the dominant force
for recent warming trends. Figures 4.6, 4.7 and 4.8 do not look
like Figure 4.9. Note also that the total range of these values is
only on the order of 1–2 W/m2. Certainly these variations can
have some influence, but they bear little similarity to the obser-
vations of air temperatures, ocean temperatures, and sea levels
plotted above. We need to dig deeper.
What about results from the latest generation of
global climate change models? To support climate change ana-
lyses, like the Intergovernmental Panel on Climate Change
Assessment reports, international climate modeling groups from
all over the world build and run state-of-the-science climate
change models. Each modeling effort in this assemblage is actu-
ally a combination of models (or modules really), with each
model/module dedicated to representing a piece of the Earth
system. There is a model for the atmosphere, and another model

for the ocean, and another model for land surface processes (like
plants and rivers), and another model for the “cryosphere.”
Cryosphere sounds like a snow globe to me, but these critical
models relate variations in temperature, precipitation, and sea
surface temperatures to changes in snowpack, glaciers, and sea
ice extent.
Multiple modeling centers all over the world develop
their best (or several best) models. Typically, each model is
initialized using land cover and greenhouse gas conditions
representing the state of the world in 1850. When a simula-
tion begins, all the variables representing atmospheric and
oceanic temperatures and motions have initial values. Then
they get busy perturbing each other. Winds blowing across
ocean waters induce changes in sea surface temperatures.
Changes in sea surface temperatures affect the patterns
of winds. Changes in the winds can produces storms.
Rainfall from storms can release heat in the atmosphere.
This atmospheric heating can influence wind patterns.
Changes in surface winds can affect sea surface temperatures.
And so on.
These models are only constrained by the underlying
physics of the Earth systems involved and a few external factors
such as changing the amount of greenhouse gasses, the amount
of particulate pollution in the atmosphere (aerosols), and the
type of vegetation covering the Earth surface (farm, forest, city).
At each time step, often four times a day, the models simulate
the “coupled” ocean, atmosphere, and land surface. “Coupling”
between the models ensures physical realism, so that energy and
mass balance between the various components of the climate
system. Storm systems in the simulations move east to west and
west to east, tracking north and south with the passing of the
sun and seasons.
To produce the Intergovernmental Panel on Climate
Change Assessments, this process was carried out for many
simulations for the many different models being run by the
many modeling groups. While the weather on any given day, in
any two different simulations, is expected (guaranteed in fact)

to be very different, the average behavior of the simulations
gives us insight into how our world is responding to human-
induced changes in our atmosphere and land surface.
Slowly, between the mid-nineteenth and early twenty-
first century, the external bounding conditions (greenhouse
gasses, aerosols, and land cover) are changed to match the ways
in which we have altered our world. Carbon dioxide concen-
trations change from about 287 parts per million in 1850 to
about 415 parts per million in late 2020.17 Over time, especially
in the early nineteenth century, forests are replaced by cropland
and human habitations.
I know from personal experience what a massive transi-
tion this could be. I grew up in northern Indiana, the flat land
where soybean and corn crops stretched for thousands of square
miles, filling the horizon. In the early 1990s, I was lucky to get
the chance to help my grandfather put together a book describ-
ing this history (Sod, Seed, and Sacrifice).18 The Funk family
came to northern Indiana in the 1850s as homesteaders from
Germany. To do research, Grampy and I drove around rural
northern Indiana, visiting graveyards and historical societies in
little towns like Fowler. There were a few pit stops at some
American Legion Outposts along the way too. Grampy loved
talking with people and telling stories. The story of sod,
seed, and sacrifice is the story of the taming of the American
prairie, whose “sod” stretched across entire states, deep into the
soil, and high up into the air. Breaking that sod was a gargan-
tuan task, often requiring massive plows pulled by a dozen oxen
to cut through deep root systems developed over decades of
growth. In America, corn replaced the prairies. All across the
globe, forests shrank as the global population increased from
1.2 billion in 1850 to 1.6 billion in 1900. Now we can use
global climate models to assess the impacts of these changes.
17
www.scripps.ucsd.edu/programs/keelingcurve/.
18
www.amazon.com/seed-sacrifice-William-Everett-Funk/dp/B0006PED5M.
Figure 4.10 Annual observed and modeled global surface

temperatures. The observed temperatures were obtained from the
NASA Goddard Institute for Space Studies (GISS). The modeled
time series is based on 221 climate change simulations. Future
temperature projections are created using simulations from the
RCP4.5 and RCP8.5 scenarios.
This data are accessible to you if you’d like to verify this analysis
for yourself.19
Figure 4.10 shows mean global surface temperature
estimates from the latest generation of models from the IPCC’s
Phase 5 Coupled Model Intercomparison Project (CMIP5) arch-
ive, together with observed global temperatures. This CMIP5
time series summarizes a tremendous human effort – 221
simulations performed by 39 different models or model combin-
ations. While there are differences among the various simulation
results, they tend to be fairly small. It is amazing how well the
19
I like to use Geert von Oldenburg’s excellent Climate Explorer site, climexp.
knmi.nl. There are many other excellent portals that are also available.
Figure 4.11 Predicted and observed global temperatures,

averaged over thirty-year periods.
simulation results track closely with the observed changes in

global temperatures, from 1861 to today. Remember, there are
no observations used in the climate change simulations, beyond
changes in greenhouse gasses, aerosols, and land cover. If the
model physics were not basically correct, it is very unlikely that
the simulated global temperatures would track so closely with
the observations. The CMIP5 time series is a prediction, based
on the physics of radiation transfer and the fluid dynamics of the
atmosphere and oceans. At a global level, the predictions and
observations agree very well. Figure 4.11 shows thirty-year
averages of the observed and predicted global temperatures
expressed as a scatter plot.
Since graduating from college in 1989, my entire working
career has centered around different aspects of data analysis and
prediction: econometric modeling in Chicago followed by drought
early warning to support famine prevention in Africa. Predictions

don’t get better than those shown in Figure 4.11. Almost all the
variance (96%) of the thirty-year average temperature changes
is explained by the model predictions. The slope between the
predictions and observations, furthermore, is almost exactly 1.
Remember, the models don’t use any observed climate informa-
tion. At a global/decadal scale the models capture very well the
long-term changes in the Earth’s temperature.
The models and observations converge on a story of
rapid change, over the course of a few generations. For example,
my grandfather’s grandfather, Bernard “Barney” Funk volun-
teered for service in the Union army in 1863 during the Civil
War. He served as a mule driver in the Cumberland Army, and
fought in the terrible Battle of Chickamauga20 on September 20,
1863, in which 34,624 soldiers were killed. Returning alive to
Indiana, Barney received forty acres of land from the US gov-
ernment and started to farm. Between that bygone era and
today, the combined impact of land cover changes and green-
house gas emissions has warmed the Earth by about 1C. The
last time the Earth was this hot (125,000 years ago, during the
last interglacial period), sea levels were about 20–30 feet higher
than they are today21 and hippos22 lived where Great Britain is
today.23 Measuring this change in generations, that’s about 0.2C
of temperature increase per Funk clan member (Bernard begets
Edward begets William begets Robert begets Christopher) or 1C
in 150 years. If we run the models forward, they predict that if we
don’t cut the addition of greenhouse gasses to the atmosphere
very soon, then the next 1C (or more) increase is coming in just
the next thirty years. The sins of us fathers will be visited upon
20
Chickamauga is an old Cherokee word that means “River of Death.”
21
Hoffman, Jeremy S., et al. “Regional and global sea-surface temperatures during
the last interglaciation.” Science 355.6322 (2017): 276–279.science.sciencemag
.org/content/355/6322/276.full.
22
www.iflscience.com/plants-and-animals/last-time-earth-was-hot-hippos-lived-
britain-s-130000-years-ago/.
23
The temperature maximum during the last interglacial (125,000 years ago) was
due to changes in solar radiation and the relative location of the Sun, not
greenhouse gas forcing.
the heads of our children and their children. As we discussed

in Figure 4.6, terrestrial air temperatures have already climbed
0.6C between the 1990s and 2015–2019.
Sidebar: The Thomas Fire

Saturday, December 16, 2017, en route from New Orleans to
Chicago. Instead of flying to Washington, DC for my planned
meetings with the US Agency for International Development and
NASA, I have rebooked my flights and am scrambling to get home
to my wife, kids, invalid mother, and in-laws: all suddenly poten-
tially in the path of the just reinvigorated Thomas Fire, which
suddenly, after days of laying low, erupted early Friday morning
into a cacophony of flame. We are not supposed to have forest
fires in winter, our wet season. To keep the fire from burning Santa
Barbara, firefighters have driven it up into the hills, and due west
along the ridge, right toward my house. Crap. In hours, despite a
virtual army of firefighters, bulldozers, and fire engines, driven by
70 mile-an-hour winds, very high fuel loads, critically low fuel
moistures, and single-digit relative humidities, the Thomas Fire
jumped miles to the west. Now directly behind Santa Barbara, it
sent fingers of flame and burning coals questing down into the
edges of Montecito and the Riviera, right by my in-laws’ house.
A few days ago I had made a difficult – and poor – choice,
deciding to put business ahead of family, attending a science meet-
ing in New Orleans, thinking that the risk of fire had abated. Boy,
was I wrong, and wow was my wife rightly pissed off. Here’s the
transcript from my cell phone group chat from that morning. Sabina
is my wife. Yesi and Matt are a couple, and good friends. She is a
volunteer firefighter. Anneli and Nicholas are also good friends. We
all live in the mountains behind Santa Barbara, and have kids about
the same age. Rochelle is Sabina’s mom. Ami is my daughter.
Around 6:30 AM PST, December 16, 2017
Yesi : 8369 firefighters, 1,102 fire engines,

78 dozers, 32 helicopters.
Sabina : And the helicopters ! Sikorskys,
skelecopters, big ones little ones and all
colors . I want to collect them all .
Annili : Wow
Yesi : Sab – as the crow flies, the fire edge is
about 3.4 miles away from your Dad’s
house ( ☹)
Sab : That’s what it looks like ! Watching them
drop – it’s continuous and mesmerizing.
Matt : Reminds me of the gap fire . We were
watching the helos from Camino Rio
Verde as the were dropping directly
north of our house ( ☹)
Sab : Opening this channel yes – and including
my folks.
Anneli : How is air
Sab : (pastes picture of horrible cloud of
smoke). air is great now but take a look
at this! that will crash down on
us soon.
Me : Wow . I have my fingers crossed for
everyone.
Nicholas : Is the picture taken from the Riviera?
Sab : Pict is from our house.
Yesi : Yes . It’s not good . I have a mtg with
Trout Club crew at 8:30. I’ll let you all
know . These winds are supposed to be up
all day.
Me : omg.
Now about 8:00 AM PST
Yesi: Jumped San Ysidro Creek now. Montecito winds are

29 miles per hour, gusting to 70.
Nicholas: Stay out of harms way.
Yesi: Mandatory evac for east of highway 154 to Mission
Canyon.
Sab: (pictures) Montecito Peak burning.
Matt: Scary
Me: Very
Nicholas: Where are you Sabina?
Me: At her parents house I think.
Sab: I’m fine on way to Trout Club.
Rochellle: Is Sabina safe?

Matt: Yes, Yesi is talking with her now. We’re going to head
off the mountain.
Yesi : We are taking Ami to my father-in-laws in Hope
Ranch. We are only volunteer evacuation, but I’d
rather they all be safe. I need to be with the volunteer
firefighters.
Me : I’ve rebooked my flight to DC to Los Angeles.
Will get into LAX about 9 and Santa Barbara
at 11-12.
Now I’m aboard United flight 430. I use the onboard WiFi to
download the 12:00 AM Mountain Standard Time MODIS fire
imagery.
Me – via e-mail from the plane:
Looking at latest MODIS Fire imagery - low connectivity,

but the fire seems to have reached Casa de Herrera in
Montecito, and jumped west of the Montecito fire depart-
ment. The museum in the center left is the Museum of
Natural History.
As freaked out as I was, it was nothing compared to my wife
Sabina. Here’s the picture she took from the deck of her parents’
house that afternoon (Figure 4.12).
A week earlier, on Saturday, December 9, I drove from my
house (“a” on the map) up along East Camino Cielo. East
Camino Cielo is a beautiful, winding skinny road that rests right
on top of the transverse range behind Santa Barbara. It offers
stunning views of the Pacific Ocean to the south and the Santa
Ynez river valley to the north. After about an hour’s drive, I made
it past Montecito, to about the place marked “b” on the map
below. This figure shows the official evacuation zones on
December 6, 10, and 16. On December 8, as I stared eastward
down the transverse range, I could only see distant rising smoke
clouds to the east and north of Carpentaria. Then, on December
10, a second flare-up occurred, and the fire spread rapidly to the
west. I have lived in Santa Barbara for a long time, and the way
this fire moved was beyond startling – in aperiodic cataclysmic
waves. On December 5 and December 10, the fire expanded by
Figure 4.12 This photo was taken on December 16, 2017, at

about 3 PM PST from near the Gibraltar Road and East
Camino Cielo intersection. It is looking east toward the
community that sits near the top of Gibraltar Road. Image
credit: Ryan Cullom/RLC Photography
more than 60,000 acres (each day).24 Sixty thousand acres is

94 square miles – an area larger than Boston. On December 10,
up in the hills, all hell was breaking loose as Santa Anna winds
picked up, combining with winds generated by the ferocity of the
fire itself.25 Towering pyrocumulous clouds, or flammagenitus,
formed above unstable columns of rising air. Capable of collaps-
ing at any time, these towers could send showers of sparks and
whipsaw winds hundreds of yards, posing a serious threat to
firefighters. My lovely in-laws, Maurizio and Rochelle
Barattucci, live in the foothills at the base of Santa Ynez
Mountains (“c” on the map). Saturday night (December 10) was
a horrid surreal experience as the fire, driven by 70 mile-an-hour
winds, suddenly expanded at an insane rate, eating an acre a
second. As I sat on a westbound plane watching satellite fire
24
upload.wikimedia.org/wikipedia/commons/a/af/2017_12_11–08.57.46.111-CST
.jpg.
25
en.wikipedia.org/wiki/Thomas_Fire.
Figure 4.13 Major advances of the Thomas Fire.
imagery from space, the fire first tore north up into the transverse
(east–west) mountain range north of Ventura. Then, in the wee
hours, the witching hours, about 3 AM, the wind changed again,
blowing from east to west. The fire quickly spread west, first
gobbling the chaparral covering the hillsides. The next to fall were
the nurseries, avocado orchards, and outlying homes near the city
of Carpentaria (Figure 4.13).
On December 11, though, the fire lines held. Forty miles away
from Ventura and twenty-five miles from Carpentaria, the ash fell
thick on our driveway, covering the cars, the porch, the sidewalks,
and the succulents. We wore masks. The crows cackled from the
broken oak branches. Their cries echoed across a gray mezzotint
landscape drawn by Edgar Allen Poe. Then again, on December
16, the fire expanded rapidly again. Amazingly, a literal army of
firefighters managed to protect Montecito and Santa Barbara.
TEMPERATURE EXTREMES – IMPACTS
5 AND ATTRIBUTION
Shocks, Exposure, and Vulnerability
Our planet is warming rapidly. Global warming is

already hurting people, contributing to widespread increases in
health risks associated with extreme temperatures.1 The impacts
of climate change are arising through the potentially explosive
interaction of weather shocks, exposure, and vulnerability
(Figure 5.1). The shock-exposure-vulnerability construct is a
common framework for understanding the impact of extremes.
Shocks represent weather-related extreme events. Exposure quan-
tifies the potential losses associated with these events. Population,
economic investment, and ecosystem productivity are commonly
associated with increased exposure. Vulnerability represents a
lack of capacity to resist negative impacts. Almost everywhere
we see very rapid increases in impacts due to multiplicative
combinations of expanding shocks and exposure. More frequent
heat waves and precipitation extremes impact rapidly expanding
populations. More expansive fires threaten growing rural and
semi-rural settlements. Cyclones and tidal surges inundate
1
Watts, Nick, et al. “The 2019 report of The Lancet Countdown on health and
climate change: ensuring that the health of a child born today is not defined by a
changing climate.” The Lancet 394.10211 (2019): 1836–1878.www.thelancet
.com/journals/lancet/article/PIIS0140-6736(19)32596-6/fulltext.
93 / Temperature Extremes – Impacts and Attribution
Figure 5.1 Dimensions of hazards or risks associated with

weather and climate extremes. Impacts arise through the
interaction of three distinct factors: the magnitude and frequency
of climate and weather shocks, the underlying level of exposure
in the human and natural systems experiencing those shocks,
and the underlying vulnerability of these systems.
increasingly populated, valuable, and vulnerable coastline prop-

erties. And in each of these cases, the poor and vulnerable tend to
bear the brunt. Adopting a card game analogy, weather deals the
cards, exposure measures how much you have to lose, and vul-
nerability quantifies how well you can absorb that loss.
Of all weather hazards, heat waves tend to be the most
immediate, and potentially the most deadly. All living organisms
must maintain subtle balances with their surrounding ecosys-
tem. Extreme temperatures, especially under humid conditions,
can lead to heat stress and heatstroke, acute kidney injury, and
dramatically increase the chance of dying due to heart problems,
especially for older people.2 Heatstrokes occur when rising
internal body temperatures impact your nervous system. When
the air surrounding your body reaches extreme levels, you may
2
Székely M., Carletto L., and Garami A. “The pathophysiology of heat exposure.”
Temperature (Austin) 2 (2015): 452.
not be able to sweat fast enough to cool down. In response,

seeking to maintain a balanced temperature, your body moves
more blood toward your skin, away from your internal organs.
This can damage these organs while also putting strain on your
heart, potentially leading to heart problems. The brain receives
less oxygen. More toxins enter your bloodstream and kidneys.
Cities, which are often warmer than the surrounding
countryside, can expose millions of people to heat waves. The
elderly and children are generally much more vulnerable. In
particular, during periods of extreme heat, young children have
a greater risk of electrolyte imbalance, fever, respiratory disease,
and kidney disease.3 Extreme temperatures can also increase the
chance of interpersonal violence,4 collective civil conflict,5 and
dramatically decrease worker productivity, which can nega-
tively impact national-level productivity. Extreme temperatures
can stunt childhood growth and stress pregnant mothers and
their babies, contributing to increased frequencies of low-
birthweight babies that often have serious health problems. In
crop-growing regions, extreme temperatures can increase water
demand, alter growth rates, and directly damage plants, espe-
cially during the critical germination and grain-filling periods in
the middle of a growing season. In the ocean, marine heat waves
pose severe threats to coral reefs, fisheries, and the human and
nonhuman populations that depend on them.
Many warm regions of the globe are densely populated,
resulting in high levels of exposure to heat extremes. Historically,
many of the most ancient civilizations centered around hot, low
subtropical river basins: the Nile, Tigris-Euphrates, Indus, and
Ganges River systems helped give birth to humanity by providing
stable sources of water. Terrestrial plants and oceanic phyto-
plankton are light-absorbing life-making miracle machines,
3
Xu Z., Sheffield P. E., Su H., Wang X., Bi Y., and Tong S. “The impact of heat
waves on children’s health: a systematic review.” International Journal of
Biometeorology. 58 (2014): 239–247.
4
Sanz-Barbero B., Linares C., Vives-Cases C., González J. L., López-Ossorio J. J.,
and Díaz J. “Heat wave and the risk of intimate partner violence.” Science of the
Total Environment. 644 (2018): 413–419.
5
Levy B. S., Sidel V. W., and Patz J. A. “Climate change and collective violence.”
Annual Review of Public Health. 38 (2017): 241–257.
constantly transforming carbon dioxide, water, and incoming

solar radiation into sugars and carbohydrates. Given abundant
water and carbon dioxide, this transformation process tends to be
the most productive where and when we receive the most
incoming solar radiation. This is also where and when tem-
peratures tend to be the warmest. So, on land and in the
oceans, many of the most biologically productive tropical and
subtropical regions face high levels of risk from rising tempera-
tures. These risks may impact the organisms absorbing the
most carbon dioxide from the atmosphere, creating potential
“positive” climate change feedbacks that can help drive the bus
even faster over the cliff. Many of the warmest regions, like
topical forests or coral reefs, team with life that helps sequester
carbon dioxide. Yet these very warm regions can also be dra-
matically damaged by modest increases in temperatures.
Focusing on human impacts, we will see that billions of
humans face annual threats associated with the north–south
passage of the summer monsoons and the areas of maximum of
incoming solar radiation.
Small Changes Can Dramatically Increase

the Chance of Extremes
Before going forward, we need to discuss further the
interaction between weather and climate extremes and hazards.
Weather and climate extremes are either short-term (weather) or
long-term (climate) anomalies that are “extreme” based on the
historical behavior of some phenomena. For example, in early
November 2019, the New South Wales region of Australia and
the city of Sydney faced catastrophic fire danger fueled by very
poor rains over the last twelve months, exceedingly rapid off-
shore winds blowing from the interior of the continent toward
the Tasmanian Sea, and exceptionally warm air temperatures (see
sidebar). High winds and warm temperatures qualify as extreme
events, and their combination can be exceptionally dangerous.
The potential impact of these weather conditions can be exacer-
bated by lower-frequency climate conditions, such as a year’s
worth of very warm air temperature and very low rainfall.
Climate scientists often use “percentiles” to describe

weather and climate extremes. At a given location and time, a
weather variable, such as daily maximum air temperatures, will
have a distribution of possible values. A bell curve–like normal
distribution is one common distribution that many people are
familiar with. The “percentile” values within this distribution range
from the lowest possible (the 0th percentile value) to the greatest
possible (the 100th percentile value). A percentile value of 50 would
correspond to the median (i.e., middle) value; half the observed
values would be expected to be above and below this value. When
dealing with extremes, we often deal with values like the 99th
percentile maximum air temperature. If we measure each day’s
warmest air temperature for a year, and sort the 365 warmest days,
then the average of the third and fourth warmest days would
correspond to the 99th percentile value for that distribution.
When discussing climate change, we often examine changes in the
values associated with a given percentile. Alternately, we can pick a
threshold (say, a temperature threshold of 40.6C/105.1F) and
look at changes in the frequency of values exceeding this threshold.
To understand the influence of climate change on
extremes requires an understanding of both the Earth’s sea-
sonal cycle and how small changes in average conditions can
translate into big changes in weather extremes. When you
watch or read about this month’s extreme temperatures, think
about how the locations of these extremes align with the pass-
ing seasons. Every year the latitude of maximum incoming
solar radiation moves north and south with the tilting of our
planet in its annual rotation about the Sun. Areas near the
equator receive the most radiation, but also tend to be cloudy.
These clouds reflect the sunlight back into space. So the
warmest maximum daily air temperatures tend to be located
in the subtropics, between about 17 and 37 north and south,
in the summer hemisphere. During summer in the northern
hemisphere, we find that the warmest (99th percentile or
greater) air temperatures have very warm average values of
~44C (111.2F). As the point of maximum incoming solar
radiation dips south of the equator, a second band of max-
imum heating arises in the southern hemisphere.
Figure 5.2 Normal distribution curves for a hot location

on Earth.
Putting these concepts together, we can imagine that a

typical very warm region in the northern or southern subtropics
might have a mean value of ~42C for the warmest (99th per-
centile) days in a year. Typical values ranging about this mean
may go from about ~39.5C to 46.5C (103.1 –115.7F). The
warmer half of this distribution is shown in Figure 5.2. The most
probable value (42C) is shown on the left, and then the prob-
ability curve drops rapidly. The shaded (black) region on this
curve corresponds to the 1 percent of distribution corresponding
to values exceeding the 99th threshold (46.3C). Now we add
2C and redraw the corresponding bell curve. The chance of
exceeding the 99th percentile value of 46.3C jumps 23-fold.
A relatively small change in the mean can result in a very large
increase in the number of extreme events.
How Extreme Have Temperatures Been in

2015–2019?
The amount of daily station data in many parts of the
world is quite limited. But we can use the NASA Goddard Space
Flight Center’s 1880–2019 monthly temperature data set to
Fraction of Earth with 99th > percentile T [%]

0.20 37.0
Fraction of Earth that is exceptionally warm [%]
GHCNDEX Max Tmax 36.5
GHCNDEX Max Tmax [C]

0.15
36.0
35.5
0.10
35.0
34.5
0.05
34.0
0.00 33.5
1950 1960 1970 1980 1990 2000 2010
Figure 5.3 Time series of the fraction of the Earth with

exceptionally warm temperatures (solid black line) and globally
averaged maximum warmest-day temperatures (dashed line).
Circles identify the 2015–2019 period.
explore the changes in temperature extremes.6 This data set has

sea surface temperature estimates over the ocean and two-meter
air temperature estimates over land (both measured in degrees
Celsius). This data set is routinely used to assess global tempera-
tures. The 2015–2019 annual global land-sea temperature
anomalies were each warmer than all the previously observed
values. To track areas with exceptionally warm conditions, for
each month and for each year we can identify areas with ocean
or air temperatures in the top 1 percent of the 1880–2019
record. Then for each year, we can translate this information
into a fraction between 0 and 1, corresponding to the fraction of
the Earth with temperatures in the top 1 percent of the historical
record. The black line in Figure 5.3 displays this data. Black
circles denote values for our years of interest.
This time series should deeply disturb you. The fraction
of the Earth with exceptionally warm ocean or air temperatures
is increasing very rapidly. When I was born, it was about
2 percent. By the time I started graduate school, it was around
5 percent. When my children were born, it hovered around
6
data.giss.nasa.gov/gistemp/.
7 percent. Since that time the area of exceptional warmth has

doubled again to around 17 percent, increasing rapidly just
between the beginning and end of our focus period
(2015–2019). Now almost one-fifth of the globe experiences
extreme temperatures at any given time. Extreme temperatures
are rapidly becoming much more frequent. According to
Maximiliano Herrera, a Costa Rican climatologist dedicated
to rigorously tracking global temperature extremes using station
data,7 the 2015–2019 period has brought record-breaking tem-
peratures at 384, 337, 201, 430, and 619 weather station loca-
tions, year to year.
These temperature extremes are having severe health
impacts. Between 2015 and 2019, the Emergency Events
Database (EM-DAT),8 maintained by the Centre for Research
on the Epidemiology of Disasters, identified 71 extreme-
temperature disasters, related to 9,116 deaths, 90,014 injuries,
affecting 4.5 million people and resulting in $1.8 billion (€1.64
billion) in damages. According to EM-DAT, in 2015 tempera-
ture extremes in France, India, and Pakistan claimed 3,275,
2,248, and 1,229 lives, respectively. In 2016, 2017, and 2018,
extreme heat waves in Morocco affected 750,000, 1,650,000,
and 70,000 people, respectively. In 2018, extreme heat waves
injured 49,000 people in Japan. In 2016, a heat wave in China
resulted in $1.6 billion dollars in losses. In 2019, 173 people
died in Japan and 112 people died in India due to extreme
temperature–related causes.
As we now routinely experience, the impact of this
warming often follows the maxima of solar insolation. During
the northern hemisphere summer of 2018, extreme temperatures
meant that drought and wildfires threatened Europe. In January
2019, extreme heat stretched across the southern hemisphere,
and forty-four fires, fueled by extremely hot and dry conditions,
burned for weeks across Tasmania.9 In June and July 2019,
Europe experienced record-breaking heat waves, and extremes
7 8
www.mherrera.org/temp.htm. www.emdat.be.
9
earthobservatory.nasa.gov/images/144486/fortnight-fires-in-tasmania.
were linked to some 1,800 deaths in France.10 In May and June

2019,11 a severe heat wave struck India and Pakistan.
Temperatures in northern India, in Rajasthan, reached 50.8C
(123.4F). Consulting India disaster statistics in EM-DAT, the
international emergency database, reveals this “extreme” to be
part of a pattern of the new normal. According to EM-DAT,
112 deaths were associated with the 2019 heat wave. Heat
waves were linked to 264 deaths in 2017 and 300 deaths in
2016. And in 2015, a terrible heat wave was tied to 2,248
deaths. As the locus of maximum solar radiation shifted south,
Australia faced, yet-again, record-breaking temperatures and
widespread wildfires (sidebar).
Using daily weather station data, we can also calculate
the maximum daily temperatures at each weather station loca-
tion, and then average these values over the globe. Several
groups of researchers are committed to the important task of
monitoring these types of climate extremes, and we can access
many of their valuable datasets online at www.climdex.org/.
The dotted line in Figure 5.3 shows estimates of global
averaged Global Historical Climate Network Index
(GHCNDEX)12 annual maximum temperatures. This dotted
line may be even more concerning than the solid black line in
Figure 5.3. Between the 2000s and our 2015–2019 period of
interest, the maximum terrestrial air temperature values appear
to have increased by something like +1.5C. Daily temperature
maximums tend to occur in the afternoon, as the daily heating
from the Sun accumulates. Seasonally, these maxima tend to
occur in the late spring or summer, when again, the heating by
the Sun tends to reach maximum values. In just a few pages, we
will provide model-based evidence suggesting that this trend
will continue. Such increases in extreme temperatures will be
very dangerous for people, crops, forests and livestock. Such
10
www.bbc.com/news/world-europe-49628275.
11
www.thelancet.com/journals/lancet/article/PIIS0140-6736(19)32596-6/fulltext.
12
Donat, M. G., L. V. Alexander, H. Yang, I. Durre, R. Vose, and J. Caesar.
“Global land-based datasets for monitoring climatic extremes.” Bulletin of the
American Meteorological Society. 94 (2013): 997–1006.
warming will shift the distributions of extreme events to the

right, as illustrated in Figure 5.2.
Quantifying Increases in Exposure to Extreme

Air Temperatures
How many people may be exposed to exceptionally
warm temperature extremes? This important question can be
difficult to address with rigor, especially because accessible
weather station data tends to be very sparse in many countries.
To help overcome this limitation, my research group has
developed a new maximum temperature product13 that uses a
long time series of geostationary weather satellite observations14
and Berkeley Earth15 weather station data to provide accurate
estimates of maximum air temperatures. These estimates are
particularly well suited to monitoring conditions in places like
Africa and India, regions where vulnerability is high, exposed
populations are large and rapidly increasing, and typical
extreme temperatures are already very high. As context, we
can examine a map (Figure 5.4) showing the mean daily 99th
percentile air temperatures, based on a new dataset I have
created with my team at the University of California Santa
Barbara. This new high resolution (0.05 degree ~ 5 km2) data
set is referred to as the Climate Hazards Center Infrared
Temperature with Stations Tmax (CHIRTSmax) dataset. At pre-
sent it extends from 1983 to 2016 and is in the process of being
updated. It benefits from lots of station data augmented by long
continuous set of satellite-based observations of the Earth’s
surface emission temperatures. This makes it particularly well-
13
chc.ucsb.edu/data/chirtsmonthly, Funk, Chris, et al. “A high-resolution 1983–
2016 T max climate data record based on infrared temperatures and stations by
the Climate Hazard Center.” Journal of Climate 32.17 (2019): 5639–5658.
https://www.chc.ucsb.edu/data/chirtsdaily. Verdin, Andrew, et al. “Development
and validation of the CHIRTS-daily quasi-global high-resolution daily
temperature data set”, Scientific Data, 7.30 (2020). https://www.nature.com/
articles/s41597-020-00643-7.
14 15
www.ncdc.noaa.gov/gridsat/ berkeleyearth.org/data/
Figure 5.4 Population-weighted averages of the warmest 20 days

of each year. Based on the CHIRTSmax dataset and
2020 Gridded Population of the World population estimates.
suited to look at changes in data-sparse areas – those areas

where people also tend to be highly vulnerable.
Broadly following the approach taken in a recent Lancet
article on the increasing extreme heat risks, we can use the daily
CHIRTSmax data set and gridded population estimates to
explore the effects of changes in air temperatures during each
year’s warmest days (Figure 5.4). Before even asking what this
time series means, linger for a moment considering what it
involves. The population data involves an incredible effort
involving hundreds of national censuses, remote sensing, and
sophisticated integration and modeling. Swap in weather station
observations for census estimates, and the same statement holds
for air temperature estimates. These datasets, compressing hun-
dreds of lives worth of human effort into collected sets of infor-
mation, provide the basis for ‘science haikus’ like Figure 5.4.
More such figures will follow.
Figure 5.4 tells us it is getting a lot hotter quickly in

places where people are living, during the hottest ~5 percent of
the year (~20 days). A simple trend line associated with this data
explains most (64 percent) of the year-to-year variability. This
trend (about 0.3C per decade), works out to a ~+1.2C increase
between 1980 and 2020. That is a lot of warming. Extending
this trend to 2030 or 2040, we would find additional tempera-
ture increases of +1.5 and 1.8C. These outcomes are marked
with gray polygons.
The current mean of the warm temperatures is around

36 C, but there are many places that are much warmer. And
many of these places are heavily populated, tropical or subtrop-
ical, and at relatively low elevations – the perfect recipe for high
levels of exposure combined with frequent extreme heat waves
Many of these regions are historic river-based cradles of civil-
ization. The Niger River flows east across western Africa,
through Guinea, Mali, Niger and Nigeria. The Nile flows north
from Ethiopia and Uganda through South Sudan, Sudan and
Egypt. The Tigris-Euphrates river system that begins in the
Zagros Mountains of eastern Turkey flows through Syria and
Iraq on its way to the Persian Gulf. The Indus River flows south
from the Hindukush mountain range through Pakistan, while
the Ganges River system flows east across northern India to
drain into the Bay of Bengal. For thousands and thousands of
years, generation after generation, humans have lived in these
hot valleys, giving birth to civilization along the way. In 2019,
2.3 billion people lived in these countries, almost one out of
every three humans. By 2050, United Nation’s population pro-
jections anticipate an additional billion people in these regions,
a 50 percent increase. Those 3.3 billion people will be exposed
to much warmer extreme heat waves.
To help visualize the billions of people in harm’s way,
we can examine a simple map of daily 2000–2016 99th percent-
ile maximum temperature values (Figure 5.5). In many parts of
India and Pakistan, temperatures reach incredible levels,
exceeding 44C (111F), and reaching as high as 47C (117F).
These very warm regions tend to align with river basins, like the
Figure 5.5 Map of 99th percentile 1983–2016 air temperatures

for 2000–2016 and 2050 (right), based on a ‘business as usual’
climate change change scenario based on the 8.5 Wm 2
Representative Concentration Pathway scenario
Indus and Ganges watersheds, which are the most densely popu-
lated locations on the planet. In between these watersheds we
find higher elevations and cooler temperatures. The United
Nation projects that India and Pakistan’s 2050 population will
be two billion. Boxes around the Indus Basin in Pakistan and the
Indo-Gangetic plain in northern India highlight these high risk
regions. Climate change projections (discussed below) indicate
imminent increases in mean temperatures that will place these
extremely large populations in great peril.
We can quantify the risks associated with heat waves by

using the high resolution daily CHIRTSmax data to identify heat
waves, and then counting the number of people exposed to these
extremes. This provides a numerical measure of the interaction
of extreme weather and exposure, two of the main dimensions
of climate hazards (Figure 5.1). There are many approaches to
identifying heat waves. Most definitions require temperatures to
exceed a given threshold for several days (often three or more).
Persistent warm conditions are much more harmful biologically.
Thresholds may be based on daily minimum or daily maximum
temperatures. Daily minimum temperatures occur at night, and
very high nighttime temperatures reduce the body’s ability to
recover from heat stress. It is also common to use measures of
“apparent” temperature that take into account the effect of
atmospheric humidity. Higher humidity levels make it harder
for humans to cool their bodies through perspiration. Here we
will use a very simple definition of a heat wave as a three day
period in which the temperature for each day is warmer than
40.6C (105.1F). Using this threshold we can calculate the
number of heat waves at each location in each year between
1983 and 2016. Using current gridded population estimates, we
can calculate the population weighted average of these counts.
This time series of global heat wave counts (not shown) exhibits
a strong upward trend of ~0.5 events per decade, increasing by
about 40 percent between the early 1980s and late 2010s, from
four to about five-and-a-half events per year.
To estimate exposure to heat wave risks, we can com-
bine our heat wave counts with gridded annual population
estimates for every year between 2000 and 2016 (Figure 5.6).
These results are very concerning, and indicate very large
increases in the number of people exposed to heat extremes.
Our analysis begins in 2000 because gridded population esti-
mates prior to that year are less reliable. For each year, we
multiply the number of heat wave events at a given location
by the population in that location. The result of this multipli-
cation produces values in units of “people-events,” capturing
the interaction between population growth and increases in
Figure 5.6 Time series of annual global estimates of total heat

wave exposure (gray circles). The dashed line shows just the
estimated exposure values associated with variations
in population.
air temperatures. Then for each year we add up the total

people-events for the planet. The time series marked with grey
circles in Figure 5.6 indicates very rapid increases in people-
events. A trend of ~7.2 billion people-events per decade
means that a relatively modest shift in extreme temperatures
(Figure 5.4) is substantially increasing the frequency of very
warm events (depicted schematically in Figure 5.2), and inter-
acting with population growth to create billions more expos-
ure events every decade.
By fixing the population at 2000 levels we can estimate
the temperature component and remove it, isolating the
population-driven component of the increases in exposure
people-events. The population component is identified by the
dashed line in Figure 5.6. If the total number of people-events

increased by around 14 billion since 2020, population growth
alone accounted for about 8 billion while increases in maximum
air temperatures accounted for about 6 billion. The interaction
between atmospheric warming and population increases is
already creating dramatic expansions in human exposure.
These trends will absolutely continue, for at least the near-
term future.
Both the climate and human socioeconomic systems
exhibit strong levels of persistence. Over the next several
decades we are certain to see continued warming and popula-
tion growth. Greenhouse gasses build up in the atmosphere and
persist for many years, while the ocean continues to store more
and more thermal energy. At present there are almost a billion
young women on the planet,16 and most of these women will
become mothers, so demographers anticipate rapid population
growth. And progress toward reducing greenhouse gas emis-
sions remains very slow. If the observed trend in heat exposure
continues, we might see a 2000–2030 increase in exposure of
24 billion people-events, and 2040 increases exceeding 30 billion
people-events a year. Many more people will experience many
more extreme heat waves over the next two decades. These heat
waves are already having serious impacts on health, especially
for children, pregnant mothers, and the elderly. Dangerous
climate change is not a “what if” but a “happening now” with
“a lot more coming soon.”
Can We Attribute a Portion of These Impacts

to Climate Change?
Attribution science involves analyzing the root causes of
phenomena, such as human diseases or extreme weather pat-
terns. Such analysis can lead to prediction and prevention. For
16
social.un.org/youthyear/docs/fact-sheet-girl-youngwomen.pdf.
example, in 1854, the British physician John Snow correctly

attributed the outbreak of cholera in London to a contaminated
pump at the intersection of Broad and Cambridge streets.
Removing the handle to this pump stopped the epidemic.
Motivated by a similar desire to predict and protect, meteorolo-
gists have spent decades exploring the mechanisms that cause
weather and climate extremes. Unlike medical researchers, who
can often use structured medical trials complete with placebos
and control groups, Earth scientists face severe experimental
limitations. We only have one world, and can’t run experiments
on hurricanes and heat waves. Extreme events, furthermore, are
by definition rare, and typically arise through complex non-
linear interactions. Yet as described in a recent report by the
United States National Academy of Sciences17: “Effective, rigor-
ous, and scientifically defensible analysis of the attribution of
extreme weather events to changes in the climate system not
only helps satisfy the public’s desire to know but also can
provide valuable information about the future risks of such
events to emergency managers, regional planners, and policy
makers at all levels of government.” The new science of extreme
weather and climate event attribution seeks to provide
such analysis.
Framing attribution questions correctly is a critical first
step; we need to avoid the question “did climate change cause
X,” where X is an extreme event. Climate change alters the
amount of heat in the atmosphere, land, and oceans, the distri-
bution of greenhouse gasses and aerosols in our atmosphere,
sea levels, sea ice, and ocean acidity. These changes can modify
the frequency or intensity of weather and climate change
events, but are not typically the primary cause of anything.
To return to our analogy of playing cards, climate change
might shift the distribution of cards, making some hands more
likely than others.
17
www.nap.edu/catalog/21852/attribution-of-extreme-weather-events-in-the-
context-of-climate-change.
Understanding the influence of such distribution

changes on temperature extremes is relatively straightfor-
ward. Of the range of weather and climate extremes (e.g.,
droughts, floods, fires, hailstorms, tornadoes, hurricanes,
etc.), the influences of climate change on extreme heat events
are the best understood, and we have the highest level of
confidence when examining these types of events. This confi-
dence arises from both the relative simplicity of extreme heat
events as well as our ability to model these influences. For
example, we are all familiar with the idea that weather
systems are produced by the movement of high- and low-
pressure systems. High-pressure cells typically produce sub-
sidence (sinking motions) and cloudless skies that allow more
sunshine to reach the Earth. These and other naturally occur-
ring factors are represented well by our current generation of
climate models, and our models do a good job of predicting
day-to-day and decade-to-decade variations in air tempera-
ture. We can contrast this predictability with wildfires and
cyclones, which arise through very complicated small-scale
and highly nonlinear interactions. Both our theoretical under-
standing and our ability to model and predict these highly
complex systems is limited. Hence, it is much easier to under-
stand how one degree of warming might change the distribu-
tion of extremely warm days than how one degree of
warming might change the number, extent, or intensity of
wildfires or hurricanes.
By combining our observed high-resolution temperature
and population data with climate model–based estimates of
temperature changes, we can explore two important questions:
(1) What would temperature exposure levels look like in a world
without climate change? (2) What will likely future temperature
exposure levels look like in a world with climate change? We
address both of these questions using a climate analysis tech-
nique called “perturbation.” We start with a huge amount of
data that describes the world as we know it over the 2000–2016
time period, the data summarized in Figure 5.6. Then we per-
turb the observed temperatures by model-based temperature
deviations, based on large collections (ensembles) of climate

change simulations.18
Climate models describe the physics controlling the
motions of the atmosphere, ocean, and land. These model com-
ponents influence each other. Today’s wind patterns and cloud
distributions will subtly affect tomorrow’s ocean temperatures.
These ocean temperatures will then influence future weather
patterns. And so on. Day in and day out, year after year,
these models can be used to simulate our weather and climate
with a pretty exciting degree of realism, thanks to decades of
careful improvements made by a global collection of modeling
centers. Massive efforts in support of initiatives like the
Intergovernmental Panel on Climate Change (IPCC) have gen-
erated petabytes of simulations that can be used to assess the
influence of climate change.
One great advantage of model simulations is that scien-
tists can alter the components that control the Earth’s “radiative
forcing.” Radiative forcing refers to those aspects of Earth–Sun
system that control our basic radiative balance. There are import-
ant natural factors such as the magnitude of incoming solar
radiation (modulated by Sun spots on short time scales) and the
cooling effect of volcanic eruptions. There are also important
anthropogenic (human-caused) variations such as the distribution
of greenhouse gasses and aerosols, and alterations to the Earth’s
land surface. Greenhouse gasses and aerosols change how energy
moves through the atmosphere. Alterations to the Earth’s surface
can alter how much solar radiation gets reflected back into space.
Here we explore three sets of climate change simulations repre-
senting different global warming. One set, based on the 4.5 W/m2
Representative Concentration Pathway (RCP4.5), represents a
future in which we effectively reduce our carbon emissions.
A second set, RCP8.5, represents a future in which we continue
18
These results are based on a large set of simulations from thirty-eight to forty-one
different models obtained from the KNMI climate explorer: climexp.knmi.nl/
start.cgi. Temperature changes were based on the modeled changes in the
warmest month of the year.
+2C
+2C
Figure 5.7 Time-series of population-weighted averages of

changes in modeled daily maximum temperatures. The time
series are expressed as changes from the 2000–2016 time period.
to emit large amounts of greenhouse gasses. A third set, RCP6.0

represents an intermediate warming scenario. The RCP scenarios
(described in Chapter 13) correspond to different projections of
future emissions.
Figure 5.7 presents the population-weighted average
change in daily maximum temperatures, during the warmest
month of the year, for RCP4.5, RCP6.0, and RCP8.5 scenarios.
These results are presented as changes from a 2000–2016 base-
line period, chosen to correspond to our heat wave analysis
(from Figure 5.6). In addition to future warming projections,
the models have also been forced with historical radiative
forcing conditions, so the time series begin in 1861. For each
year, we have multiple estimates from approximately forty
models (it varies a little by RCP). Figure 5.7 shows the average
of these simulations.
There are four important things to be learned from this
figure. First, the simulations suggest that we have already
experienced about 1C of human-induced warming. Second, no

matter what, we are very likely to experience about another
degree of warming by 2050. Even if we adopted vigorous reduc-
tions in emissions now, the persistent effect of our accumulated
greenhouse gasses will continue to warm the Earth for the next
thirty years. But (third very important point) even a modest
effort to mitigate our emissions (corresponding to the RCP6.0
scenario) will reduce the amount of 2050 warming by about
half, limiting it to around 1C. On the other hand, continuing
on our current rapid (RPC8.5-like) emissions trajectory will see
our children mature in a world likely to be two degrees warmer
than present conditions. Which leads us to point four. The
RCP8.5 scenario is estimated to warm the world inhabited by
our childrens’ children by 4C. Committing ourselves to mitiga-
tion now can head of this catastrophic increase.
What might life have been like in a world without climate
change? We have the tools to answer this question quite precisely,
at least for relatively simple phenomena like temperature
extremes. We can estimate the spatial pattern of human-induced
warming, remove it from the observational record, and estimate
the number of people exposed to extreme heat events. This allows
us to attribute a proportion of the observed extreme heat waves to
climate change. We can perform this attribution by perturbing our
2000–2016 temperature data, adjusting it to align with a world
without human-induced warming, and then recalculating our risk
metrics. In this “counterfactual” experiment we construct a world
that might have been by taking the observed 2000–2016 tempera-
ture observations and cooling them by the difference, at each
location, between the 1861–1900 and 2000–2016 temperatures
simulated by our climate change models. This gives us a reason-
able approximation of a world like today but without the influ-
ence of human-induced warming. We can then recalculate the
frequency of extreme heat waves and the billions of people-events
experienced, on average, between 2000 and 2016. The bottom left
star in Figure 5.8 shows these results.
The answer we find implies that climate change has
nearly doubled the number of extreme heat waves. Between
People-Counts of 3-day heat waves [billion people-events]

150 2080 RP
C8
projected .5--
pop
2080 RC
P4
100 projected .5--
pop
2050 RC
P8
projected .5-- 2050 RC
pop P8
2050 RC current p .5--
P4.5-- op
projected
pop
50
--2080
--2050 RRCP4.5, curren
CP8.5, cu t pop
--2050 R rrent pop
CP4.5, cu
rrent pop
--2000-2016 observations
--2000-2016 minus human-induced warming
0
0 5 10 15 20
Counts of 3-day heat waves [#]
Figure 5.8 Estimates of annual heat frequencies and the number

of heat wave people-events (#people x #heatwaves).
2000 and 2016, we observed about five three-day heat waves

per person. Subtracting an estimated human-induced warming
(about 1C), we found that we would expect to see about three
extreme heat waves per person per year. So, climate change has
increased the number of extreme heat waves by about 70 per-
cent. This means that climate change is negatively impacting
hundreds of millions of people now, enhancing the frequency
of severe health risks, increasing conflict, feeding climate migra-
tion, and reducing worker productivity. Now.
But “climate change” is just shorthand for “our collect-
ive actions.” We all are implicitly hurting, increasing conflict,
feeding crises, and reducing productivity. This may be a fourth
moral dimension of climate hazards. In addition to shocks,
exposure, and vulnerability (Figure 5.1), we may need to include
a dimension associated with climate change attribution – moral
culpability. In Chapter 4 we discussed how the incredibly rapid
rate of greenhouse gas increases (Figure 2.3) was produced an
incredibly rapid rate of global warming (Figures 4.6 and 4.7). In
the warmest tropical oceans, marine heat waves are killing coral
reefs. In the warmest locations on land, atmospheric heat waves
are already killing thousands of people. There are billions of
people in harm’s way, and human biology is not going to be able
to evolve rapidly enough to keep the most vulnerable of these
populations from massive increases in exposure.
Explosive Interactions with Future

Population Growth
We can next use our projections to explore how changes
in exposure might increase in the future. The “2050 RCP4.5,
current population” projection, shown with a light gray star,
was calculated by warming our observed 2000–2016 climate by
the average modeled RCP4.5 increase. This would correspond
to the Scenario 1 in Figure 5.7 if our 2050 population remained
the same as today. The lower gray stars to the right of this value
would correspond to Scenarios 2–4. The RCP4.5 2080 world
(Scenario 3) has about the same amount of warming as RCP8.5
2050 (Scenario 2).
These results tell us that if we ignore the impacts of
population growth, and just examine the impacts of climate
change, we will still see massive increases in the amount of heat
exposure. Even the relatively modest 2050 RCP4.5 scenario
indicates that the frequency of extreme heat waves will almost
double by 2050. With unchecked emissions growth, we could
see an almost fourfold increase in the frequency of very extreme
heat waves by 2080.
Factoring in both climate change and a moderate popu-
lation growth scenario,19 we find huge potential increases in
exposure. These results are marked with dark stars. The most
modest outlook ‘RCP4.5 2050, projected population’ indicates
19
These population change estimates were based on gridded population projections
corresponding to Shared Socioeconomic Pathway 2. Jones, B., and B. C. O’Neill.
2017. Global Population Projection Grids Based on Shared Socioeconomic
Pathways (SSPs), 2010-2100. Palisades, NY: NASA Socioeconomic Data and
Applications Center (SEDAC). doi.org/10.7927/H4RF5S0P.
a massive increase in heatwave exposure, with the number of

events increasing from an observed 2000–2016 baseline of
about 20 billion to a 2050 baseline of about 70 billion people-
events a year. Even under a scenario of modest population and
emissions increase, we are likely to see very large increases in the
number of exposure events. By 2080, this same scenario identi-
fies another ~20 billion people-event increase, a magnitude of
exposure consistent with the total level experienced today. An
RCP8.5 scenario would be catastrophic. The number of heat
wave exposure events might increase by 670 percent.
Conclusion
Looking beyond the numbers, we need to extend our
imagination, employ our empathy. Billions are currently
exposed to extreme heat, in burgeoning cities, often without
recourse to air conditioning. As our planet warms, how many
will fear the Sun’s passage, knowing that it may bring remorse-
less heat to our hottest cities for days on end? Have you walked
the streets of Cairo, Khartoum, Niamey, Bangalore, Bhadgad,
or Hyderabad on a very hot summer day? How about Chicago’s
South Side, or one of the poor banlieues ringing Paris? Block
after block, the stands, markets, and apartments spread, each a
microcosm of human life. But increase the frequency of extreme
heat events, and these burgeoning cities may become extremely
hazardous, especially for the poor and elderly who can’t afford
air conditioning.
The results presented here are deeply concerning. They
suggest that the best we can realistically hope for is a doubling
in the number of extreme heat waves by 2050. If we adopt
rapid reductions consistent with RCP4.5, we may see about
nine extreme heat waves per person per year, compared to the
five events in recent years. This is three times the frequency in a
preindustrial world without climate change.
Fast forward yet another thirty years, and the seeds of
our decisions today will truly bear fruit, benign or no. If we
adopt a moderate mitigation strategy now, our RCP4.5
estimates indicate that by 2080, conditions will only get a little

worse, with nine heat waves per person in 2050 increasing to
eleven in 2080. But if we fail to mitigate, and follow the RCP8.5
pathway, we find a stunning increase to seventeen heat waves
per person per year, about three times the current frequency and
about six times the number in a world without climate change.
But this only begins to tell the story, because these
extremes are often happening in areas with rapidly growing
populations. When we factor in moderate population projec-
tions, we find explosive increases in the number of estimated
annual people-events (y-axis, Figure 5.8). By 2050, the optimis-
tic RCP4.5 scenario appears associated with a 300 percent
increase in exposure. The pessimistic RCP8.5 scenario indicates
almost a fourfold increase in people-events compared to
2000–2016, to a total of 81 billion people-events per year.
Going forward to 2080, population growth slows but still
continues to increase. Even the modest RCP4.5 scenario calls
for some 93 billion exposure events per person per year, and for
RCP8.5 our estimate is 150 billion. Currently, the overall level
of exposure (22 billion) is about three times the population of
the planet. By 2080, this level of exposure may be ten to four-
teen times greater. These are our children we are talking about.
Our bodies, especially the bodies of young children, preg-
nant mothers, in utero babies, and old people, experience severe
stress when exposed to consecutive days of very warm weather.
These extremes also feed conflict and substantially reduce worker
productivity. Here, in this chapter, we have used some of the best
available data to estimate the frequency of extremely warm days,
and the overall level of exposure, measured in people-events.
These observations are deeply concerning. Projecting into the
future, we find even greater reason for concern.
Sidebar – Examining Extreme Temperatures in India and Pakistan

Each year, the Bulletin of the American Meteorological Society
publishes a special issue presenting assessments of how human-
caused climate change may have affected the strength and likeli-
hood of individual extreme events.20 While a detailed assessment
of these special issues is beyond the scope of this book, describing
a single study can help elucidate how expert climate attribution
scientists do their work. To assess the influence of human-induced
climate change on extreme events, scientists borrow techniques
from epidemiology. In the broadest strokes, we can divide epi-
demiological analyses into situations where we can perform con-
trolled experiments and those where we cannot. For example, if
we are studying the efficacy of a pain medication, we might
perform a controlled experiment using a large sample of patients.
Some patients will take the real medicine. Some will take a pla-
cebo. Sometimes, however, we are morally prevented from such
experimentation. We would not want to perform such a study on
children facing a severe, life-threatening disease. In such a setting
we typically have to rely on using statistics. In climate attribution,
the first type of experiment is usually performed using climate
models. The second type of experiment is performed using
historical data.
The India 2015 heat wave attribution study by Drs. Michael
Wehner, Daithi Stone, Hari Krishnan, Krishna AchutaRao and
Federico Castillo21 provides a good example of both types of
analyses. In May and June 2015, severe heat waves struck New
Delhi, Allahabad, Jharsuguda, Hyderabad, and Khammam.22 In
New Delhi, a city with 30 million people, extreme temperatures
reached 45.5C (113.9F) and the city streets literally melted.23
A brutal heat wave, exacerbated by warm “loo” winds blowing
south from Pakistan, exposed hundreds of millions of people to
extreme heat stress. In the area around Hyderabad, five million
chickens perished, and farmers faced desiccated conditions.
20
www.ametsoc.org/ams/index.cfm/publications/bulletin-of-the-american-
meteorological-society-bams/explaining-extreme-events-from-a-climate-
perspective/.
21
Wehner, Michael, et al. “The deadly combination of heat and humidity in India
and Pakistan in summer 2015.” Bulletin of the American Meteorological Society
97.12 (2016): S81–S86. journals.ametsoc.org/doi/10.1175/BAMS-D-16-0145.1
22
en.wikipedia.org/wiki/2015_Indian_heat_wave.
23
www.telegraph.co.uk/news/worldnews/asia/india/11636124/Indias-extreme-
heat-wave-in-pictures.html.
Dr. Wehner et al.’s attribution study used both statistical analy-

sis and climate models to examine the potential role played by
climate change. They built a statistical model linking the amount
of atmospheric CO2 to the frequency of very warm days, showing
that increasing levels of CO2 were statistically related to increases
in the frequency of heat extremes. The study then used an atmos-
pheric model to study the frequency and magnitudes of extreme
events. This atmospheric model was similar to those used to
predict global weather patterns. The model was forced with
observed May 2015 sea surface temperatures, sea ice extent, and
atmospheric greenhouse gasses and aerosols. This process was
repeated ninety-eight times, to give ninety-eight independent
weather predictions. The chaotic nature of the weather ensured
their independence. The India analysis team then simulated “a
world that might have been” by removing estimates of human-
induced changes to sea surface temperatures, sea ice distributions,
greenhouse gasses, and aerosols. Such a simulated world is often
called a “counterfactual” experiment representing the climate
system had humans not altered the composition of the
atmosphere.
As with the actual experiments, the weather model was run
ninety-eight times, but forced with cooler “counterfactual”
boundary conditions. So now the research team had ninety-eight
sets of May 31, 2015 daily heat index values from the actual
world, and ninety-eight sets of daily heat index values from a
counterfactual world without climate change. Because human
health impacts are much greater when we experience persistent
heat stress, they next converted these daily values into running
five-day averages. The figure below shows the results of these
experiments. The dashed line shows the “Counterfactual” distri-
bution of five-day heat index values in a world with climate
change. The solid black line shows the “Actual” distribution from
a world with climate change. Climate change made the 2015 heat
waves much more probable.
The information contained in these distributions can be used to
examine changes in extreme event intensity or frequency. To
explore changes in the magnitude of extreme heat events similar
to those experienced in 2015, the India study identified the “return
period” of the event. Such events occurred about every three years
0.3
Counterfactual
Distribution--
Fraction of Events
--Actual Distribution
0.2
0.1
0
36 37 38 39 40 41 42 43 44 46
Daily Heat Index (°C)
Figure 5.9 Figure showing histograms of five-day averages of

simulated May 2015 daily maximum heat index values in
Hyderabad for the counterfactual (dashed line) and actual
(solid black line) simulations. Recreated based on results
shown in Wehner et al.’s 2016 article. © American
Meteorological Society. Used with permission
in the observational record. By contrasting the one-in-three year

warmest heat index values in the Actual and Counterfactual dis-
tributions, Wehner et al. identified a large change in magnitude
(+1.7C). Climate change made Hyderabad heat extremes much
larger. To explore changes in the frequency of the extreme heat
events, the team calculated the probability of the observed peak
five-day heat index value occurring in a world with and without
climate change. In a world without climate change, the observed
extreme event was very unlikely, only happening once every
ninety-two years. The increase in the probability of such an event
happening can be expressed as the “risk ratio.” The risk ratio is
calculated by dividing the probability of an event happening in a
world with climate change by the probability of the event occur-
ring in a counterfactual world without. The data presented in
Figure 5.9 indicates that climate change made it much more likely
(32.8 times) for Hyderabad to experience deadly heat waves simi-

lar to those that occurred in 2015. Implication? Human-induced
climate change almost certainly contributed to thousands
of deaths.
Sidebar: Catastrophic Fire Danger for the Greater

Sydney Area
As I updated this chapter in early November 2019,
catastrophic fires raged, for the first time ever, through
Figure 5.10 Time series of average annual Sydney daily

maximum air temperatures from the Observatory Hill
weather station. Data from www.bom.gov.au/climate/data/
Sydney, Australia.24 Fueled by low rainfall levels, dry vegetation,

and extremely warm temperatures, more than 120 fires burned
across two states, stretching 1,000 km from Sydney to the Gold
Coast near Brisbane.25 Sydney has extensive weather data stretch-
ing back to the mid-nineteenth century, and this gives us an
excellent opportunity to look at just how warm conditions have
been in Sydney. Figure 5.10 shows a continuous time series of
January–December annual average daily maximum air tempera-
tures. Daily maximum air temperatures are recorded each day in
the early afternoon, when the daily temperature cycle reaches its
peak. Each dot in Figure 5.10 represents the average of a year’s
worth of these values. The year 2019 was the warmest year on
record. We also note a very strong upward trend, with the
2015–2019 data more than 1.2C warmer than the 1990s average.
24
en.wikipedia.org/wiki/2019%E2%80%9320_Australian_bushfire_season.
25
www.bbc.com/news/world-australia-50365131.
PRECIPITATION EXTREMES
6 Observations and Impacts
Introduction
On Monday, November 29, 2019, about 4:15 PM, the
Cave Fire broke out two miles from my house in the woods
behind Santa Barbara, within spitting distance from the Painted
Cave community. Just a week prior, on Tuesday, I had discussed
climate change and fire risk with my friends Ted Adams and
Mike Williams on their disaster preparedness radio show. Now
Ted’s house was near the front lines of the flames. Thirty mile-
an-hour downhill winds conspired with bone-dry brush to fuel
the conflagration, which raced downhill, away from Painted
Cave but toward the densely populated city of Santa Barbara.
From below, Santa Barbarites watched the fire descend at insane
speeds. It being winter, the sun rapidly set. With the darkness
came the grounding of the fire planes, conditions being too
dangerous to fly. Wind-blown embers raced downhill ahead of
the walls of flames, producing a series of spot firs that helped
feed the fire’s rapid advance. By 8 PM the Cave Fire stretched
hungry fingers toward Santa Barbara’s house-filled foothills.
Mandatory evacuations were called for more than 5,000 people.
Throughout the night a few hundred hardy first responders
stood between the city and disaster, ultimately stopping the fire
and saving the city.
123 / Precipitation Extremes: Observations and Impacts
It just so happened that I had a prescheduled telephone

call with Aurora Almendral, a National Geographic reporter on
her way to Kenya to work on a story on climate change and
migration. I stood on a hill above and behind the Santa Barbara
with my wife, kids, and neighbors, watching the flames spread
toward our homes, and tried to explain to Aurora how a
warming atmosphere intensifies droughts, floods, and fires:
basic physics plus warming temperatures equals more extreme
rainfall when the atmosphere is moist and enhanced moisture
extraction from land surfaces when the atmosphere is dry.
On December 4, 2019, the Global Carbon Budget pro-
ject released their annual report.1 In 2019, once again, about
half our carbon emissions remained in the atmosphere, while
about a quarter were absorbed by the ocean and a quarter by
the land; carbon dioxide concentrations in the atmosphere now
exceed 410 parts per million. Once again, we emitted unpreced-
ented amounts of CO2.
On December 18, the East African Food Security
and Nutrition Working Group released a report detailing
the dangerous impacts of catastrophic flooding on eastern
Africa. According to rainfall observations from the Climate
Hazards Center (where I work), much of the region experi-
enced the wettest or second-wettest October–December
rainy season on record (since 1981). Floods impacted more
than 2.8 million individuals, directly killing 280 people and
leading to widespread outbreaks of cholera and acute watery
diarrhea.
In December 2019, the impact of drought and consecu-
tive extreme heat waves also contributed to catastrophic condi-
tions supporting ninety-eight fires in New South Wales,
Australia.2 Fires burned outside every major Australian city,
not to mention additional fires in Singapore, Indonesia, and
Papau New Guinea (Figure 6.1). Two volunteer Australian
1
www.globalcarbonproject.org/index.htm.
2
twitter.com/NSWRFS/status/1208717082012946432/photo/1.
Figure 6.1 Fire locations on December 22, 2019.

Map obtained from myfirewatch.landgate.wa.gov.au/
firefighters, both young fathers (Figure 6.2), died when their

truck was hit by a falling tree.3
A Conceptual Model of Temperature: Water

Vapor Relationships
Linking all these events, and of course all life on Earth,
is our atmosphere.
Our thin, thin atmosphere.
If the Earth were a basketball, the atmosphere would be
0.03 inch or 0.8 mm thick, literally whisker deep.
3
www.rfs.nsw.gov.au/news-and-media/general-news/featured/support-for-
firefighter-families.
Figure 6.2 Australian volunteer firefighters Andrew O’Dwyer

(age 36) and Geoffrey Keaton (age 32) lost their lives fighting fires
southwest of Sydney on December 20, 2019.
The vast majority of this atmosphere is made up of

nitrogen and oxygen atoms. Nitrogen accounts for approxi-
mately 78 percent of the atmosphere, and is boring (chemically
inert). Kind of like a climate scientist at a party, nitrogen just sits
there without much to say. Oxygen, which makes up 21 percent
or so of the atmosphere, is much more active, supporting key
processes like photosynthesis and combustion. While oxygen
levels have varied dramatically on geologic time scales, on cli-
mate and weather time scales they remain essentially constant.
Unlike nitrogen or oxygen, the number of atmospheric
water molecules (aka water vapor) varies substantially from
place to place and time to time. U these variations is critical to
understanding climate change. These insights are accessible to

anyone with a good comprehension of basic scientific principles.
Air molecules are disconnected from each other, and free to
bounce around in very random ways. A molecule, most fre-
quently nitrogen or oxygen, will careen in some random direc-
tion until its electron cloud encounters another molecules’
electron cloud, at which point they exchange momentum, prom-
ise to call each other in the morning, and head in disparate
directions. The velocity of these molecules is directly related to
the air’s temperature and density. Cooler air with slower mol-
ecules will tend to be more dense. The molecules are packed
together more closely. Warmer air with faster molecules will
tend to be less dense. The molecules will be, on average, farther
apart. Figure 6.3 shows this relationship schematically. Typical
atmosphere molecules will usually be paired nitrogen (N2) or
oxygen (O2) atoms. Nitrogen and oxygen make up 99 percent of
the atmosphere. Slower-moving cold air parcels will tend to
have smaller distances between them, making it harder for the
Figure 6.3 Schematic diagram of cold air (left) and warm air
(right). On average, the molecules in cold air are closer together,
making less room for water molecules. In warm air, the molecules
are farther apart, creating more space for water molecules.
air to hold water vapor molecules (shown as a big oxygen atom

with two small oydrogen atoms attached). Warming the air
causes the nitrogen and oxygen to move more quickly and
spread out, creating more space for water vapor.
Ironically, the fact that a warmer atmosphere can hold
more water vapor can lead to increased drought intensities and
drier fire fuels. At small scales near the surface of a plant or a
clod of earth, water vapor molecules may move up from the
surface into the air, but this process can slow or even stop
completely depending on the air’s temperature and total amount
of preexisting water vapor in the lowest level of the atmosphere.
If we keep adding water vapor to any parcel of air at any
temperature, it will eventually become “saturated, with a rela-
tive humidity of 100 percent.” The parcel is no longer able to
absorb more water vapor. But warmer air can hold a lot more
water vapor; increasing the temperature by one degree Celsius
increases the potential water-holding capacity of air by about 7
percent. This relationship is commonly referred to as the
Clausius-Clapeyron relationship. Under dry conditions, like
during the gap between one rainy season and the next, increases
in air temperatures will often lead to air that is less saturated,
making it easier for vegetation to lose moisture. A modest level
of warming, supporting modest enhancements of plant water
loss month after month, can produce substantial decreases in
plant and fuel moisture.
Decreases in plant and fuel moisture can in turn lead to
increased fire risks in, for example, places like California and
Australia. Greater vegetative moisture loss can also dry out grass-
lands, providing less food for cattle, sheep, goats and camels.
These moisture losses have large impacts on both lucrative ranches
in the US4 and on poor, food-insecure pastoralists in Africa.5
4
Williams, Emily, et al. “Quantifying human-induced temperature impacts on the
2018 United States Four Corners Hydrologic and Agro-Pastoral Drought.”
Bulletin of the American Meteorological Society 101.1 (2020): S11–S16.
https://journals.ametsoc.org/doi/pdf/10.1175/BAMS-D-19-0187.1.
5
Pricope, Narcisa G., et al. “The climate-population nexus in the East African
Horn: Emerging degradation trends in rangeland and pastoral livelihood zones.”
Global Environmental Change 23.6 (2013): 1525–1541.
https://www.sciencedirect.com/science/article/pii/S095937801300173.8
But the same simple physical principle illustrated in

Figure 6.3 (the Clausius-Clapeyron relationship) can also lead
to increased precipitation extremes. Nitrogen and oxygen mol-
ecules are very well mixed in our atmosphere. Everywhere you
go, there they are. Not so with water vapor. Warmer air holds
more water vapor, so we usually find more water vapor in the
tropics. But even at a fixed latitude we will see huge variations
over time and place. When air is close to saturation (100 percent
relative humidity), warming increases the total amount of water
available for precipitation.
A simple conceptual model of rainfall (Figure 6.4) can
help us understand why climate change is making extreme pre-
cipitation even more extreme. Heavy precipitation occurs when
moist air rises quickly into the upper atmosphere and cools. It
gets colder as you go up in the atmosphere because the air
becomes less dense. This cooling reduces the air’s ability to hold
water vapor. When the rising parcel of air becomes saturated, the
water vapor it contains condenses, becoming liquid droplets or
ice crystals: clouds and precipitation. Warmer air will contain
more water vapor, supporting heavier rainfall.
A little algebra can help us express this mathematically.
We can describe rainfall rates as the multiplicative interaction of
two terms – water vapor (W) and the vertical velocity (V). W is
simply the amount of gaseous water vapor in a parcel of air. V is
the vertical (upward when raining) velocity of that parcel. Rain
happens when moist air moves up and cools, causing the water
vapor in the air to cool, condense, and precipitate. So we can
assume that rain rates (RR) will be proportional to W times
V (RR ¼ W V) W V describe the vertical mass flux of
water, which is strongly related to precipitation intensities (i.e.,
millimeters or inches of rainfall per day).
Now, if we do a little thought experiment (Figure 6.4), we
can relate this simple precipitation model to increases in air tem-
perature. On the bottom left and bottom right we have parcels of
air with the same relative humidity, but the right-hand parcel is
warmer. It therefore has substantially more water vapor mol-
ecules. Now we subject both parcels to the same vertical velocity,
Figure 6.4 A thought experiment explaining the impact of

increased water vapor on precipitation rates.
whisking them up into an ever-colder atmosphere. Before long,

the water in each parcel will condense. Assume the velocity (V) is
the same in both cases. Since the warmer parcel holds more water
vapor, it would likely produce more precipitate. W would be
bigger. This simple relationship provides a good starting point

for understanding changes in precipitation extremes.
Climate change impacts on these two terms, W and V,
are generally called thermodynamic and dynamic6 controls.
Thermodynamic control is just a fancy term to describe how
warmer air holds more water vapor, i.e. the Clausius-Clapeyron
relationship. As atmospheric moisture content increases, precipi-
tation increases.7 For our simple model, if vertical velocity stays
the same but water vapor increases, then precipitation will
increase in proportion to increases in W. A general rule of thumb
is that if relative humidity values remain constant, then the
amount of water vapor will increase by about 7 percent per
degree Celsius of warming. But changes in observed and modeled
average precipitation values increase much less, only about 3.4
percent per degree of global warming.8 Why? The answer is that
mean precipitation values face a general energetic constraint;
when water condenses, it releases energy, and there is a balance
of energy between the cooling of a column of air by radiation and
the heating of a column by precipitation. Hence mean precipita-
tion changes are constrained to be substantially less than those
expected due to changes based just on the relationship between
air temperatures and the water holding capacity of a warming
atmosphere. This has implications for droughts, because our
warming atmosphere’s ability to dry out the land may be increas-
ing faster than the atmosphere’s average precipitation rates.
At short time scales, however, thermodynamic controls
on rainfall extremes can increase by more than 3.4 percent or
even 7 percent per degree Celsius. At small spatial and temporal
scales, the “thermodynamic” control supplied by the Clausius-
Clapeyron relationship predicts a ~7 percent increase in heavy
precipitation rates per degree of warming. As we see in Figure 6.4,
6
Emori, Seita, and S. J. Brown. “Dynamic and thermodynamic changes in mean
and extreme precipitation under changed climate.” Geophysical Research
Letters 32.17 (2005).
7
Trenberth, K. E., A. Dai, R. M. Rasmussen, and D. B. Parsons, “The changing
character of precipitation.” Bulletin of the American Meteorological Society 84
(2003): 1205–1217.
8
Allen, Myles R., and William J. Ingram. “Constraints on future changes in climate
and the hydrologic cycle.” Nature 419.6903 (2002): 224.
a warmer atmosphere is like a bigger sponge. When this sponge

rises and cools, more water can condense and fall as rain.
But as we will see when discussing cyclones and hurri-
canes, warming impacts on the most extreme rainfall events
could become even more severe. In extreme rainfall events mois-
ture doesn’t just move up; it also moves side to side. Superstorms
like Hurricane Harvey draw in winds from hundreds of miles
away, collecting additional moisture, to potentially explosive
effect. This extra moisture increases rainfall rates. But condensa-
tion also releases a tremendous amount of energy, potentially
creating a weather feedback that leads to more rapid moisture
convergence and accelerated vertical motions.
Examining the Data

While a detailed analysis is beyond the scope of this
book, I always want to present you with real data to underscore
the topics we are discussing. Here we briefly examine five-day
precipitation totals from the archive of the Climate Hazards
Center for Infrared Precipitation with Stations (CHIRPS). This
is the precipitation product produced by our group (the Climate
Hazards Center, CHC) at the University of California, Santa
Barbara.9 My friend Pete Peterson and other members of our
team have spent years making this one of the best global rainfall
products. CHIRPS combines rainfall estimates from geostation-
ary infrared weather satellites with observed rain gauge obser-
vations. This rapidly updated product was designed to support
humanitarian relief efforts. Here we analyze global CHIRPS
data following an approach commonly used to examine changes
in extremes.10 For each year and location, an annual maximum
rainfall accumulation total is identified, and then divided by that
9
chc.ucsb.edu.
10
Donat, Markus G., et al. “More extreme precipitation in the world’s dry and wet
regions.” Nature Climate Change 6.5 (2016): 508; Donat, Markus G., Oliver
Angélil, and Anna M. Ukkola. “Intensification of precipitation extremes in the
world’s humid and water-limited regions.” Environmental Research Letters 14.6
(2019): 065003.iopscience.iop.org/article/10.1088/1748-9326/
ab1c8e#erlab1c8ef1.
Figure 6.5 Time series of running five-year averages of global

five-day terrestrial extreme precipitation rainfall totals expressed
as percentages of the 1981–1985 value. Also shown is a line
describing the data’s strong linear trend along with a 2050
projected value.
location’s long-term average total annual precipitation. This is a

measure of the intensity of each location’s most extreme precipi-
tation for that year. We can then calculate five-year global
averages of these values to look for trends in extreme precipita-
tion. To facilitate analyzing changes over time, Figure 6.5 dis-
plays these values expressed as the percentages of the first
1981–1985 five-year value. While there are natural year-to-year
fluctuations, we also a see a strong upward trend in the data.
Between the first five-year period (1981–1985) and the last
(2015–2019), we see about a 9 percent increase in the observed
extreme rainfall events.
A 9 percent difference may not sound like much, but it
can really be substantial, increasing the intensity of very extreme
rainfall events. Flooding is all about the extra water. Runoff is
what we call the water that is left over after rainfall is absorbed
by the soil and evaporated up into the atmosphere. In simple
terms, Runoff ¼ Rainfall Evaporation. At low precipitation
intensities, the Sink can be as big as the Source. This relationship

tends to change in a nonlinear way when a watershed receives a
lot of water, so a 9 percent increase in a rainfall extreme may
increase runoff (and flooding) by much more. If the observed
trend persists, we will likely see another large increase in the
wettest five-day precipitation events (star in Figure 6.5). This is a
recurring theme in this book. We have already seen a substantial
increase in precipitation-related hazards that are harming
people today. More increases are on the way. While the atmos-
phere is a complicated place, the relationship between atmos-
pheric warming, increases in water vapor, and associated
enhancements to extreme precipitation are basic physics that
everyone can understand (Figures 6.3 and 6.4).
2015–2019 Impacts of Extreme Precipitation

While we will deal with hurricanes and cyclones in the
next chapter, here we will briefly discuss some of the most severe
floods and storms that happened between 2015 and 2019.
Between 1998 and 2017, floods and storms (including the
impact of hurricanes) affected more people than any other type
of disaster, impacting 2.7 billion people overall and resulting in
1.99 trillion dollars of recorded economic losses.11 Most of the
wettest extreme rainfall events tend to occur in Asia and the
Maritime Continent (the region of Southeast Asia that com-
prises, among other countries, Indonesia, Philippines, and
Papua New Guinea), with heavy precipitation also occurring
in northern Australia, southeast Africa and Madagascar, South
America, Central America, and the southeastern United States.
We can explore the potential impact of the extremes by digging
into the international emergencies database (EM-DAT), which is
maintained by the Centre for Research on the Epidemiology of
disasters (www.cred.be/). This is the database used by the
United Nations Office for Disaster Risk Reduction. Annual
11
In 2017 US dollars. United Nations report, Economic Losses, Poverty &
Disasters: 1998–2017, www.unisdr.org/files/61119_credeconomiclosses.pdf.
disaster statistical reviews and other interesting analyses are

available on the CRED site (www.cred.be/publications).
According to EM-DAT, in 2015–2019, severe storms (including
cyclones, typhoons, and hurricanes), floods, and landslides have
affected some 405 million people, resulting in more than 31,000
deaths and $569,437,073,000 ($569 billion!) in damages. While
quantifying such impacts with great precision is difficult, it is
clear that flooding and rainfall extremes are expensive
and dangerous.
Focusing just on large flood events, we can explore the
data in more detail. Table 6.1 summarizes recent (2015–2019)
EM-DAT12 disaster statistics for big flood events. Big flood
events are defined as those impacting at least 1.5 million people.
As you can see, most of these big-impact events occur in Asia.
This spatial selection arises from an interaction between human
development and natural climate variability. This is the most
densely populated region on Earth. But it also contains many
regions that customarily receive abundant and sometimes
extreme rains. Just over five years, the total number of people
being affected is pretty staggering: 223 million. Associated with
these floods were a total of 9,036 deaths and economic losses of
approximately $78 billion.
To put a human face on these statistics, we can consider
some of the worst flood disasters from 2018. A deadly flood
occurred in the Indian state of Kerala, which is on the western
side of the Ghats Mountains in southwestern India. The western
side of the Ghats is one of the rainiest places on Earth, because
warm moist monsoon winds blow right against the mountain
range. This produces strong vertical motions. Strong vertical
motions combined with high levels of water vapor produces
intense rainfall events (RR ¼ W V). In Kerala in 2018 torren-
tial rains quickly accumulated, dams needed to be opened to
release the water, and more than 500 people are believed to have
directly died. Extensive inundation displaced more than a mil-
lion people and impacted more than 23 million people, causing
12
www.emdat.be/
Table 6.1. EM-DAT disaster statistics for non-cyclone 2015–2019

floods affecting at least 1.5 million people.
Total Total Total Damage

Year Country Name Deaths Affected (Millions US$)
2016 China 816 61,297,200 31,793
2019 India 1752 23,830,060 unknown
2018 India 710 23,307,698 2,865
2017 India 1046 22,271,843 2,117
2015 India 839 16,413,459 2,880
2017 China 231 12,333,608 8,446
2019 Iran 75 10,001,096 2,580
2017 Bangladesh 144 8,086,025 628
2019 China 152 6,375,861 unknown
2019 Bangladesh 119 4,000,000 unknown
2017 Thailand 150 3,790,498 1,308
2018 China 244 3,698,400 4,570
2016 Dominican Republic 15 2,792,000 unknown
2016 Philippines 45 2,563,098 9
2017 Peru 200 2,188,505 3,200
2018 Nigeria 300 1,938,204 275
2016 Bangladesh 106 1,900,000 150
2017 Philippines 22 1,842,000 8
2016 India 666 3,806,000 1,499
2017 Nepal 187 1,706,134 595
2015 Myanmar 172 1,635,703 119
2015 Pakistan 367 1,577,490 1
2018 Japan 246 1,500,102 9,500
2016 Viet Nam 91 1,428,585 161
2015 Bangladesh 31 1,411,901 40
Table 6.1. cont’d

Total Total Total Damage
Year Country Name Deaths Affected (Millions US$)
2015 China 310 1,374,350 6,992
2017 Ghana 1,000,000 unknown
Total 9,036 223,069,820 79,737
Source: EM-DAT: The Emergency Events Database – Universite catholique de

Louvain (UCL) – CRED, D. Guha-Sapir – www.emdat.be, Brussels, Belgium. Empty
cells indicate missing values. Damage amounts present reported values in US dollar
equivalents from the year of the disaster. Created December 23, 2019.
an estimated $4 billion in damages and leading to at least

319 fatalities.
Let’s look at another example. The summer storms that
struck Japan between June 28 and July 9 in 2018 were even more
financially devastating. On June 28, heavy rains began falling
across southwestern Japan. On July 3, during Typhoon
Prapiroon, moist air and an extra-tropical low-pressure system
combined to produce very intense rains. Daily rainfall totals
exceeded 250 millimeters. In some places, eleven-day rainfall
totals quickly exceeded 1,850 millimeters (Figure 6.6). The inten-
sity of these totals is amazing. In the village Umaji of Umaji, in
Kōchi prefecture, four consecutive days received rainfall totals
exceeding 250 millimeters. A 1 meter 1 meter 1,850 milli-
meter volume of water would weigh 1,853 kilograms. Japan
reeled from the impact of this flooding, which contributed to a
~1.2 percent contraction of the Japanese economy.13 According
to EM-DAT, this deadly natural disaster resulted in 246 mortal-
ities and a billion dollars in economic losses.
Climate Change Projections

Are we heading toward more such extreme precipita-
tion events? To explore this question we can use a set of
13
www.bbc.com/news/business-46203864.
Figure 6.6 Top: 2018 daily precipitation in Umaji, Kōchi Japan.

Bottom: A picture of Umaji taken by Saigen Jiro.
annual simulated climate extreme statistics hosted at the

Koninkliijk Nederlands Meteorologisch Instituut (the Dutch
national weather service) on the great Climate Explorer
website (climexp.knmi.nl). These climate extremes were
obtained from the Canadian Centre for Climate Modeling
Figure 6.7 CMIP5 five-day precipitation extremes, for Asia and

Oceania, expressed as a fraction of the average
annual precipitation.
and Analysis14 using a large multi-model ensemble of climate

change simulations.15 Because so many extreme precipitation
events happen over Asia and Oceania, I have calculated aver-
ages over land between 65E and 152E and 20S and 40N.
As in Figure 6.5, the annual maximum five-day precipitation
total is expressed as a fraction of the 1950–1979 mean annual
total precipitation. Four time series are presented in
Figure 6.7. If we continue along our current rapid emissions
scenario (RCP8.5), we are likely to see continued rapid
increases in precipitation, with the models anticipating that
typical five-day maximum rainfall totals will be about
14
climate-modelling.canada.ca/climatemodeldata/climdex/climdex.shtml.
15
These results are based on the Phase 5 Coupled Model Intercomparison Project
(CMIP5) ensemble.
35 percent more intense than they were in the 1950s. Given

that we are already experiencing large increases in rainfall
extremes with impacts costing hundreds of billions of dollars,
resulting in serious disruptions in development,16 continued
increases in extreme precipitation will be costly and deadly.
Note, however, that even modest reductions in green-
house gas emissions, such as those associated with the RCP4.5
and 6.0 pathways, really make a big difference by mid-century.
According to these simulation results, Asian precipitation
extremes increased by about 5 percent between 1950 and
2018. If we act now to curb emissions, we may see a similar
increase by mid-century. Following a rapid growth (RCP8.5)
trajectory means that we may see rainfall extremes increasing
twice as fast. The decision on which path we take is being made
now; whether we choose to decide or not, we still have made a
choice.17 There is solid observational (Figure 6.5) and model-
based (Figure 6.7) evidence supporting the link between a
warming atmosphere and more intense precipitation extremes,
and clear evidence that these extremes are having deadly and
costly impacts today (Table 6.1). Relatively modest (RCP4.5) or
aggressive (RCP2.6) reductions in emissions will limit increases
in these human catastrophes and attendant human suffering.
16
gar.unisdr.org/sites/default/files/gar19distilled.pdf.
17
Based on the band Rush’s song, Freewill.
HURRICANES, CYCLONES,
7 AND TYPHOONS
Solon, the ancient Greek philosopher famed for giving
the city of Athens its laws, famously told the rich, successful,
and powerful King of Lydia, “Count no man lucky until his end
is known”1 (Figure 7.1). How true. Each of us, no matter how
surrounded by kindly kin, treasure, glory, and the good opinion
of others, could be wracked with ruin in an instant. Failing
brakes, the cough of a sick child, cancer. So fragile we are. So
intertwined with other. So ultimately enmeshed in the universe
around us. This is a good thing, our existential thingness, which
gives us an opportunity to be good. And that goodness knows
that the potential for catastrophe – drought, flood, fire – links us
in a common web of humanity.
Life tends to be conservative because there are so many
ways to not be it (alive). Between the ultimately cool and
absolutely hot, life hangs in a delicate balance. We live in a
narrow temperature band, and as climate change shifts our
planet’s narrow temperature regime, we may experience more
extreme hurricanes and cyclones, which often bring cata-
strophic damage.
At absolute zero – zero degrees Kelvin or 273.15
degrees Celsius – the motion of molecules ceases. This is the
1
classicalwisdom.com/philosophy/count-no-man-happy-end-known/.
141 / Hurricanes, Cyclones, and Typhoons
Figure 7.1 Solon before Croesus, by the Dutch painter Gerard

van Honthorst, painted in 1624. https://en.wikipedia.org/wiki/
Gerard_van_Honthorst#/media/File:Honthorst_solon_and_
croesus.jpg
absolute minimum temperature, absolute cold. Then there is

absolute hot, about which scientists are somewhat unsure.
According to quantum theory derived in the late 1980s, the
maximum obtainable temperature would be the Planck tem-
perature, or ~1.416808cc1032 Kelvin. But let’s be conserva-
tive, and dial back to the temperature of a supernova, which is
about 100 billion Kelvin. So temperatures range from
~ 273.15C to ~1011C, or ~0 to 100,000,000,000 Kelvin.
Between these extremes, life hovers. Let’s sketch out the
potential range of temperatures on Earth, beginning at the
hypothetical coolest: Miles Davis’s Sextet playing “Kind of
Blue” in Antarctica ( 98C or 175K). And then for hot, you
pretty much can’t beat the emission temperatures (~50C or
~323K) coming from the parking lot of the Devil’s Tower
National Monument in northeastern Wyoming in August,
during the annual Sturgis Motorcycle Rally.2 While to me the

Antarctica-to-Sturgis spread seems exciting, exotic, and better
than Monday, from a galactic scale we are looking at mere
digits, about 200K in a universe that ranges from 0 to ~1011K.
Pessimists should observe that we have a lot more room on the
upside, where global warming abides.
Of course, those perceptive souls who might wish to
believe in a benign universe or caring higher power could men-
tion that we seem to be luckily ensconced within a few magic
decimal places conducive to watery phases and associated life-
sustaining metabolic processes. Out of this many decimal places,
we happened happily in the two bold ones a few places from the
rightmost digit: 10000000 00 00.
Maybe it is totally random that our planet has gyrated
for billions of years, almost (in cosmic age) since the birth of the
universe, between temperatures supporting freezing liquid and
gaseous water. On the other hand, if I were playing God, and
looking forward to the imminent and immanent evolution of
human life, I would set the temperature dial to that small
fraction of the spectrum between ~175 and 323 Kelvin (~ 98
to +50C). And there we are. Water, in its myriad of forms,
makes complex life possible on our Blue Planet (Figure 7.2).
Our Blue Marble’s temperature range uniquely (at
least as we so far know) supports water in all its three glorious
phases (solid, liquid, gas). Near your house there might be a
nice cool pond or lake or river. Above this liquid water body
you will find moist air. In this moist air, water vapor (water in
its gaseous state) freely mixes with our atmosphere’s more
boring constituents, like nitrogen and oxygen. Imagine, unseen
to you, a raging house party, with all those wild and crazy
2
Please note the following from www.sturgismotorcyclerally.com/faq-and-stats/ on
December 26, 2018: “To be married in Sturgis® (Meade County) you must obtain
a marriage license at the Register of Deeds office located in the Erskine Building at
1300 Sherman St., Suite 138. (1 block south of Main St.) Both applicants must be
present with identification and $40 in cash or travelers’ checks. There is no waiting
period, the license is good for (20) days and same day marriages are legal. You are
responsible for locating your own wedding official.”
Figure 7.2 Our Blue Marble. A view of most of North America

taken from a low orbit of about 826 km altitude. This composite
image uses a number of swaths of the Earth’s surface taken on
January 4, 2012. By NASA/NOAA/GSFC/Suomi NPP/VIIRS/
Norman Kuring, www.nasa.gov/multimedia/imagegallery/
image_feature_2159.html.
H2Os bouncing around. Folks in the pond would like to come

to the party, but there is only so much room. The room is
packed. If someone comes in the door, someone falls out the
window. Unless air temperatures increase. Then the spaces
between the molecules of nitrogen and oxygen increase,
making more room to party.
Now assume our pounding party pad is pendulous,
suspended beneath a giant hot air balloon. As we drift higher,
air temperatures quickly drop. The warmth of the air around us
is strongly influenced by the weight of the air above, packing the

atmospheric molecules closer together. The lowest 85 percent of
the atmosphere (the troposphere) stretches up about 12 kilo-
meters where it reaches the “tropopause.” The tropopause
region is where incoming radiation causes the very thin air of
the upper atmosphere to warm. As temperatures cool, there is
less room in the “inn.” As our pendulous party pad ascends,
declining air temperatures quickly cool our mini-mighty party
partners. They stop to check their disco-watches (digital of
course) and realize it’s time to chill.
Now something important happens. Slowing down,
they aggregate, undergoing a “phase transition” – condensing
into water droplets. These droplets, denser than the surrounding
air, drop – i.e., precipitate. As our pendulous party pad rises
higher, the disco palace reaches freezing temperatures. Panicked
partiers leap from windows, freeze together in clumps, and drift
slowly down to Earth. None of us is truly alone. Even snow-
flakes are compound creatures.
Precipitation processes are super-critical to supporting
life on our planet. Sometimes, though, we can get way too much
of a good thing. Increasing air temperatures can contribute
to more extreme hurricanes and floods. As air temperatures
increase, the water-holding capacity of the atmosphere goes up
in a predictable systematic way, as we saw in the previous
chapter. This additional water vapor is expected to directly
increase the intensity of extreme precipitation. An increase in
the size of our pendulous party pad means more precipitating
panicking partiers.
But there are also important indirect dynamic effects.
Water phase transitions can release or absorb a lot of energy. It
takes a lot of energy to boil or freeze liquid water. When gaseous
water vapor turns liquid or freezes, a tremendous amount of
energy can be released – which can have important dynamic
impacts. This released energy can rapidly heat the atmosphere,
causing rapidly rising columns of air. Which in turn causes
precipitation, triggering precipitating partying panickers that
release more energy – a positive feedback loop.
The “collective” aspect of atmospheric water vapor

exacerbates both of the direct (7 percent per degree) and indirect
dynamic effects of water vapor increases. Unlike atmospheric
temperatures and pressures, which tend to be evenly distributed
in space and time, water vapor can vary greatly from place to
place in the atmosphere as winds collect converging flows.
Sweeping across warm waters, winds can gather water from a
vast fetch, sometimes to catastrophic effect.
Take, for example, Hurricane Harvey, which pummeled
the coast of Texas and Louisiana in late August 2017. The
greater Houston area received 30 to 40-plus inches of rain, with
several totals well above 50 inches (127 cm) of rain from
Harvey,3 making it the largest rainfall event on record in the
lower forty-eight states. As reported in the Washington Post,4
John Nielsen-Gammon, Texas state climatologist, found that
“Harvey’s total rainfall concentrated over a 20,000-square-mile
area represents nearly 19 times the daily discharge of the
Mississippi River.” According to the Post, this is equivalent to
about 9 trillion gallons of water,5 equivalent to a cube of water
two miles long, two miles wide, and two miles high; this would
be enough water to fill the Great Salt Lake in Nevada, twice. In
Houston, more than five million people lived in areas receiving
more than 36 inches of rain.6 More than 32,000 people were
forced to abandon their homes and sleep in shelters.7 Something
to always remember, however, is that water isn’t just wet – it’s
also energy. When 9 trillion gallons of water are evaporated
from the ocean, the ocean is cooled, and a lot of energy is stored
up in the atmospheric water vapor. The total energy associated
3
www.weather.gov/hgx/hurricaneharvey.
4
www.washingtonpost.com/news/capital-weather-gang/wp/2017/08/29/harvey-
marks-the-most-extreme-rain-event-in-u-s-history/?utm_term=.3a6689ffe5b3.
5
www.washingtonpost.com/news/capital-weather-gang/wp/2017/08/27/texas-
flood-disaster-harvey-has-unloaded-9-trillion-tons-of-water/?tid=a_inl&utm_
term=.066934a16c9d.
6
https://mashable.com/2017/08/29/harvey-houston-flood-by-the-numbers-worst-
flood/.
7
www.npr.org/sections/thetwo-way/2017/08/30/547227788/harvey-makes-
landfall-again-in-louisiana.
Figure 7.3 The most expensive US hurricanes, expressed in terms

of adjusted 2017 US dollars. Based on data provided NOAA’s
National Hurricane Center.
with evaporating 9 billion gallons of water is about 76 1018

Joules (76 quintillion Joules; I had to look that one up). When
water vapor condenses during precipitation, this same amount
of energy is released. This energy rapidly warms the atmosphere,
fueling a hurricane’s gale-force winds. Over Texas, peak wind
gust speeds reached 132 miles (211 km) per hour. In 2017, all
the humans around the world consumed about 630 quintillion
Joules of energy – or just 8.3 times the amount of energy
released in Texas by Hurricane Harvey.
According to NOAA,8 the economic impacts of Harvey
were massive – $125 billion, making it the second most expensive
hurricane after Katrina (Figure 7.3). Note also that two other very
expensive hurricanes occurred in 2017: Maria, which struck
Puerto Rico and the Virgin Islands; and Irma, which struck
Florida. Maria devastated Puerto Rico,9 contributing to the deaths
of at least 2,975 people and leaving the island without power for
months. Irma was the strongest observed hurricane in the Open
Atlantic, causing catastrophic damage to Florida and the
8
www.nhc.noaa.gov/news/UpdatedCostliest.pdf.
9
www.nationalgeographic.com/magazine/2018/03/puerto-rico-after-hurricane-
maria-dispatches/.
Figure 7.4 Total cost of the biggest hurricanes summed by

twenty-year time periods.
Caribbean Islands along its path. In 2018, hurricanes Florence

and Michael struck the southeastern United States, causing some
$33 million in damages. These three cyclones (Harvey, Maria,
Irma) alone inflicted some $260 billion in damages, according to
the NOAA data shown below (Table 7.1). To put this in perspec-
tive, the 2003–2010 US–Iraq war cost US taxpayers about $822
billion in direct military expenditures, according to the November
2018 Brown University Cost of War report.10
We can rearrange the data from Figure 7.3 by summing
up impacts over twenty-year time periods (Figure 7.4). This
emphasizes the huge jump in economic costs incurred by the
United States since the beginning of the twenty-first century. I’m
in the disaster business, and used to looking at bad news, but
even I was shocked by Figure 7.4. Together, the biggest hurri-
canes inflicted about $740 billion in damages between 2000 and
2018. So, so far, twenty-first-century hurricanes have been
almost as costly as the Iraq War. There are many factors related
to these cost increases, and a very important contributing factor
is that people like to live next to the ocean. According to the
10
watson.brown.edu/costsofwar/files/cow/imce/papers/2018/Crawford_Costs%
20of%20War%20Estimates%20Through%20FY2019.pdf.
Wall Street Journal,11 between 1980 and 2017, population

density in coastal areas of the US Gulf and eastern states
increased from about 350 to 550 people per square mile, while
densities in the rest of the US remained very steady. And coastal
real estate and repairs are expensive. While there are debates
regarding the true inflation-adjusted costs of these extremes,
most assessments indicate that, indeed, there has been a very
large recent increase in US hurricane-related losses.12
Complexities Surrounding Cyclone Attribution

and Detection
It is important to note, however, that increasing cata-
strophic losses do not necessarily implicate climate change. The
relative complexity and rarity of hurricanes and cyclones makes
their detection and attribution very difficult. According to an
authoritative National Academies report, “The frequencies and
intensities of tropical cyclones and severe convective storms are
related to large-scale climate parameters whose relationships to
climate change are understood to varying degrees but, in gen-
eral, are more complex and less direct than are changes in either
temperature or water vapor alone.”13 Luckily, for this chapter
we can draw on two recent multiauthored synthesis reports
prepared by the World Meteorological Organization Task
Team on Tropical Cyclones and Climate Change. The first
report focuses on observed cyclone detection and attribution
results.14 The second study15 assesses model-projected changes
in tropical cyclone activity for a 2C anthropogenic warming.
11
www.wsj.com/articles/the-rising-costs-of-hurricanes-1538222400.
12
www.pnas.org/content/116/48/23942.short.
13
www.nap.edu/catalog/21852/attribution-of-extreme-weather-events-in-the-
context-of-climate-change.
14
Knutson, Thomas, et al. “Tropical cyclones and climate change assessment:
Part I. Detection and Attribution.” Bulletin of the American Meteorological Society
2019 (2019). journals.ametsoc.org/doi/pdf/10.1175/BAMS-D-18-0189.1.
15
Knutson, Thomas, et al. “Tropical cyclones and climate change assessment:
Part II. Projected response to anthropogenic warming.” Bulletin of the American
Meteorological Society 2019 (2019). journals.ametsoc.org/doi/pdf/10.1175/
BAMS-D-18-0194.1.
To understand these studies we need to expand our

arsenal of understanding to encompass two sets of terms: detec-
tion and attribution; and type I and type II errors. In climate
change assessments, according to the IPCC, “Detection of change
is defined as the process of demonstrating that climate or a system
affected by climate has changed in some defined statistical sense
without providing a reason for that change,” while Attribution is
“the process of evaluating the relative contributions of multiple
causal factors to a change or an event with an assignment of
statistical confidence.”16 Some aspects of climate change are
much easier to detect than others. At both global and local scales,
for example, it is generally quite possible to detect increases in
temperature extremes. In Chapter 5 we assessed the global vari-
ations of heat extremes, where we found a strong detectable
signal, but even when we zoomed in to a local weather station
on Observatory Hill in Australia, we found a clearly visible
increase in air temperatures. Such increases are much harder to
see in extreme precipitation records and in time series of data
associated with cyclones. These very rare and complicated events
have very limited data, which results in small sample sizes, which
can make detection and trend analysis difficult.
Which brings us to the idea of type I and type II errors.
These kinds of errors are often discussed in the context of
medicine or legal matters. Type I errors involve accepting some-
thing as true when it is not. Type II errors involve incorrectly
rejecting a hypothesis when it is true. Different analytical
approaches tend to favor one type of error over the other. As
Elisabeth A. Lloyd and Naomi Oreskes point out,17 the most
common approaches used by climate scientists tend to minimize
type I errors. While rigorous, these approaches risk underesti-
mating the role of global climate change in extreme events and
missing connections that are really there. A series of recent
16
https://www.ipcc.ch/sr15/chapter/glossary/.
17
Lloyd, Elisabeth A., and Naomi Oreskes. “Climate change attribution: When is it
appropriate to accept new methods?” Earth’s Future 6.3 (2018): 311–325. doi.
org/10.1002/2017EF000665
“storyline”-based studies focus on minimizing type II risks.

According to LLoyd and Oreskes, “The storyline method is like
an autopsy: it gives an account of the causes of the extreme
event – the flood or storm – and can indicate whether climate
change was one of these causes.” The approach, however, may
overstate the role of climate change.
Lloyd and Oreskes argue that climate scientists should
be willing to look at climate extremes through both lenses, and
this is what the recent Task Team assessment of observed cyc-
lone changes does. In general, cyclone data tends to show rela-
tively few clear, well-understood trends. The Team assessment
finds little clear evidence, in the observations, for changes in
cyclone frequency, intensity, or number of landfalls. Such evi-
dence would involve being quite sure that such a result was
unusual when compared to natural variability.
However, when viewed from a type II (risk avoidance)
perspective, a large proportion of the team members agreed on
the following:
There has been a detectable observed northern migration of

the latitude of maximum intensity in northwest Pacific basin,
and anthropogenic forcing has contributed to this movement.
There has been a detectable increase in the global proportion
of tropical cyclones reaching category 4 or 5 intensity in
recent decades, and anthropogenic forcing has contributed
to this increase.
There has been a detectable increase in the global average inten-
sity of the strongest tropical cyclones since the early 1980s, and
anthropogenic forcing has contributed to this increase.
There has been a detectable long-term increase in the occur-
rence of Hurricane Harvey–like extreme precipitation events
in the Texas region, and anthropogenic forcing has contrib-
uted to this increase.
Interestingly, the type II framework also yielded some negative
findings. All the authors, for example, agreed that there has not
been a detectable increase in the frequency of land-falling hurri-
cane frequency since the late 1800s. So what emerges from this
study is that while it is still hard to detect changes in hurricane

behavior with a high degree of certainty, the balance of evidence
does suggest that the strongest storms may be becoming more
intense and associated with heavier precipitation. These results
are broadly consistent with the Team’s +2C model-based
analysis, which found that with “medium-to-high confidence
projections include increased tropical cyclone rainfall rates,
intensity, and proportion of storms that reach Category 4–5
intensity globally.”15
Putting this all in layperson’s terms, we might summarize
these results as: to date, things look concerning, and while it is
hard to be certain, model projections and the balance of observa-
tional evidence converge on an increased frequency of very strong
and very rainy hurricanes and cyclones. We have clear evidence
of the increasing cost of extreme hurricanes and cyclones
(Figure 7.4). Using the Munich Re Catastrophe Database, we
can also see a large increase in the number of hydrologic disasters
(Figure 7.5). According to the database, the number of expensive
($25 million or more in 2014 US dollars) hydrologic disasters has
increased from around 100 per year in the early 1990s to almost
400 per year in the 2015–2019 period. This increase is only
partly due to climate change; the expansion of human settle-
ments, especially along the coasts, puts more people in harm’s
way. But the rapid increase in risk is irrefutable.
It doesn’t take a cyclone to wreak havoc. Many floods
and disasters, in fact, are associated with non-cyclonic extreme
precipitation events. In 2017, numerous catastrophic flood events
occurred, most frequently in Asia.18 Approximately 27 million
people were affected by severe flooding in India, Nepal, and
Bangladesh. Extensive August rains in upstream areas of Nepal
and India helped fuel severe flooding in Bangladesh, with the
International Federation of Red Cross and Red Crescent
Societies estimating that 41 million people were affected, leading
to the loss of some 950,000 homes and 1,200 lives.19 Then in late
18
2017 Annual Disaster Statistical Review, cred.be/sites/default/files/adsr_2017.pdf
19
www.cnn.com/2017/09/01/asia/bangladesh-south-asia-floods/index.html.
Number of relevant hydrological events worldwide 1980–2018

400
350
300
Number of events
250
200
150
100
50
20 0
20 01
02
19 7
19 8
19 9
19 0
19 1
92
19 3
19 4
19 5
19 6
19 7
19 8
99
20 3
20 4
20 5
20 6
07
08
20 9
2010
2011
2012
2013
2014
2015
2016
2017
19 0
19 1
19 2
19 3
19 4
19 5
19 6
18
0
8
8
8
9
9
9
9
9
9
9
9
0
0
0
0
0
8
8
8
8
8
8
8
19
20
20
20
20
19
Figure 7.5 The number of large hydrologic disasters from the

Munich Re catastrophe database.
Source: natcatservice.munichre.com/
June and early July 2018, successive extreme rainfall events struck
Japan, triggering landslides and massive flooding. Some 158
Japanese people lost their lives, and costs to repair infrastructure
are estimated at around 270 billion yen (~US$2.4 billion).20 The
impact of the floods, along with the impact of a September earth-
quake in Hokkaido, have been linked to a ~1.2 percent decline in
Japan’s 2018 gross domestic product (GDP). Japan’s 2017 GDP
was US$4.872 trillion. One percent of US$4.872 trillion is
US$48.72 billion. Japan’s economic slowdown, in turn, has com-
bined with economic reductions in Germany and China to contrib-
ute to a potential global recession in 2019. Extreme precipitation is
having substantial impacts on national and global economies.
Storyline-Based Attribution Studies: An Example

for Houston
For the past several years, I have chaired sessions of
talks focused on climate extremes at the annual meeting of the
20
asia.nikkei.com/Politics/Japan-faces-2bn-price-tag-for-flood-rebuilding.
American Geophysical Union. Together with my co-chairs,

I have been honored to hear presentations from some of the
world’s most eminent climate scientists. In this section I draw on
talks by Dr. Michael Wehner from 2018 and 2019, because
these talks provide excellent examples of extreme event
attribution focusing on very harmful events. In general, this
work tends to follow the “storyline” approach to attribution.21
The storyline approach does not focus on questions like “will
climate change make hurricanes more frequent,” but instead
begins with a specific event and asks the question “did climate
change make this event more extreme.”
Dr. Wehner’s 2018 talk (“Causality and extreme event
attribution – Or, was my house flooded because of climate
change?”)22 began with a humorous homage to the “Rocky
and Bullwinkle” cartoon, delved briefly into the epistemo-
logical foundations of climate attribution, and then presented
some very powerful and important attribution analyses of
hurricanes Harvey, Katrina, Maria, Irma, and Florence.
Michael has kindly shared his presentation with me, and it is
definitely worth our while to work our way through it –
because it provides you with a front-row seat to the developing
science of climate attribution.
Dr. Wehner began by noting that the causality of cli-
mate extremes is inherently complex. Extreme events are rare
because multiple factors have to be aligned just right. In more
ways than one, questions like “does smoking cause cancer”
rhyme with “do CO2 emissions enhance extreme climate
events.” Borrowing from epidemiology, climate scientists often
ask one of two questions: (1) Has the probability of an event
changed? (2) Has the magnitude of an extreme event changed?
We answer these questions by considering two worlds: a world
21
Trenberth, K. E., J. T. Fasullo, and T. G. Shepherd (2015). “Attribution of
climate extreme events.” Nature Climate Change, 5, 725–730. doi.org/10.1038/
nclimate2657.
22
The next several paragraphs follow very closely the material presented by
Dr. Wehner.
with climate change and a world without industrialized humans.

If we are interested in probability, we can fix the magnitude of
an event and examine the change in the risk ratio of a given
extreme. The risk ratio is one common measure of such a
change: risk_ratio = [Probability in a world with climate
change]/[Probability in a world without climate change]. For
example, we might ask how climate change influences the prob-
ability that Houston would receive 50 inches of rain. Or, if we
are interested in changes in magnitude, we can fix the probabil-
ity and examine changes in quantity. We might ask, for
example, how climate change alters the magnitude of precipita-
tion in a one-in-twenty-year flood event.
Since we only have one planet, a planet with climate
change, we have to rely on models to approximate a world
without climate change. There are two basic approaches to
doing this: climate models and statistical models. Attribution
analyses using climate models typically follow an approach to
causal theory developed by Judea Pearl, a professor at the
University of California, Los Angeles. Just as in a medical trial,
climate simulations are used to produce one group that gets a
“drug” (climate change) and another “placebo” group that has
boundary conditions representing a world without climate
change. The first world is warmer and has increased green-
house gasses. The second world is cooler and lacks human-
induced increases in greenhouse gasses and aerosols.
Attribution analyses can also be based on Granger Causality,
named after Sir Clive Granger (1934–2009). These analyses
are based on statistical inference, and are often used in medi-
cine when explicit interventions are unethical. For example,
feeding children high doses of sugar to see if they develop
diabetes would be wrong. To explore such situations, we often
use statistical models. In general, Pearl statements of causality
are stronger than Granger statements, because we have carried
out controlled experiments that can reduce the possible influ-
ence of hidden covariates.
Dr. Wehner presented examples of both types of causal
analyses. Examining Hurricane Harvey (Figure 7.6), he reported
Figure 7.6 A reproduction of Figure 1 from Risser and Wehner’s

2017 paper, “Attributable Human-Induced Changes in the
Likelihood and Magnitude of the Observed Extreme
Precipitation during Hurricane Harvey.” Precipitation totals
(mm) for the Houston, Texas, region from August 25 to 31,
2017 based on station observations (left), interpolated station
observations (middle), and a radar station combination (right).
in a paper published in collaboration with Mark Risser.23 This

work used a statistical model (Granger attribution framework)
to suggest that climate change increased the probability of such
an extreme event by 350 percent (3.5 times). Examining changes
in magnitude, Risser and Wehner found that climate change
probably increased the magnitude of Harvey by something like
38 percent. In a world without climate change, Harvey might
have brought 31 inches of rain, not 50.
Wehner next presented another compelling study,
“Anthropogenic Influences on Major Tropical Cyclone
Events,” published by Christina Patricola and Michael Wehner
in the journal Nature.24 In this elegant “Pearl Causality” study,
the scientists used a sophisticated high-resolution convection-
resolving model to examine the behavior of hurricanes Katrina,
23
Risser , Mark D., and Michael F. Wehner. “Attributable human-induced changes
in the likelihood and magnitude of the observed extreme precipitation during
Hurricane Harvey.” Geophysical Research Letters (2017).
24
Patricola, Christina M., and Michael F. Wehner. “Anthropogenic influences on
major tropical cyclone events.” Nature 563.7731 (2018): 339. www.nature.com/
articles/s41586-018-0673-2
Irma, and Maria in worlds with and without climate change

(Figure 7.7). Big science, big money, big impacts; remember
these cyclones had costs totaling at least $446 billion and caused
more than 5,000 deaths (Tables 7.1 and 7.2). While Patricola
and Wehner’s study didn’t find big changes in wind speeds, they
did find significant changes in precipitation totals. The left
column of Figure 7.7 shows estimates of how much stronger
the rainfall rates of Katrina, Irma, and Maria were because of
human-induced warming of the oceans. The right column shows
the change between the simulated hurricane intensities, based on
the observed conditions, and simulated rainfall intensities at
the end of the twenty-first century if current emission patterns
continue.
As bad as Katrina, Irma, and Maria were, such cyclones
are likely to get substantially worse if we do not curtail our
emissions. Patricola and Wehner’s study, in fact, found large
(8–42 percent) increases in the end-of-the-twenty-first-century
rainfall intensity in eleven cyclones (Katrina, Irma, Maria, Bob,
Floyd, Gilbert, Ike, Iniki, Haiyan, Yasi, and Gafilo), assuming
continued rapid greenhouse gas emissions. We are already
seeing a rapid rise in the economic costs of hurricanes (Figure
7.3–7.5) as coastal development coincides with more intense
rainfall rates. The impacts of flooding and intense rainfall,
furthermore, can be nonlinear. Houston might be able to cope
with a doubling of rainfall if a hurricane brings 24 inches of rain
as opposed to 12 inches (61 cm versus 30.5 cm). The doubling
from 24 to 48 inches (from 61 to 122 cm) brings with it much
greater economic and societal destruction. What if future
Harvey-type storms bring 75 inches (191 cm) of rain?
Investing in reducing emissions now, while also making our
energy economies more efficient, resilient, and effective, will
likely be a smart investment.
Conclusion: Climate Change Is Hurting People Now

The Greek philosopher Solon (Figure 7.1) gave the city
of Athens its famous set of laws, which proved so popular that
Figure 7.7 An extract of figure 3 from Patricola and Wehner’s

Nature paper. The rows correspond to Katrina, Irma, and
Maria. The left-hand column shows the estimated change in
rainfall rates in a world with and without climate change. The
right-hand column shows results for the world as it was and
the world as it will be if we continue our current pattern
of emissions.
Table 7.1. Description of Harvey, Katrina, Irma, and Florence from

the US Billion Dollar Disaster Database (www.ncdc.noaa.gov/bil
lions/events). Losses have been adjusted for inflation and are shown
2018 US dollars.
Economic
Loss
Hurricane Description (billions) Deaths
Katrina Category 3 hurricane initially $168.8 1,833

August impacts the US as a category 1 near
2005 Miami, Florida, then as a strong
Category 3 along the Louisiana–
Mississippi coastlines, resulting in
severe storm surge damage
(maximum surge probably exceeded
30 feet) along the coasts, wind
damage, and the failure of parts of
the levee system in New Orleans.
Harvey Category 4 hurricane made landfall $130 89

August near Rockport, Texas, causing
2017 widespread damage. Harvey’s
devastation was most pronounced
due to the large region of extreme
rainfall producing historic flooding
across Houston and surrounding
areas. More than 30 inches of rain
fell on 6.9 million people, while
1.25 million experienced over
45 inches and 11,000 had over 50
inches, based on 7-day rainfall
totals ending August 31. This
historic US rainfall caused massive
flooding that displaced over 30,000
people and damaged or destroyed
over 200,000 homes and businesses.
Maria Category 4 hurricane made landfall $93.7 2,981

September in southeast Puerto Rico after
2017 striking the US Virgin Island of St.
Croix. Maria’s high winds caused
widespread devastation to Puerto
Table 7.1. cont’d

Economic
Loss
Hurricane Description (billions) Deaths
Rico’s transportation, agriculture,

communication, and energy
infrastructure. Extreme rainfall up
to 37 inches caused widespread
flooding and mudslides across the
island. The interruption to
commerce and standard living
conditions will be felt for a long
period as much of Puerto Rico’s
infrastructure is being rebuilt.
Maria was one of the deadliest
storms to impact the US, with
numerous indirect deaths in the
wake of the storm’s devastation.
Irma Category 4 hurricane made landfall $52 97

September at Cudjoe Key, Florida, after
2017 devastating the US Virgin Islands –
St John and St Thomas – as a
category 5 storm. The Florida Keys
were heavily impacted, as 25% of
buildings were destroyed while 65%
were significantly damaged. Severe
wind and storm surge damage also
occurred along the coasts of Florida
and South Carolina. Jacksonville,
Florida, and Charleston, South
Carolina, received near-historic
levels of storm surge causing
significant coastal flooding. Irma
maintained a maximum sustained
wind of 185 mph for 37 hours, the
longest in the satellite era. Irma also
was a category 5 storm for longer
than all other Atlantic hurricanes
except Ivan in 2004.
Total $445.5 5,000
Table 7.2. Large cyclone disaster statistics from the Munich Re

catastrophe database. Losses are shown in inflation-adjusted
2014 US dollars.
Economic Loss
Cyclone Date Location (millions) Deaths
Enawo March 2017 Madagascar $200 81
Ava January 2018 Madagascar $130 51
Okchi November 2017 India $400 89
Gaja November 2018 India $780 57
Roanu May 2016 Sri Lanka $500 96
Vardah November 2015 India $3,500 597
Titli October 2018 India $920 85
Damrey November 2017 Vietnam $650 108
Doksuri September 2017 Vietnam $480 14
Mujigae October 2015 China $4,000 20
Hato August 2017 China $3,600 21
Meranti September 2016 China $3,000 29
Nepartak July 2016 China $1,400 83
Soudelor August 2015 China $2,2000 26
Chan-Hoim July 2015 China $1,4000 1
Lionrock August 2016 South Korea $100 138
Jebi September 2018 Japan $13,000 11
Trami September 2019 Japan $4,000 4
Pam March 2015 Vanuatu $300 11
Cuba $3,500 12
Irma September 2017 Virgin Islands $9,000 5
Mathew Cuba $1,600

October 2016 Haiti $1,400 546
Maria Puerto Rico $65,000 2,975

September 2017 Dominica $1,200 31
Solon had to flee town or suffer reprisal. His many peregrin-

ations brought him before wealthy Croesus, whom he told,
“count no man lucky until his end is known.” Not long after
meeting Solon, Croesus was struck blind by the gods for hubris,
his son was killed in a hunting accident, his empire was crushed
by the Persians, and he ended his life on the business end of a
funeral pyre while he was still breathing. We modern humans
might very well be described as Croesian, but with the twist that
our demise stems from our rise and will also primarily descend
upon our descendants. Our tragedy follows the arc of Daedalus,
the famed Greek craftsman whose wings fatally propelled his
son Icarus too close to the Sun. As our human wealth explodes,
we spew greenhouse gasses into our atmosphere, supporting an
inevitable increase in both air temperatures and atmospheric
water vapor. This increase in water vapor increases the intensity
of extreme precipitation events. Hurricanes, cyclones, and
typhoons are the most extreme and complex class of extreme
rainfall events. While it is still difficult to detect changes in these
events in the observational record, attribution studies such as
those presented here for Harvey, Katrina, Irma, and Maria
imply that we are already paying a steep economic and humani-
tarian cost for our greenhouse gas emissions. Table 7.1 shows
the economic losses and fatalities from these five hurricanes. The
attribution results presented here imply that a substantial frac-
tion of these impacts resulted from human-induced warming.
While detailed data at a global scale is more difficult to obtain,
Table 7.2 shows similar statistics for large 2015–2018 catas-
trophes from the Munich Re database. Comprehensive attribu-
tion of these events is not available at present, but the balance of
evidence suggests that there has been a detectable increase in the
proportion of tropical cyclones reaching high levels of intensity,
and anthropogenic forcing has contributed to this increase
(Figure 7.8).
Between absolute zero and absolute hot, our fragile
Earth inhabits a small magic range of temperatures capable of
supporting water in its frozen, liquid, and gaseous phases.
Water transitioning from one phase to another supports the
Figure 7.8 Jacob Peter Gowy’s The Flight of Icarus (1635–1637).

en.wikipedia.org/wiki/Icarus#/media/File:Gowy-icaro-prado.jpg
development of tropical storms and cyclones. Cyclones collect

water vapor from vast expanses and focus it into punishing
rains, which in turn fuel intense winds and storm surge. The
latest science suggests that a warming of a degree or two can
lead to increases in tropical cyclone rainfall rates, intensity, and
the proportion of storms that reach Category 4–5 intensity and
enhance the severity of hurricanes like Harvey, Katrina, Maria,
and Irma. Continued warming to the end of the twenty-first
century may lead to further increases in the precipitation
intensities of very strong cyclones. The delayed impact of our

actions complicates Solon’s definition of a happy life. We may
die happy, but what of those who come after? Can we eat our
vegetables, be kind, pleasant, and respected, and still die happy
knowing that our actions set the stage for more catastrophic
weather for Earth’s future descendants?
CONCEPTUAL MODELS OF CLIMATE
8 CHANGE AND PREDICTION, AND HOW
THEY RELATE TO FLOODS AND FIRES
Introduction
Late 2019 brought catastrophic floods to East Africa.
October to December (OND) 2019 rainfall totals were among
the highest of the past nearly four decades in many areas in
East Africa. Flooding and other related disasters impacted
3.4 million people in the region.1 Nearly a million people were
impacted in South Sudan, a country where extreme food inse-
curity is impacting almost the entire nation. In five other coun-
tries (Ethiopia, Somalia, Kenya, Uganda, and Djibouti), between
250,000 and 570,000 people were impacted. In early 2020, fires
raged across Australia. As of January 4, 2020, hundreds of
wildfires in Australia had scorched more than 12 million acres2;
in New South Wales, the epicenter of the crisis, more than
480 million animals are estimated to have perished,3 according
to Sydney University ecologist Charles Dickman. Australia’s
1
OCHA January 2020 report, reliefweb.int/report/south-sudan/eastern-africa-
region-regional-floods-and-locust-outbreak-snapshot-january-2020.
2
www.usatoday.com/story/news/world/2020/01/03/australia-fires-map-animals-
evacuations/2803057001/.
3
sydney.edu.au/news-opinion/news/2020/01/03/a-statement-about-the-480-
million-animals-killed-in-nsw-bushfire.html.
165 / Conceptual Models of Climate Change and Prediction
environment minister estimates that 30 percent (more than

8,400) of the koalas living in New South Wales’ mid-north coast
may have perished.
Connecting these floods and fires is a very strong posi-
tive Indian Ocean Dipole event. These events occur when the
western Indian Ocean is very warm and the eastern Indian
Ocean is very cool.
Understanding how climate change contributes to
exceptional sea surface conditions will enhance our ability to
understand, anticipate, and predict climate extremes like
droughts and floods.
Stories matter. Stories shape our perception of the
world. Stories mold our interpretation of the past. They inform
our predictions of the future. In climate attribution studies,
“storyline” approaches are being used to describe precipitation
extremes.4 Climate change increases the amount of water vapor
in the atmosphere, which may enhance the severity of natural
phenomena such as hurricanes.5 As the atmosphere warms, we
can expect more moisture to converge during many of the most
extreme events.
We can consider extreme sea surface temperatures in a
similar way, and such considerations can guide climate hazard
predictions. Heat in the ocean, like water vapor in the atmos-
phere, is transported by currents. And like water vapor in the
air, heat in the ocean can converge and dissipate due to circula-
tion changes. But unlike the atmosphere, conditions in the
world’s oceans vary slowly. They persist. This persistence pro-
vides opportunities for prediction. Slow variations of sea surface
temperatures will influence wind and water vapor patterns,
offering opportunities for forecasts. When climate change con-
tributes to extremely warm ocean conditions, we have windows
4
appropriate to accept new methods?” Earth’s Future 6.3 (2018): 311–325. doi.
org/10.1002/2017EF000665.
5
Trenberth, K. E., J. T., Fasullo, and T. G. Shepherd. “Attribution of climate
extreme events.” Nature Climate Change 5 (2015): 725–730. doi.org/10.1038/
nclimate2657.
of opportunity for early warning. To make the most of these

opportunities, we need a clear understanding of how energy
moves through the Earth’s climate system. Understanding how
climate change is actually altering the spatial distributions of
energy and water vapor, week to week and month to month,
can literally help us save lives.
Contrasting “Bathtub” Warming with Energy

Convergence Patterns
Most of the “extra” energy provided by climate change
goes into the oceans, and this extra energy is accumulating very
rapidly.6 The energy stored in the ocean accounts for about
93 percent of the total increase in the climate system since
1971. Figure 8.1 shows a time series describing the total amount
of energy in the top 700 meters of the world ocean, expressed as
anomalies (differences from the long-term mean). I look at this
kind of data all the time, but I am yet again stunned by the rapid
rate of change we are seeing in the top layers of the oceans. It
took us about thirty years to go from an anomaly of about 3
1022 Joules in the 1960s to about +3 1022 Joules in the late
1990s. The current anomaly in 2019 is more than five times this
amount. Between 2014 and 2019, we see another 5 1022 Joule
jump. A 5 1022 Joule increase in energy is equivalent to the
energy released by about 12 million one-megaton nuclear bombs.
To understand and anticipate climate extremes, it is crit-
ical to understand how this “extra” heat behaves. Energy moves
around the oceans in complex ways. These movements can cause
energy to concentrate in some places and dissipate in others. Such
anomalies can produce climate hazards – persistent seasonal
climate anomalies that put people and ecosystems in harm’s
way. But when the ocean is exceptionally warm in one place, it
also tends to be cool somewhere else. Climate change will not
remove the natural processes that lead heat in the oceans and
6
www.climate.gov/news-features/understanding-climate/climate-change-ocean-
heat-content.
Figure 8.1 Global 0–700 m ocean heat content anomalies, based

on differences from the long-term average global ocean heat
content (1955–2006) in the top 700 meters of the ocean.7
water vapor in the atmosphere to accumulate in one region one

day and another region the next. Importantly for early warning
applications, the energy in the ocean changes very slowly because
it takes such a huge amount of energy to heat water. This pro-
duces sea surface temperature gradients that persist at weekly and
monthly time scales. These persistent gradients create opportun-
ities for climate predictions and effective life-saving interventions.
Focusing on a conception of climate change as the aver-
age of collections of climate simulations can cause us to miss
these opportunities.
Climate scientists often talk about “internal” and
“external” sources of variability. “External” variations are rep-
resented by averages taken across a large number of climate
simulations. These are variations that can be easily attributed
to changes in greenhouse gasses, aerosols, land cover, and solar
7
Global 0–700 m global ocean heat content anomalies, based on differences from
the long-term average global ocean heat content (1955–2006) in the top 700 m of
the ocean. data.nodc.noaa.gov/woa/DATA_ANALYSIS/3M_HEAT_CONTENT/
DATA/basin/3month/ohc_levitus_climdash_seasonal.csv.
radiation. “Internal” variations are then identified as the devi-

ations between the individual simulations and the ensemble.
While this decomposition is useful, especially when applied to
very broad area estimates, like the average temperature of the
globe, or low-frequency changes over decades, it can rapidly
lose its utility as we move to smaller and smaller spatial and
temporal scales. Here, we will refer the spatially similar
warming pattern produced by averaging many simulations as
“bathtub” warming. Imagine a bathtub in which all the water
warms at a uniform rate. While such a warming is a common
way to think about climate change, we need to recognize the
limitation inherent in this conception.
Global warming arises through the combined action of
an incredible number of individual photons. Each ray of light that
passes through a column of atmosphere with more greenhouse
gasses leaves a little more heat behind. But the oceans absorb,
assimilate, and translate this heat energy from place to place. So,
if we are interested in understanding climate extremes, then the
“bathtub” warming signal that we get from averaging across all
of our climate change simulations, across all the surface of the
Earth, provides a misleading “search pattern” – misleading in the
sense that it is unlike each of the individual simulations that were
actually produced by our climate change model – and misleading
in the sense that there is no physical mechanism in the models or
in the real world that accounts for an even flat “bathtub”
warming signal across all the world’s oceans. This “bathtub”
pattern arises from averaging many individual simulations, but
it does not actually exist from the perspective of actual physics.
According to the bathtub warming paradigm, sea sur-
face temperature gradients will be similar to those in the past,
since everywhere in the ocean will be warming at about the same
rate, at the same time. If, on the other hand, human-induced
warming increases the ocean heat transports associated with
natural climate variations, enhancing the temperatures of natur-
ally occurring, very warm sea surface temperatures, then what
we will likely see are more pockets of very warm sea surface
temperatures, and associated severe climate extremes – which
Figure 8.2 Two conceptual models of climate change. The top

schema depicts a “bathtub” warming scenario where two ocean
areas (one warm and one cold) each warm at the same rate. The
bottom schema depicts an energy convergence model in which
“extra” heat energy converges into a warm region of the ocean.
will provide opportunities for prediction. This is why philoso-

phy matters. We need a conceptual model that assumes that heat
energy will build up in the oceans and move around. This is
increasing the frequency of extremely warm sea surface tem-
peratures, which can induce climate extremes, droughts, and
floods, giving us an opportunity for early warning. We need to
be on the lookout for areas of extreme tropical or subtropical
sea surface temperatures that lie next to naturally occurring
regions of anomalously cool waters.
Figure 8.2 describes these competing conceptual models
of climate change. In the top schema, an evenhanded “bathtub”
warming affects a cold and warm region in the tropical and
subtropical ocean. Please note that in this conceptual model,
I am really referring to anomalously warm or cold regions of the
tropical and subtropical oceans, not the difference between

polar regions and the tropics. In the tropics, changes in sea
surface temperatures are closely linked to changes in atmos-
pheric circulation, and associated variations in rainfall and
air temperatures.
Under a bathtub interpretation, the greenhouse gas
warming (GW) would be about the same in both regions,
leaving the sea surface temperature gradient between the two
regions unchanged. In the bottom schema (the Energy
Convergence Model), extra energy from greenhouse gas
warming is transported by natural current variations into the
warm region. This can lead to stronger temperature gradients.
Temperatures in the cold region might resemble those in a world
with less global warming, and those in the warm region could be
much warmer than we might expect, given the average of many
climate change simulations.
Thought experiments can help us understand why such
conceptual models matter.
Let’s set the mood. Smooth Jazz plays in the back-
ground. You don a swimsuit and slide into the experimental
hot tub. Bath salts make the water pleasantly buoyant, and you
float contentedly on your back.
Experiment 1.
The bathtub warms slowly. You relax, adjusting to the
heat, and it is 2080 before you realize the water is
scalding hot. You are now a boiled frog.
Experiment 2.
You rest contentedly in a bathtub, gently dozing off
toward sleep, the soulful burr of sappy Jazz saxophone
pulling you slowly toward somnolence, when suddenly
after 6,000 soapy soothing seconds — ow! Burning hot
waters burn your feet. Then ow! Burning hot waters burn
your bum. Then ow! Burning hot waters excoriate your
hand. It is 2020 and you recognize that human-induced
increases in hotness are hurting people. Now. You act.
Experiment 2 describes one important manifestation

of climate change. Rather than a bathtub warming, we are experi-
encing limited areas of exceptionally warm water that move from
place to place to place, from month to month, lying alongside
ocean waters with relatively cool “normal” temperatures.
Despite this (hopefully) humorous description, this
topic is absolutely deadly serious. In the tropics, since 2015,
pockets of exceptionally warm ocean waters have destroyed
tropical coral reefs8 and contributed to severe droughts in
Eastern and Southern Africa.9 On December 5, 2016, such an
interpretation of climate change, together with a decade of
research and cross-validated statistical models, was used to
predict a high probability of back-to-back droughts in East
Africa.10 This work contributed to an effective and early
humanitarian response, helping prevent a repeat of the
2011 Somali famine.11 Here, however, we examine October–
November 2019 conditions, and how they related to flooding
and extreme air temperatures in East Africa and Australia.
Exceptional Flooding and Temperatures in 2019

In late 2019 (October and November), East Africa
experienced severe flooding while Australia experienced
extreme air temperatures and drought, conditions associated12
with the Indian Ocean Dipole (IOD),13 the gradient in tropical
8
journals.ametsoc.org/doi/pdf/10.1175/BAMS-ExplainingExtremeEvents2016.1,
Chapters 2, 9, 28.
9
Funk, Chris, et al. “Examining the role of unusually warm Indo-Pacific sea-
surface temperatures in recent African droughts.” Quarterly Journal of the Royal
Meteorological Society 144 (2018): 360–383. rmets.onlinelibrary.wiley.com/doi/
full/10.1002/qj.3266
10
Funk C., G. Husak, D. Korecha, G. Galu, and S. Shukla. (2016). Below normal
forecast for the 2017 East African long rains, December 5, 2016. blog.chg.ucsb.
edu/?m=201612.
11
Funk, Chris, et al. “Recognizing the famine early warning systems network: over
30 years of drought early warning science advances and partnerships promoting
global food security.” Bulletin of the American Meteorological Society 100.6
(2019): 1011–1027.journals.ametsoc.org/doi/pdf/10.1175/BAMS-D-17-0233.1.
12
www.bbc.com/news/science-environment-50602971.
13
www.bom.gov.au/climate/iod/.
Figure 8.3 Time series of western and eastern Indian Ocean sea
surface temperatures. Based on the NOAA Extended
Reconstruction version 5 data set. The selected western Indian
Ocean region stretched from 40E to 80E and 20S to 3N. The
eastern Indian Ocean region stretched from 80E to 110E and
15S to 3N.
sea surface temperatures between the western and eastern

Indian Ocean.14 When the western Indian Ocean is extremely
warm and the eastern Indian Ocean is extremely cold, eastern
Africa receives torrential rains, while Australia experiences
drought. In Figure 8.3, I display time series of sea surface
temperatures from the tropical western Indian Ocean (20S–
3N, 40E 80E) and the tropical eastern Indian Ocean
(15S–3N, 80E 110E) during October–November. These
boxes are a little different than the boxes typically used to define
the typical IOD index, and were selected to correspond with the
2019 anomaly pattern, which featured an exceptionally strong
Indian Ocean sea surface temperature gradient just south of
the equator.
14
Saji N. Hameed. (2018). The Indian Ocean Dipole, Oxford Research
Encyclopedias, February 2018. DOI: 10.1093/acrefore/
9780190228620.013.619. oxfordre.com/climatescience/view/10.1093/acrefore/
9780190228620.001.0001/acrefore-9780190228620-e-619.
As you can see from Figure 8.3, historically, the eastern

Indian Ocean has been substantially warmer than the western
Indian Ocean. In the tropical oceans, regions with warmer sea
surface temperatures tend to be much rainier than regions with
relatively cool sea surface temperatures. So, under normal con-
ditions, the eastern Indian Ocean and countries like Indonesia
receive copious amounts of rainfall. The western Indian Ocean
and countries like Somalia and Kenya receive relatively
little rain.
But Figure 8.3 indicates a very large deviation from
normal conditions in 2019. According to this data set, for the
first time since 1900, the western Indian Ocean was warmer, in
absolute magnitude, than the eastern Indian Ocean. The excep-
tionally warm western Indian Ocean sea surface temperatures
occurred alongside atypically cool eastern Indian Ocean waters.
We can observe an upward trend in the time series from both
regions. Given this trend, and the natural historical variability,
the exceptionally warm western Indian Ocean conditions in
2019 (and 2015) are not unexpected, but still exceptionally
warm. The 2015 and 2019 values were the warmest on record,
and much warmer (~+0.5C) than any previous values.
But the behavior of the eastern Indian Ocean time series
is also notable and important. It indicates that despite global
warming, eastern Indian Ocean temperatures can still be quite
cold – in this case, attaining values similar to those experienced
in the 1960s. These results align with the Energy Convergence
Model (Figure 8.2). We should expect pockets of exceptional
warmth accompanied by adjacent cool areas. While both the
western and eastern Indian Ocean are warming, there is no
magical warming process that is heating both ocean regions by
the same amount on a week-to-week or month-to-month basis.
Taken together, the combination of exceptionally warm
western Indian Ocean sea surface temperatures and atypically
cool eastern Indian Ocean inverted the Indian Ocean tempera-
ture gradient (Figure 8.3), producing a strong east-west pressure
gradient that drove winds and water vapor westward across the
equatorial Indian Ocean, feeding floods in East Africa but
Figure 8.4 Scatterplot showing the relationship between

1981–2019 October–November Climate Hazards center Infrared
Precipitation with Stations (CHIRPS) data from Uganda, Kenya,
Somalia and Ethiopia south of 11N and the east-west gradient of
tropical Indian Ocean sea surface temperatures.
reducing rainfall and increasing air temperatures over Australia.

Australia experienced the driest September-to-November spring
rains on record (since 1900), according to the Bureau of
Meteorology.15
Historically, there has been a strong relationship
between October and November East African rainfall and the
gradient between the western and eastern Indian Ocean sea
surface temperatures (Figure 8.4). The strength of this relation-
ship is primarily determined by behavior at the extremes. On the
left of this scatterplot, we find a cluster of strong negative
15
www.abc.net.au/news/2019-12-02/bureau-of-meteorology-declares-spring-
2019-the-driest-on-record/11755848.
Figure 8.5 Scatterplot showing the relationship between 2002

and 2019 October–November Australian MODIS Land Surface
Temperature data and the east-west gradient of tropical Indian
Ocean sea surface temperatures.
gradient values and very low East African rainfall values. These
years (2010, 1996, 1998, 1990, 1995, 2005, 2016) correspond
with very dry East African rainy seasons. On the right of
Figure 8.4, we find 1997 and 2019, two exceptionally wet years
associated with extremely weak Indian Ocean sea surface tem-
perature gradients. Figure 8.5 shows a similar but weaker rela-
tionship between satellite observations of Australian Land
Surface temperatures.
Appreciating how exceptional these climate conditions
were might have enhanced disaster preparedness and response.
The stories we tell ourselves matter. While accurate numerical
predictions are always an important aspect of effective early
warning, the “holy moly” dimension is important as well (“holy
moly” = an exclamation to express surprise or astonishment).
When we recognize that the tropical oceans are in an excep-

tional state, we are more likely to be on the lookout for poten-
tially related weather and climate extremes. This is why
identifying exceptional sea surface states can be so important.
Consider the following illustrative statement that could have
been made on November 5.
October sea surface temperatures in the western Indian

Ocean were exceptionally warm (27.5C), the warmest
on record, tying the previous extreme experienced in
2015. At the same time, October sea surface tempera-
tures in the eastern Indian Ocean were extremely cool,
on par with conditions experienced during the very
strong Indian Ocean Dipole/La Niña event in 1997.
Taken together, these anomalies produced an extremely
strong Indian Ocean Dipole event, historically associ-
ated with flooding in East Africa and drought
in Australia.
October rainfall totals in East Africa were exception-
ally wet, 162 millimeters, more than twice the
1981–2010 average and similar to the extremely wet
1997 rainy season. In Australia, October Land Surface
temperatures were the warmest on record. They were
also extremely warm in an absolute sense (45.8C or
114.4F), +2.9C warmer than the 2000–2018 mean.
Please note that these are estimates of the emission
temperature of the land surface, not 2-meter air tem-
peratures, which are typically a little cooler. Still, the
fact that an entire continent had a land surface tempera-
ture of 45.8C is truly exceptional and concerning.
Historically, for both East African rainfall and
Australian land surface temperatures, regional time
series of October and November conditions are very
well correlated. Correlations of about 0.8 indicate very
strong levels of persistence from one month to the next.
October Indian Ocean gradient values also exhibit rea-
sonably high levels of correlation with November East
African rainfall and Australian land surface tempera-

tures (correlations of R = 0.7 and 0.5). These relation-
ships, combined with an appreciation of the
exceptional October–November ocean state, support
predictions of exceptionally wet East African and
warm Australian conditions in November 2019.
Figures 8.6 and 8.7 show the regression-based forecasts
supporting these statements.
These relatively simple results illustrate how monitoring

extreme rainfall, land surface temperatures, and sea surface
temperatures can inform skillful one month outlooks of poten-
tial weather extremes.
Figure 8.6 Scatterplot showing predicted and observed

1981–2019 November East African precipitation. Based on
Climate Hazards center Infrared Precipitation with Stations
(CHIRPS) data from Uganda, Kenya, Somalia, and Ethiopia
south of 11N, October CHIRPS data and the October east-west
gradient of tropical Indian Ocean sea surface temperatures.
Figure 8.7 Scatterplot showing predicted and observed

2002–2019 November Australian MODIS Land Surface
Temperature values along with October forecasts based on the
east-west gradient of tropical Indian Ocean sea surface
temperatures and October Australian land surface
temperatures.
Climate Change?
Was the October–November 2019 Indian Ocean gradi-
ent event caused by climate change? Almost certainly not.
Historically, we see that there have been strong east-to-west
gradient variations for as far back as the data goes. But did
climate change increase the strength of the 2019 gradient event?
I would say almost certainly yes. Recent studies have used
climate change simulations to examine the frequency of strong
positive (warm western Indian Ocean) Indian Ocean Dipole
events in a warming climate. Even under a very modest 1.5C
Figure 8.8 Time series of the gradient between October and

November western and eastern Indian Ocean sea surface
temperatures. Signature flat gradient years associated with
flooding in East Africa are shown with squares. Strong negative
gradient years associated with East African droughts are
identified with triangles. The year 2019 is the only one since
1900 in which the west was warmer than the east.
warming scenario, these models indicate a doubling in the

number of extreme IOD events.16
Looking at the observed 1900–2019 sea surface tem-
peratures (Figure 8.3) we can see that the 2019 and 2015
western Indian Ocean temperatures were truly exceptionally,
about 0.5 warmer than even the warmest prior values. We
can see a strong tendency toward warmer conditions in the
western Indian Ocean, which is one of the fastest warming
ocean regions, with more warming projected by models.
These historically unprecedented warm conditions com-
bined with “naturally” cool eastern Indian Ocean conditions
combined to produce a stunning reversal of the equatorial
Indian Ocean sea surface temperature gradient (Figure 8.8).
16
Cai, Wenju, et al. “Stabilised frequency of extreme positive Indian Ocean Dipole
under 1.5 C warming.” Nature Communications 9.1 (2018): 1419.www.nature
.com/articles/s41467-018-03789-6.
Typically, the eastern Indian Ocean is warmer than the west-

ern region (Figure 8.3), so this gradient has a 1900–2019
mean value of 1C. Recent La Niña seasons like 2010 and
2016 were associated with dry October–November condi-
tions in East Africa. Before 2019, only one year (1997) was
associated with a very weak (near zero) gradient. According
to this data set, 2019 is the only year since 1900 in which
the western Indian Ocean was actually warmer than the
eastern Indian Ocean. The climate change–enabled excep-
tional warming in the west (Figure 8.3) combined with nat-
urally occurring cool conditions in the east to produce a
truly historic reversal of the Indian Ocean sea surface tem-
perature gradient and extreme East African precipitation
(Figure 8.6).
Conceptual Models Can Lead to Rapid

Transformation
How we think about climate change influences how we
see the world. If we are unduly wed to a fixed concept of climate
change as a bathtub-like warming of the land and oceans,
then we may miss the opportunities for prediction provided by
exceptional warming events. Western Indian Ocean tempera-
tures (Figure 8.3) and the associated inversion of the Indian
Ocean sea surface temperature gradient (Figure 8.8) provide a
compelling example.
There are times when averaging can obscure the truth.
Climate change involves much more than just the average trends
produced by large collections of climate change model simula-
tions. Extreme temperature gradients, such as those shown in
Figure 8.8, might arise within these simulations, and are antici-
pated by climate change simulations. Averaging across simula-
tions would obscure these patterns.
Consider Figure 8.9, which shows images of individual
snowflakes. Each flake represents a unique and beautiful mani-
festation of complex processes.
Figure 8.9 Pictures of individual snowflakes – each unique.
Now consider Figure 8.10, which shows the average of

all twelve snowflakes. In a very real way, this image is wrong. It
creates an image of a snowflake that never really existed.
Averaging men and women to find a “typical” human also does
not make sense. We would not average all the animals in a zoo
to create a “standard animal” and use that average as our
expectation for what the next animal we meet is going to look
like. In the same way, averaging across climate simulations can
Figure 8.10 An average of twelve snowflakes. Something that

never happens in nature.
obscure important extreme climate states, such as stronger El

Niños, La Niñas, and positive Indian Ocean Dipole events.
Weather events, like snowflakes, arise through the complex
nonlinear interplay of fluxes and forces. Averaging across these
patterns does not necessarily represent a “true” manifestation of
nature. While taking averages over large numbers of climate
change simulations may provide insights into the low-frequency
tendencies of sea surface temperatures, these averages may also
obscure the important behavior of the extremes.
For effective early warning, it is important to remem-
ber that there is no “external forcing” that acts to ubiquitously
heat the ocean surface in a consistent manner. The actual
greenhouse longwave radiation heating at the ocean’s surface
is quite small (~2–3 Watts per meter squared), so it is the net
accumulation of heat, and its tendency to concentrate in one
place or another that tends to drive many twenty-first-century
climate hazards.
Sharpening our ability to recognize extreme ocean tem-
peratures, and their predictable influence on climate, will
enhance our capacity to mitigate disasters. Such recognition
can also reveal the tragic moral consequences of our greenhouse
gas emissions, but may also lead toward a sustainable future.
We are what we pay attention to. This focus can alter

quickly, triggering rapid societal change. For example, during the
Renaissance, science grew up alongside the humanities and
advances in technology. Scholars and thinkers began looking at
the world as it was, not just the world as it was written. Art and
architecture helped trigger scientific revolution. Inspired by
Humanist thinkers and the art of ancient Greece and Rome,
fifteenth-century artists like Brunelleschi and Alberti invented
linear perspective, achieving convincing representations of depth.
Leonardo Da Vinci observed the old world through new eyes.
Strolling through a museum, like flicking on a light, we
can see the Renaissance’s rapid transition in how painters began
to see and render the human form. What was flat before became
full of life. Marvelous architecture and universities sprung up in
cities such as Florence, Bologna, Venice, and Milan. At the
University of Pisa, Galileo Galilei and his pupils studied “dis-
egno” (drawing) and later taught perspective and chiaroscuro,
the new Renaissance technique of using strong contrasts in
lighting to create striking representations of three-dimensional
forms. Galileo also pursued an avid interest in mathematics, and
in 1592 became a professor of geometry, mechanics, and astron-
omy at the University of Padua.
Like many Renaissance scientists, Galileo sought to
uncover God’s mysteries as written in the “book of the world,”
a source of truth as important as the Bible. Unique to Galileo,
however, was the newfound desire to express that knowledge in
the new language of mathematics. Or, to quote Galileo’s Assayer,
Philosophy [nature] is written in that great book which

ever is before our eyes – I mean the universe – but we
cannot understand it if we do not first learn the lan-
guage and grasp the symbols in which it is written. The
book is written in mathematical language, and the sym-
bols are triangles, circles and other geometrical figures,
without whose help it is impossible to comprehend a
single word of it; without which one wanders in vain
through a dark labyrinth.
Figure 8.11 Fra Angelico’s and Sandro Botticelli’s depiction of

the Madonna circa 1440 and 1481.
Once attuned to this new tome, humanity advanced quickly.

Isaac Newton (1642–1726) literally brought the heavens down
to Earth, showing that the same force of gravity that worked on
Earth could describe and predict the motions of the planets.
Modernity, science, and progress ensued, liberating mil-
lions and then billions from aching toil and mind-numbing
repetition. But now we face the painful impacts of our own
success as climate change contributes to more extreme climate
anomalies. But together we can face this challenge by getting
better at seeing the impacts of climate change, and getting better
at seeing the world around us.
History suggests we can do this. Consider, for example,
Fra Angelico’s depiction of the Madonna from circa 1440
and Sandro Botticelli’s painting from 1481 (Figure 8.11). Like
flicking on a light, we can see the rapid transition in how
painters began to see and render the human form. In just 40
years, the way artists perceived and presented the human form
blossomed. A similar focus on improving our perception and
prediction of climate extremes will help us survive the rest of the
twenty-first century. Keeping an eye out for areas of exception-
ally warm sea surface temperatures will help us provide effective
predictions. The “fingerprints” of climate change already form a
heavy hand. But climate-smart early warning systems may help
us mitigate some of the worst impacts.
CLIMATE CHANGE MADE THE
9 2015–2016 EL NIÑO MORE EXTREME
The story of this book began with a dream.
August 1995, Chicago, 2 AM. Air so wet and warm it
blends undistinguished from the waters of Lake Michigan, which
lap against my thighs. Facing the fullest moon, the city lights fan
behind me north, south, and west. Behind me waves lap gently
at the shore, and late-night cars roll along Shoreline Drive.
Laughing and drinking cheap American beer, we are young,
awake, and partying while others snooze away the night in fetid
dampness. Perfect moment. But then begins the strangest pull.
Gently against my legs comes a soft insistent tug, a soft push
against the back of my things. The push grows stronger. My first
thought: tide must be going out. Tug turns to tow, intensifies. My
second thought: Oh crap. . . Lake Michigan is not tidal.
I could have raced to a pay phone. I should have flagged
down a policeman.
Instead I leapt aboard my motorcycle, fired it up, and
raced to my office in Chicago’s south loop financial district.
Land-locked Lake Michigan is not tidal. The rapidly
receding waters foretold an imminent tsunami, a catastrophic
flood. From my office on the ninth floor I phoned in my stock
options, shorting the stock market. Sell sell sell, as disaster for
my fellow Chicagoans loomed. Man, was I going to be rich!
Upon waking, this dream lingered. Like a limping
shadow twin it followed me, haunting. Who had I become?
187 / Climate Change Made the 2015–2016 El Niño More Extreme
By the next fall I was leaving my job as a programmer.

Leaving my great group of friends, my rock band, and all my
stuff. With a hammock, sleeping bag, and cook pot strapped to
the back of my little Honda motorcycle, I headed west, like so
many before me.
Riding without a windscreen, I laughed into the wind
of America’s blue highways, spitting back the mosquitoes.
Heading south across Missouri, every day’s advance matched
the Sun’s autumnal retreat, so I seemed to race frozen in time,
each day bringing the same turning of the leaves from green to
gold. Fall itself seemed timeless, I moving with it, a southward
amber traveling wave, frost following right behind.
By day leaning into the wind, by night camping under the
stars, hammock slung between two trees, I came west the south-
ern way. In north Texas I pulled over and helped some cowboy
types kick out the makings of a lightning-strike forest fire beneath
the pines. In northern New Mexico I stopped to see my dad.
Delicious hikes high in the Sangré de Christo Mountains with the
first snows deep along the rushing streams. Delicious pozole
spiced with New Mexican chilies. Then came the beautiful drive
across northern Arizona. Then in California, a three-day back-
country trek in mysterious Joshua Tree National Park, filled with
twisted, blasted, barren beauty. Finally I arrived in Santa
Barbara. I was going back to school. I was going to become a
geographer. I was going to make maps of the future.
While I was studying at the University of California,
Santa Barbara (UCSB), I met two of the most important people
in my life: my wife (Sabina née Barattucci) and Jim Verdin.
Sabina got me excited about living life. Jim got me excited about
living work, about using satellites and climate information to
save lives and livelihoods. Sabina and I married in 1999.
I finished my PhD in 2002, and we brought into the world
two wonderful creatures, my La Niña and El Niño (Amelie
and Thelonious). I was working with Jim to support the
Famine Early Warning Systems Network (FEWS NET, www
.fews.net). FEWS NET provides unbiased, evidence-based analy-
sis to governments and relief agencies who plan for and respond
to humanitarian crises. The 2000–2001 southern African rainy
season had been poor, and then in the fall of 2002 an El Niño
developed. El Niño events occur when exceptionally warm sea
surface temperatures form in the eastern equatorial Pacific. They
typically produce droughts in southern Africa. At this time,
FEWS NET did not make much use of forecast information.
I had worked on this research as part of my PhD.
The time had come to refute my dark, options-buying,
stock market–shorting shadow from that dream gone by.
Already exhausted from twindom, I worked feverishly.
Alongside Jim and my Zimbabwean friend and colleague,
Tamuka Magadzire, I crafted a statistical forecast model and
accompanying report, “Forecasts of 2002/2003 Southern Africa
Maize Growing Conditions,” which called for “the possibility
of dry growing conditions in these regions, together with the
particularly dry forecasts for Northeastern Republic of South
Africa and Southern Mozambique, suggest an increased prob-
ability of a poor crop production season.” Taking a red-eye
flight to Washington, DC in December, I presented these results
and helped motivate an effective humanitarian response in
Zimbabwe, a country quickly eroding under the increasingly
erratic guidance of Robert Mugabe.
Fast forward to September 2003, six months after the
end of Southern Africa’s 2002–2003 rainy season. Poor rains
had contributed to widespread hunger and disruption in many
poor countries like Zimbabwe. FEWS NET reported, “The
2002/03 harvest is running out for most rural households, and
purchased foods are selling at prices that continue to escalate far
beyond the reach of the majority of poor households.”1
Ironically, for many regions dependent on summer rains, the
hungriest time of the year, or the lean period, arises when the
next season’s rains begin. The good news was that the World
Food Programme (WFP) was already distributing food aid to
almost two million Zimbabweans.
The timely distribution of such aid is a great human
accomplishment, an example of humanity being our best selves.
1
fews.net/sites/default/files/documents/reports/Zimbabwe.193"/>_200309en.pdf.
Reaching beyond barriers of race, religion, or nationality, we act

effectively to save the lives and livelihoods of desperately hungry
people. I thought, “This is way cooler than capitalizing on the
imminent inundation of America’s Windy City.”
To appreciate why this relief matters, and why I am so
deeply concerned about climate change intensifying droughts,
you need to understand both how horrible severe hunger can be
and how and why it remains so widespread. In disaster preven-
tion and relief, precision enables action. We spend a lot of time
quantifying how bad “bad” is. As a drought expert, one of my
specialties is producing the data and analyses underlying state-
ments like “Ethiopia‘s 2015 drought was the worst in 50
years”2 or “Southern Africa’s 2015/16 drought was the worst
in 36 years.”3 Food security experts (which I am not) have also
developed a means for quantifying extreme food insecurity. This
“integrated phase classification” supports the comparison of
food insecurity in diverse regions – supporting the comparison
of conditions in very different countries. The classes4 include:
1. Generally Food Secure; 2. Moderately/Borderline Food
Insecure; 3. Acute Food and Livelihood Crisis; 4. Humanitarian
Emergency; and 5. Famine. The determination of which class a
household or community belongs to is based on a number of
different criteria. This complexity is necessary, given the great
diversity of our societies and the multiple pathways to extreme
hunger. Under this classification scheme, famine is characterized
by at least one in five households facing an extreme lack of food,
more than 30 percent of children under five suffering from acute
malnutrition (wasting), and at least two people out of every ten
thousand dying each day from starvation and starvation-related
health complications.
One simple metric of severe malnutrition can help us
imagine the horrors of hunger. Young children are typically
2
www.telegraph.co.uk/news/2016/04/23/ethiopia-struggles-with-worst-drought-
for-50-years-leaving-18-mi/.
3
www.unocha.org/story/el-ni%C3%B1o-southern-africa-faces-its-worst-drought-
35-years.
4
fews.net/IPC.
Figure 9.1 The width of an extremely undernourished

child’s forearm.
deemed to be suffering acute malnutrition when their upper arm

circumference is less than 11.5 cm. That’s an upper arm diam-
eter of 3.7 centimeters, or 1.4 inches, about as big as this circle
(Figure 9.1).
Think about the hunger, the horrible suffering, involved
in a four- or five-year-old child with such a stick-thin forearm.
Now multiply that pain by tens of thousands of people. It is
pretty tough math. While there are multiple indicators linked to
integrated phase classes, such as mortality rates, coping strat-
egies, and food access, acute malnutrition in children under five
using a measure like upper arm circumference is one of the
easiest to measure and imagine. During a famine more than
30 percent of children under five may exhibit acute
malnutrition. At the integrated phase class 4, 15 percent of
children may exhibit acute malnutrition. At integrated phase
class 3, 10–15 percent children may exhibit acute malnutrition.
Hunger, El Niño, and the Southern Oscillation

El Niños can induce droughts5 in many regions, which
can lead to famine. For example, El Niño–related famines in
1876–1878, 1897, and 1899–1902 struck India, China, Brazil,
5
Glantz, M. H. Currents of Change: Impacts of El Niño and La Niña on Climate
and Society. Cambridge University Press, 2001.
Ethiopia, Korea, Vietnam, the Philippines, and New Caledonia.6

Some 19 million people in India and 10 million people in China
may have perished due to these El Niño–related droughts. In
1899–1900, Indian harvests in the Bombay Deccan, Karnatak,
and Gujarat provinces were only 4–16 percent of normal
according to the Bombay government’s “Report on the Famine
in the Bombay Presidency.”7 Yet Victorian Britain continued to
extract Indian cotton and wheat even as millions of children and
adults wasted, withered, and died (Figure 9.2). Indian author-
ities, held under rigid, inflexible ideological British rule, failed to
respond adequately to the extreme conditions. George
Nathaniel Curzon, First Marquess Curzon of Kedleston, served
as Viceroy. In his zeal to suppress Home Rule for India, Curzon
tightened press censorship, clamped down on education, and
pitted Hindu against Muslim. For Curzon, financing the Boer
war in South Africa was much more important than relieving
the distress of famine-stricken people of India. Writing at the
time, and quoting data from the Lancet, William Digby wrote,
“This statement by what is probably the foremost medical
journal in the world means that the loss of life thus recorded
represented the ‘disappearance’ of fully one-half a population as
large as that of the United Kingdom.”6
By 1899, India was supplying one-fifth of England’s
wheat, expanding cotton plantations, and financing Britain’s
Asian military interests as the Indian Army engaged in adven-
tures in Ethiopia, Sudan, and Egypt. As noted by Nobel Prize–
winning economist Amartya Sen, these crises were rooted in
chronic poverty and unjust economies, not just poor crop pro-
duction.8 As Davis notes, “During the famine of 1899-1900,
when 143,000 Beraris died directly from starvation, the pro-
vince exported not only thousands of bales of cotton but an
6
Davis, Mike. Late Victorian Holocausts: El Niño Famines and the Making of the
Third World. Verso, 2000.
7
Report on the famine in the Bombay Presidency, 1899–1902: Vol. I – Report,
dspace.gipe.ac.in/xmlui/handle/10973/38215.
8
Sen, Amartya. Poverty and Famines: An Essay on Entitlement and Deprivation.
Oxford and New York: Clarendon Press and Oxford University Press, 1982.
Figure 9.2 Famine relief at the Zenana Mission in Deori Panager,

on the outskirts of Jabalpur, India, March 1897.
incredible 747,000 bushels of grain.” Today, El Niño events still

pose large climatic risks, especially for relatively less well-off
countries in the tropics. But our ability to understand, predict,
and respond to extreme El Niños has increased substantially.
We will likely need all three dimensions of preparedness (under-

standing, prediction, and response) in the future, because cli-
mate change simulations anticipate more frequent and extreme
El Niños.9
One huge step forward for modern climate science – and
our ability to predict El Niño–related climate disasters –
occurred in the latter half of the twentieth century, when scien-
tists the likes of Jacob Bjerknes10 connected the dots between the
ocean phenomena of El Niños and the atmospheric variation
known as the Southern Oscillation (SO). El Niños occur when
the eastern equatorial Pacific becomes exceptionally warm. The
term “El Niño” was coined by fisherman off the coasts of Peru
and Chile during the seventeenth century. The appearance of
unusually warm waters brought increased fish catches, and these
beneficially warm waters were named “El Niños” or “Christ
Child” because they typically occurred around the month of
December. Much later, early in the twentieth century, the
British meteorologist Sir Gilbert Walker coined the term
“Southern Oscillation” to describe the inverse relationship
between sea level pressures in Darwin Australia and Tahiti.
Together, these two components describe the coupled ocean–
atmosphere phenomena known as the “El Niño–Southern
Oscillation”11 or ENSO for short. Warm East Pacific El Niño
waters occur alongside low atmospheric pressures near Tahiti,
while on the other side of the ocean the West Pacific cools and
air pressures near Darwin Australia increase. During a La Niña
event, the opposite occurs. In the aftermath of the terrible turn-
of-the-century Indian famines, Walker had been appointed as
Director General of the Indian Observatories in 1903.
A mathematician by training, Walker organized the Indian
9
Cai, Wenju, et al. “Increasing frequency of extreme El Niño events due to
greenhouse warming.” Nature Climate Change 4.2 (2014): 111–116. www
.nature.com/articles/nclimate2100.
10
Bjerknes, J. “Atmospheric teleconnections from the equatorial Pacific.” Monthly
Weather Review 97 (1969): 163–172.
11
Bjerknes, Jacob. “Atmospheric teleconnections from the equatorial Pacific.”
Monthly Weather Review 97.3 (1969): 163–172.
weather observatories and analyzed the accuracy of monsoon

forecasts. Walker pioneered the use of lagged correlations as a
means to make predictions. Such lagged correlations, as dis-
cussed in Chapter 8, remain important tools for early warning
systems. Looking for predictive anomalies that could be used to
forecast Indian droughts led Walker to analyze global weather
variations.12 This analysis led to his discovery of the Southern
Oscillation or the “swaying of pressure on a big scale back-
wards and forwards between the Pacific Ocean and the Indian
Ocean.”13
Under normal conditions, as discussed in Chapter 2, the
global Walker Circulation draws warm waters and moist air
into the tropical regions surrounding Indonesia, producing the
warmest ocean region on the planet. Winds blowing west across
the eastern Pacific have a cooling effect, resulting in a strong
gradient in average tropical sea surface temperatures, with the
eastern Pacific being typically as much as five degrees cooler
than the west.
But sometimes this pattern breaks down. The west-
blowing winds weaken, west Pacific sea surface temperatures
cool rapidly, while east Pacific sea surface temperatures increase
dramatically. These changes are caused by, and cause, large-
scale variations in the Indo-Pacific atmospheric circulation
(Figure 9.3). Torrential rains follow, drawing air up and shifting
the shape of wind patterns and storm systems around the world.
Under normal conditions a strong oceanic temperature gradient
(called a thermocline) stretches between Indonesia and the coast
of Peru. Subsurface waters are much warmer near Indonesia.
During an El Niño, a massive quantity of heat energy is shifted
east, warming the sea surface, triggering a quasi-global disrup-
tion of the Earth’s climate system.
12
Walker, G. T. “Correlation in Seasonal Variations of Weather, VIII:
A Preliminary Study of World Weather.” Memoirs of the Indian Meteorological
Department, 1923.
13
Katz, R. W. “Sir Gilbert Walker and a connection between El Nino and
statistics.” Statistical Science 17 (2002), 97–112.
Figure 9.3 Generalized Walker Circulation (December–

February) anomaly during El Niño events, overlaid on a map of
average sea surface temperature anomalies. Anomalous ocean
warming in the central and eastern Pacific (dark gray) helps shift
a rising branch of the Walker Circulation to east of 180, while
sinking branches shift to over the Maritime continent and
northern South America.
Source: NOAA Climate.gov drawing by Fiona Martin. From
“The Walker Circulation: ENSO’s atmospheric buddy,” by Tom
Di Liberto, www.climate.gov/news-features/blogs/enso/walker-
circulation-ensos-atmospheric-buddy
El Niños are the biggest natural disruption in our cli-

mate systems, shifting weather in many parts of the world from
their norm. To the west of the Pacific, El Niños can influence the
Asian, African, and Australian monsoons, producing droughts
in Indonesia, Thailand, India, Ethiopia, and Southern Africa. To
the east of the Pacific, in Central America, the Caribbean, and
northern South America, the chances of droughts are also
increased. And almost everywhere, El Niños can cause air tem-
peratures to jump upward as energy accumulated deep in the
ocean is released.
All of these potential impacts unfurled during the
2015–2016 El Niño. Because of my job as an early warning
analyst, I had a front row seat to much of this life-threatening
mayhem. During the northern hemisphere summer of 2015
(June–August), life-sustaining monsoon rains typically sweep
across the most populous regions of the globe, from northern

Ethiopia to India to Thailand. During 2015, our Climate
Hazards Center was watching Ethiopia closely. Since the
early 2000s, I have been very personally invested in developing
satellite and rain gauge–based data sets for monitoring
droughts in Ethiopia. The El Niño–related 1983–1985 famine
occurred when I was in high school, leading to more than
a million deaths. Many of you may remember Live Aid. In
1999, I joined the FEWS NET science team, hoping to do my
part to keep such a disaster from reoccurring. Being able to
rapidly quantify how bad a drought is going to be is one of the
key first steps in motivating an effective response. By combin-
ing satellite observations from space with long time series
of rain gauge observations, we were able to determine that in
the summer of 2015, northern Ethiopia had had its worst
drought in fifty years.14 FEWS NET reported on this
crisis,15 and a quick internet search of “Ethiopia’s worst
drought in 50 years” identifies articles on NPR, the New
York Times, Catholic Relief Services, Voice of America,
NBC, Time Magazine, CBC, UNICEF, Save the Children,
and BBC.
As the summer harvest failed, food prices rose and the
country plunged into one of the most severe food crises since
1984. In northeastern Ethiopia, plummeting livestock and milk
production conspired with rising food prices to push millions to
the edge of famine (Figure 9.4). Thankfully, massive inter-
national assistance began to arrive.16 The government of
Ethiopia assessed the required assistance as approximately
US$1.4 billion. USAID alone contributed more than half a
billion dollars in aid. The World Food Programme provided
targeted supplementary feeding interventions to more than
14
fews.net/sites/default/files/documents/reports/FEWS%20NET_WFP_
Ethiopia.200"/>%20Alert_20151204.pdf.
15
fews.net/sites/default/files/documents/reports/FEWS%20NET_Ethiopia%
202015%20Drought%20Map%20Book_20151217_0.pdf.
16
www.usaid.gov/sites/default/files/documents/1866/ethiopia_ce_fs07_03-30-2016
.pdf.
Figure 9.4 Many animals have died as a result of the drought.

Animal carcasses are common sight around the Fadeto and
Hariso Internally Displaced Persons centers, Siti region, Ethiopia.
The effects of a super El Niño are set to put the world’s
humanitarian system under an unprecedented level of strain in
2016 as it already struggles to cope with the fallout from
conflicts in Syria, South Sudan, Yemen, and elsewhere. Oxfam
estimates the El Niño weather system could leave tens of millions
of people facing hunger, water shortages, and disease next year if
early action isn’t taken to prepare vulnerable people for its
effects. It’s already too late for some regions to avoid a major
emergency. In Ethiopia, the government estimates that 10.2
million people will need humanitarian assistance in 2016, at a
cost of $1.4 billion, due to a drought that’s been exacerbated
by El Niño. Courtesy of Abiy Getahun/Oxfam
450,000 children and pregnant and lactating mothers experi-

encing acute malnutrition. By March almost one million people
required emergency water-trucking services to meet basic needs.
More than half a million metric tons of wheat were transported
to Ethiopia through the Port of Djibouti, and more than 11 mil-
lion Ethiopians required relief food assistance.
In 2015, due to the monster El Niño, the Asian summer
monsoon also fared poorly. In India, massive rainfall deficits
stretched across much of the country, impacting 330 million
people,17 a quarter of India’s population. A UNICEF report,

“When coping crumbles – A Rapid Assessment of the Impact of
Drought on Children and Women in India,”18 focused on an
analysis of 118 villages in drought-afflicted portions of nine
states: Maharashta, Bihar, Madhya Pradesh, Chattisgarh,
Telangana, Rajasthan, Jharkhand, and Odisha. This study
found that “nearly 90 per cent of the people faced severe food
shortage from January to June 2016.” Farm incomes were
destroyed, and most villagers relied solely on government sub-
sidies from the Mahatma Gandhi National Rural Employment
Guarantee Act.
In India and Pakistan an extreme (>45C or >113F)
heat wave led to more than 2,500 deaths19 (as discussed in
Chapter 5).
Thailand experienced extremely warm, dry weather and
crippling crop losses amounting to US$500 million in agricul-
tural losses.20 In July and October 2015, Indonesia experienced
severe drying and extreme temperatures. Some 1.2 million
people were classified as extremely food insecure, and a World
Food Programme assessment found severe reductions in
rice production.
Between June and December 2015, the El Niño grew
stronger, and reached its peak in January 2016, contributing to
yet another severe drought over Southern Africa – the worst
drought there in thirty-six years. Poor October 2015 to March
2016 rains resulted in huge crop production deficits and large
increases in food prices.21 Extreme hunger threatened more than
17
Guha-Sapir, D., et al. “Annual disaster statistical review 2016: The numbers and
trends.” Brussels, Belgium: Centre for Research on the Epidemiology of
Disasters, 2016.
18
reliefweb.int/sites/reliefweb.int/files/resources/pub_doc117.pdf.
19
Wehner, Michael, et al. “S16. The Deadly Combination of Heat and Humidity in
India and Pakistan in Summer 2015.” Bulletin of the American Meteorological
Society 97.12 (2016): S30–S32.
20
Christidis, Nikolaos, et al. “The hot and dry April of 2016 in Thailand.” Bulletin
of the American Meteorological Society 99.1 (2018): S128–S132.
21
www.sadc.int/files/3214/7806/7778/SADC_Regional_Situation_Update_No-3_
Final_011116_V1.pdf.
24 million people. Northern South America, Central America,

and Haiti also suffered severe droughts. Central America and
Haiti faced crisis levels of food insecurity.
2015–2016 Drought Testimonial

Name: Prosper Chirara; male, 28 years
Location: Yafele Village, Ward 16, Goromonzi District,

Mashonaland East Province, Zimbabwe
Date: December 12, 2017
My name is Prosper Chirara. I am 28. I come from Mutare on the

eastern border of the country with Mozambique. I have been
married for 5 years and we have two beautiful daughters.
I passed my Form Four (Ordinary Level) with 5 subjects.
Unfortunately, I could not afford to proceed with my education
(Advanced Level or vocational technical training) as I had wished,
since my mother could not afford the school fees. My father is
deceased.
I came to Harare (the capital city of Zimbabwe) 8 years ago
looking for work. I stayed for 2 years with an uncle without
finding a job in Harare, upon which I decided to relocate to
Goromonzi Business Centre, some 40km east of Harare, after a
cousin told me I could secure part-time jobs on construction
projects.
When I came to Goromonzi 5 years ago, there was a lot of
construction work. I easily got connected to some local builders
who provided me with part-time labor opportunities such as
digging foundations, working as an assistant builder mixing
mortar, providing bricks and stones, and fetching water. I would
move from one project to another quite easily and the opportun-
ities were readily available.
The most lucrative job I used to get was fetching water for
construction projects for people building in the area. Most people
constructing houses in the area have not yet sunk deep wells. So
they rely on hiring people to fetch water for them from a local
dam, river or community borehole. Fetching water to fill up a 200-
litre drum would earn $2. I would average 5 drums a day, earning
about $10 a day. With an average 5 days a week that would be

$50. Multiply by 4 weeks, that would be $200. $200 just for
fetching water! Then consider the other part-time jobs I would
do besides that. In a good month I could make between $250 and
$350 per month. That is more than what most professional civil
servants earn in a month!
Back then, I used to make good earnings that saw me looking
after my family very comfortably. My family was well-fed. I would
also send my mother money every month to buy food and other
basics. My mother is 57, and though she still able to work the
fields, she suffers from high blood pressure. I also could afford the
medication that she requires.
The first three years I was here (Goromonzi) things were really
really good for me. I also managed to buy some household items
(beds, set of chairs, bicycle, utensils, clothes etc.) for my family.
Then came the drought from 2014 to 2016! The year 2015 into
2016 was really bad. I will never forget that period. Things
changed completely, and my life changed. Even now as you see
me I have not recovered fully from the impacts of that drought.
That drought caused many households’ deep wells to dry up.
Water tables receded so much that the local dam dried off. The
local river stopped flowing.
Most people who were constructing houses stopped work on
their projects due to water challenges. Very few could afford to
hire water trucks or supply tankers. The long and ever-present
queues at the borehole ended construction projects. It would take
one up to 5 hours for one’s turn to come to draw water. And one
was only allowed two 20-litre buckets at a time. After which one
had to join the queue from the back and wait for another five or so
hours. People would literally not go to sleep at night in order to
take their stands in the queue. Some would wake up as early as
1 am or 2 am, upon which the queue would still be there.
My income dwindled so pathetically before my own eyes. This
was worsened by the cash shortages that began in early 2016.
Household incomes for casual laborers fell. For the few jobs
available, the rates offered were very low.
My friend, things went tough for me. My income was just a
trickle. Sometimes I would get a small job, but the “bosses” now
would take a long time to pay. I don’t forget the hundreds of
kilometers I walked from one boss to another pleading to be paid.

The savings that I had were cleaned up!
My family struggled from that drought. Putting food on the
table became a big challenge. I could not afford the most basic of
groceries. What made the situation worse was that prices for food
and other basic commodities were increasing regularly. My family
literally “starved”. I ended up borrowing heavily just to make sure
I got food for my family. I am still working to clear off some of the
debts I owed.
All the support I used to offer my mother just evaporated.
I could not afford to send my mother any money for groceries.
At least my mother was among the food assistance beneficiaries in
our village – so I knew she had some food from that source,
though not enough! What pained me most is that I would receive
calls from my neighbor informing me that my mother was not
feeling well, yet I could not afford to go and see her.
My friend, I am happy that the rains were good last season
(2016-17). I managed to have a good crop of maize, which stocks
we are still consuming and will likely last the next two months. It’s
not back to five years back yet, but I can make up to $50-$60 per
month. At least I am happy I can manage to send my mother
something, even if it’s small.
Prosper’s testimonial conveys so much about his

struggles. He is bright and hardworking, with an academic
prowess that could have got him into college if he and his family
could have afforded it. Like many young Africans, the lack of
opportunity led him to the capitol city, in this case Harare,
where he worked hauling water. Zimbabwe is a land-locked
nation just north of the Republic of South Africa. Once viewed
as a shining light of Africa’s future, thirty-plus years
(1987–2017) under the despotic (now deposed) rule of Robert
Mugabe reduced this country, which once hosted the greatest
Iron Age settlement south of the Sahara Desert, to an economic
shadow of its former self. Then came the El Niño–related
droughts of 2014/2015 and 2015/2016.
These droughts were part of an extensive swath of
multiyear dry conditions that covered a lot of our planet.
Figure 9.5 World Food Programme Analysis of multiyear El

Nino impacts. Graphic by author but based on a similar figure
from the WFP report El Niño: Implications and Scenarios for
2015–2016.
Analysts at the WFP analyzed the impacts of these droughts22

and highlighted the dangerous implications of the persistent
multiyear dry conditions associated with the El Niño–like cli-
mate conditions that existed from June 2014 through May
2016. To highlight these global impacts, I have produced results
similar to one of their figures (Figure 9.5). This image shows
two-year precipitation totals that begin in June 2014 and end in
May 2016. These totals are expressed as standardized precipita-
tion index (SPI) values. These values range from about 2 to +2.
Average values have a value of zero. Exceptionally wet or dry
regions will have values of less than 1.2 or higher than +1.2. In
2014–2016, areas of extreme persistent dryness stretched across
many tropical regions typically associated with dry conditions
during El Niños. Central America, the Caribbean, Amazonia,
Sudan/Ethiopia, southern Africa, India, and Southeast Asia
experienced several years of poor rain.
While conflict, civil unrest, and economic crises have
played an important role in increasing food insecurity, these El
Niño–associated droughts also contributed to a rise in both the
22
documents.wfp.org/stellent/groups/public/documents/ena/wfp280227.pdf.
number of moderately food-insecure and extremely food-

insecure people. Every year, the UN Food and Agriculture
Organization (FAO) releases a report on the state of global food
security. For several years (the years following the 2015/2016 El
Niño), these reports highlighted a concerning increase in the
number of undernourished people, with 820 million people
(more than 1 in 10) facing serious caloric deficiencies in
2019.23 Estimates of extreme food insecurity,24 focused on
identifying people facing crisis levels (or worse) of food insecur-
ity, show an 87 percent increase since 2015, with some 88 mil-
lion people likely in need of emergency assistance in 2020.
Eighty-eight million people is equivalent to the com-
bined populations of New York, Washington, Boston,
Chicago, London, Rome, Mexico, Tokyo, Delhi, Sydney,
Moscow, Bogota, Hong Kong, and Shanghai. If 88 million
people joined hands, they could circle the globe thirteen times.
Eighty-eight million people is 1 out of every 100 people. Out of
88 million extremely food-insecure people we might have 18 mil-
lion children, and about 2.7 million acutely malnourished chil-
dren. That’s an entire city of Chicago’s worth of extremely
hungry children: children with stick-thin arms.
Attributing Potential Climate Change Impacts on

the 2015/2016 El Niño
If climate change made the 2015/2016 El Niño more
intense, then climate change helped produce severe droughts
that directly contributed to massive increases in food insecurity
and human suffering. But did it? I absolutely believe the answer
is “yes” – but how one answers this question may depend on
your conceptual model of climate change. As discussed in
Chapter 8, if one simply defines climate change as the average
of a large number of climate change simulations, or the long-
term trend in some variable, then the answer might be “no.” For
23
www.fao.org/3/ca5162en/ca5162en.pdf.
24
fews.net/sites/default/files/Food_assistance_needs_Peak_Needs_2020_Final.pdf.
example, climate attribution analyses incorporating such a

framework have suggested that climate change did NOT inten-
sify the 2015 Ethiopia drought.25,26
On the other hand, however, if one begins with a con-
ceptual model that expects climate change to increase the
ocean’s heat content, and that this “extra” heat content will
converge in different regions at different times, then it follows
that human-induced warming did make the 2015/2016 El Niño
more intense, hurting a lot more people than it would have done
in the absence of human-induced climate change. Following this
basic line of reasoning, my colleagues and I have formally
assessed the contribution that human-induced climate change
made to the extreme 2015–2016 El Niño in three papers,27
formally linking this anthropogenic contribution to the associ-
ated droughts and food crises in Ethiopia and Southern Africa.
These food crises helped push more than 35 million people into
extreme food insecurity. It is fairly well accepted by scientists
that climate change will increase the frequency and magnitude
of future strong El Niños28: our research emphasized that such
impacts are happening now, hurting people like Prosper now.
You don’t have to just believe me; you can see it with
your own eyes. Figure 9.6 shows a time series of December-to-
25
journals.ametsoc.org/doi/10.1175/JCLI-D-17-0274.1.
26
www.worldweatherattribution.org/ethiopia-drought-2015/.
27
Funk, C., L. Harrison, S. Shukla, A. Hoell, D. Korecha, et al. “Assessing the
contributions of local and east Pacific warming to the 2015 droughts in Ethiopia
and Southern Africa.” Bulletin of the American Meteorological Society
(December 2016): S75–S77, doi:10.1175/BAMS-16-0167.1. journals.ametsoc.
org/doi/abs/10.1175/BAMS-D-16-0167.1; Funk C., F. Davenport, L. Harrison,
T. Magadzire, G. Galu, et al. “Anthropogenic enhancement of moderate-to-
strong El Niños likely contributed to drought and poor harvests in Southern
Africa during 2016.” Bulletin of the American Meteorological Society, 37 (2017):
S1–S3, DOI. 10.1175/BAMS-D-17-0112.2.www.ametsoc.net/eee/2016/ch18
.pdf; Funk C., L. Harrison, S. Shukla, C. Pomposi, G. Galu, et al. “Examining the
role of unusually warm Indo-Pacific sea surface temperatures in recent African
droughts.” Quarterly Journal of the Royal Meteorological Society (2018). doi.
org/10.1002/qj.3266.
28
Cai W. et al. “Increasing frequency of extreme El Niño events due to greenhouse
warming.” Nature Climate Change (2014). www.nature.com/articles/
nclimate2100.
Figure 9.6 Observed December–February equatorial East Pacific

sea surface temperature anomalies. Based on version 5 of the
NOAA Extended Reconstruction data set averaged over the
Niño3.4 region. The circles and stars mark out strong El Nino
events during the 1920s–1970s and from 1981 to 2016. The
horizontal bars denote the average of these events.
January equatorial East Pacific29 sea surface temperatures. The

December–February time period tends to be both the peak of El
Niño events and the heart of the Southern African crop-growing
season. These sea surface temperatures are commonly used to
measure the intensity of El Niño. The data are shown as “anom-
alies” or differences from the long-term average. When these
anomalies are greater than about +0.5C, we enter mild El Niño
conditions. When temperatures are above about +1.5C, we are
experiencing a strong El Niño.
29
These equatorial East Pacific sea surface temperatures were averaged over the
Niño3.4 region (170E–120W, 5S–5N).
Visually, there a few striking features of Figure 9.6. It

looks as if recent ocean temperatures may be more variable, and
that recent El Niño events (noted by very warm “spikes” in the
time series) may be substantially warmer. El Niños typically
occur once every five to seven years or so. We can isolate these
El Niño events by looking at the average of the top one-in-six
warm events over a given period of time. This corresponds to
the typical strength of a moderate to strong El Niño event. For
example, we can look at the average temperature of the six
warmest anomalies between 1981 and 2016. We can also look
at the average temperature of the ten anomalies between
1921 and 1980. These are the peaks in Figure 9.6.
The two horizontal bars in Figure 9.6 shows the average
temperature anomaly associated with these one-in-six year El
Niño events over the 1921–1980 time period and the
1981–2016 period. In this data set we find a huge difference
between “new” (1981–2016) and “old” (1921–1980) El Niños.
New El Niños are much warmer, about 0.8C warmer. While
other data sets give slightly different answers, they all agree on a
similar, and disturbing, story. We have seen a large and statis-
tically significant increase in the intensity of El Niño events. This
magnitude of change could transform a weak-to-moderate El
Niño into a strong El Niño. We will find out that that difference
can be supercritical – helping substantially increase the chance
of an El Niño–related drought.
But maybe the results shown in Figure 9.6 are just
due to chance? We have a very limited historical record.
Historically, sea surface temperature observations were taken
from ships, who used the measurements to help track their
location in the oceans. Before 1920, there were very few ships
traveling in the eastern equatorial Pacific. Strong El Niños are
also infrequent, which makes it statistically difficult to identify
changes in their extremes. Furthermore, we only have one
planet, so we only have a single time series to examine.
Using climate simulations, we can formally address
these types of questions using climate attribution methods.
Using climate models, we can create simulations over a much
Figure 9.7 Seventeen hundred years of standardized December–

February Niño3.4 sea surface temperatures from a preindustrial
simulation using version 1 of the Community Earth
System Model.
longer time period than a few hundred years. And, if we are

interested in the behavior of a world without climate change, we
can use long “preindustrial simulations.” These global ocean-
atmosphere model simulations have fixed low levels of green-
house gasses and aerosols. They are often used to characterize
“normal” behavior in a world without human emissions. These
simulations are typically run for hundreds or thousands of years
to give us a big sample to characterize 100 percent natural
preindustrial climate conditions. Figure 9.7 shows standardized
December–February El Niño sea surface temperature anomalies
from a 1,700-year climate simulation30 under preindustrial con-
ditions. We do in fact see many extreme El Niños in Figure 9.7,
so it turns out that it would be possible to see increases like that
30
Broadly following the work presented in our 2017 BAMS attribution paper,
www.ametsoc.net/eee/2016/ch18.pdf, I have used preindustrial simulations from
version 1 of the Community Earth Systems Model.
shown in Figure 9.6 purely by chance. We do in fact find

instances in a world without human-induced climate change
when a thirty-year period could experience a large increase in
strong El Niños. But these occurrences are very rare. With 1,700
years of modeled data to work with, we can quantify how likely
it might be to see a period of moderate El Niños followed by a
period of strong El Niños. It does turn out to be possible, but
highly unlikely (statistically speaking).
To estimate the likelihood of the observed behavior
we see in Figure 9.6 we can take the 1,700 years of data in
Figure 9.7 and break out all the possible 60-year-followed-by-
36-year time periods. Then for each of these time periods we
can calculate the change in one-in-six warm events between
the first 60 and the second 36 years. The dashed line in
Figure 9.8 shows the probabilities of changes in sequential
60- and 36-year moderate-to-strong Niño3.4 sea surface tem-
peratures under preindustrial conditions. The distribution is
symmetric and centered about zero. Sometimes we might see a
shift toward cooler conditions, simply due to chance.
Sometimes we might see a shift toward warmer conditions,
again simply due to chance. But over time the chances of
going up or down are symmetric, so, as you might expect,
this distribution is centered on zero. But overall, we can use
this distribution to assess what the typical range of natural El
Niño fluctuations might be, and this range is surprisingly
large. El Niños can be really warm (Figure 9.8, dashed line),
and runs of naturally occurring warm or cold events can
produce large changes in El Niño behavior just through
random chance. But when we plot the observed changes in
El Niño behavior (+0.8C) as a thick vertical line on
Figure 9.8, we see that the observed shift is extremely unlikely
given the distribution of simulated changes. The gray shading
in Figure 9.8 shows the region of the preindustrial distribu-
tion with changes as large as or larger than the observed
+0.8C change. This area covers only 3 percent of the distri-
bution, indicating that a +0.8C change would be possible but
very unlikely under natural conditions.
Figure 9.8 Preindustrial and historic distributions of changes in

one-in-six-year warm El Niño sea surface temperature anomalies.
The dashed line shows the preindustrial change distribution.
The vertical black line indicates the observed NOAA Extended
Reconstruction version 5 Niño3.4 temperature increase
(~+0.8C). The dark gray shading denotes likelihood given
preindustrial conditions (~3%). Thick non-vertical black line
indicates the probability distribution provided by the models
forced with historic greenhouse gasses and aerosols.
So, how likely might a +0.8C change be in a world with

human-induced climate change?
To answer this question, we can use a large set of forty
simulations driven with observed greenhouse gas and aerosol
changes.31 This large set of climate change simulations allows us
to derive a probability distribution function showing anticipated
31
Again, to broadly follow the work presented in our 2017 BAMS attribution
paper, I have used forty historic simulations from version 1 of the Community
Earth Systems Model.
changes between 1921–1980 and 1981–2016. Using a climate

change model, we create forty worlds similar to our world, and
subject these worlds to the same historic changes in greenhouse
gasses, aerosols, and solar radiation as our Earth. Now, instead
of just one time series like Figure 9.6, we have forty. This allows
us to assess the probability of a warming similar to that seem in
the observations, in a world with climate change. The distribu-
tion of these changes is shown with the heavy black line in
Figure 9.8. The climate change simulations predict an increase
in El Niño sea surface temperatures that is very similar to what
we see in the observations (about +0.7–0.8C). In other words,
a +0.8C change would be very likely in a world with
climate change.
Putting the two pieces of this story together, we can use
the preindustrial distribution (dashed line Figure 9.8) to say that
it appears extremely unlikely that the observed increase in El
Niño strength would have occurred in a world without climate
change. We can also use climate change simulations to suggest
that the observed magnitude of the increases appears to be
almost exactly what we are seeing in many of our climate
models and simulations (solid line in Figure 9.8). The climate
change simulations indicate that the observed warming is very
likely. Furthermore, the climate change models do a great job of
predicting the actual rate of warming.
This allows us to say with confidence that climate
change did indeed make the 2015/2016 El Niño more extreme,
and exacerbated droughts like those that occurred in Ethiopia
and Southern Africa. Think on that for a second. This attribu-
tion assessment has profound moral implications. Detailed
assessments by humanitarian agencies identified massive
increases in food insecurity and widespread economic disrup-
tion in Ethiopia and Southern Africa. Emergency assistance for
47 million people helped avert famine, but the total amount of
human suffering was immense, and climate change made it
worse. Which means we have all contributed to this suffering.
The 2015/2016 El Niño also induced droughts and temperature
extremes in many regions (Figure 9.5), and contributed to
extreme tropical sea surface temperatures, fishery destruction,

and coral bleaching.32
There is another important dimension to our attribution
analysis: prediction. By finding persuasive evidence that climate
change has already contributed to a large increase in the magni-
tude of strong El Niño events, we also deduce that climate
change will likely contribute to more severe El Niño events
and El Niño–related disasters in the (near) future. We are likely
to continue to see more frequent very extreme El Niño events,
similar to 2015/2016, accompanied by severe droughts in many
tropical regions (Figure 9.5).
I am a careful scientist. I will honestly tell you we can’t
be 100 percent sure that we will see a continued increase in
strong El Niño behavior in the next twenty years. But the odds
will not be forever in our favor. Six chambers, one bullet. Spin,
click: no boom. Double or nothing with your children’s future?
Do we have the right to gamble the future of millions of people
trying to get enough to eat? Do we have the right to gamble the
livelihoods of aspiring young men and women like Prosper
Chirara?
32
BAMS, Explaining Extreme Events of 2015 from a Climate Perspective.
journals.ametsoc.org/doi/pdf/10.1175/BAMS-ExplainingExtremeEvents2015.1;
Explaining Extreme Events of 2016 from a Climate Perspective.
journals.ametsoc.org/doi/pdf/10.1175/BAMS-ExplainingExtremeEvents2016.1.
BIGGER LA NIÑAS AND THE EAST
10 AFRICAN CLIMATE PARADOX
Standing in front of the mid-sized Washington, DC
conference room on May 19, 2016, I could feel my palms
sweat. I was pretty nervous as the thirty or so important
US Agency for International Development (USAID) analysts
and decision-makers settled into their seats. I had never given
this type of presentation to USAID before. I had flown out
from Santa Barbara to talk about how climate change had
made the 2015/2016 El Niño worse, and why they should all
now be worried about climate change making the 2016/2017
La Niña–associated droughts more intense. There were some
old USAID friends in the room – but there were also many
new faces.
My job for the next thirty minutes was to try to com-
press about a dozen research papers’ worth of insights into
thirty-two slides. These insights stemmed from an intense multi-
year, multipartner FEWS NET research effort that focused on
understanding the decline in the East African March–May rains,
and why USAID should be concerned about sequential back-to-
back droughts if a La Niña emerged, as was currently predicted
by many climate experts.
The stakes were high. The last time such a sequence of
droughts had happened was in October–December 2010, and
then in March–May 2011 a massive food crisis had broken out
213 / Bigger La Niñas and the East African Climate Paradox
in East Africa. In central and southern Somalia, the violent al

Shabab insurgency combined with drought, access limitations,
devastated crops, and skyrocketing food prices to create a per-
fect recipe for famine. More than 255,000 Somalis died, and one
out of ten children in southern Somalia lost their lives.
La Niña events occur when the equatorial eastern Pacific
Ocean becomes very cool, and the western Pacific becomes very
warm. Such events often follow strong El Niño events.
The 2011 drought was characteristic of what has been
called the East African Climate Paradox.1 This has been called
a paradox because while climate models predict increases
in March-to-May rains in East Africa, observational analyses
originally led by me2 and then confirmed by many others iden-
tify a disturbing tendency toward more frequent droughts.
Figure 10.1 provides an updated rainfall analysis describing this
tendency.3 Since 1999, there have only been a few wet seasons:
in 2010 and 2013, and the very wet 2018. Conversely, nine
years had poor rainfall: 1999, 2000, 2001, 2004, 2008, 2009,
2011, 2017, and 2019. This means that the “new normal” in
eastern East Africa has been a substantial or severe March-to-
May drought about every other year. Such sequential rainfall
deficits have had dangerous impacts, eroding resilience, eco-
nomic reserves, herd size, and human health.
On May 19, 2016, I didn’t know that the 2017 or 2019
rains were going to be so terribly poor, but I did know that
climate experts at NOAA and the International Research
Institute were predicting that a La Niña event was likely. And
I did know that climate change was likely to make that incipient
La Niña event more dangerous for East Africa. For many years
and in dozens of research papers, FEWS NET research scientists4
had been advancing a line of study connecting the increased
frequency of East African droughts to warming in the Indian
1
journals.ametsoc.org/doi/pdf/10.1175/JCLI-D-15-0140.1.
2 3
pdf.usaid.gov/pdf_docs/PNADH997.pdf. blog.chc.ucsb.edu/?m=201907.
4
Primarily Chris Funk, Park Williams, Andrew Hoell, Gideon Galu, Brant
Liebmann, Shraddhanand Shukla, and Laura Harrison.
Figure 10.1 A 1900–2019 time series of standardized

precipitation index (SPI) values, Based on Climate Hazards
Center precipitation data, for March–April–May rainfall in
central and eastern Kenya, central and eastern Ethiopia, and
Somalia. SPI values have a mean of 0 and a standard deviation of
1, so an SPI value of less than 1 indicates a potentially
dangerous drought. The thin vertical line depicts the date of my
May 2016 presentation to USAID.
and Pacific Oceans. As first advanced in 2005,5 our work sug-

gested that the basic mechanism for this drying involves increases
in rainfall over the warm tropical oceans to the east of East
Africa. Human-induced warming in these already very warm
waters can explain both the long-term decline of the East
African rains,6 and the emergent increased sensitivity to La
Niña climate conditions.7 In 2010, however, my research was
still primarily focused on warming in the Indian Ocean. Then, in
5
royalsocietypublishing.org/doi/full/10.1098/rstb.2005.1754.
6
Williams, P., and C. Funk. “A westward extension of the warm pool leads to a
westward extension of the Walker circulation, drying eastern Africa.” Climate
Dynamics 37.11–12 (2011): 2417–2435. https://link.springer.com/article/10
.1007/s00382-010-0984-y
7
Williams, P., and C. Funk. “A Westward Extension of the Tropical Pacific Warm
Pool Leads to March through June Drying in Kenya and Ethiopia.” USGS
Openfile Report 1199 (2010). https://pubs.er.usgs.gov/publication/ofr20101199.
2012, important new work by Brad Lyon and David DeWitt8

argued convincingly that the Pacific played a critical role in
driving East African droughts. This research helped lead to a
new set of FEWS NET papers emphasizing the interplay of warm
west Pacific sea surface temperatures and cool East Pacific sea
surface temperatures.9 This work moved us beyond understand-
ing the trend and into the realm of prediction. The eastern East
African rains were predictable, based on the interaction of west
and east Pacific sea surface temperatures. So – at my talk in May
2016 – I warned:
FEWS NET research suggests that both El Niños and La

Niñas may be becoming more intense, causing more extreme
impacts in some places.
For 2016/2017, a strong La Niña appears likely to be
enhanced by a warmer west Pacific.
The spatial pattern of La Niña impacts on East Africa inter-
sects with some very food-insecure regions of Somalia and
Ethiopia. La Niña–related drought could exacerbate existing
food and water shortages.
8
Lyon, Bradfield, and David G. DeWitt. “A recent and abrupt decline in the East
African long rains.” Geophysical Research Letters 39.2 (2012).
agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2011GL050337.
9
Hoell, A., and C. Funk. “The ENSO-related West Pacific sea surface temperature
gradient.” Journal of Climate 26 (2013): 9545–9562;Hoell A., and C. Funk.
“Indo-Pacific sea surface temperature influences on failed consecutive rainy
seasons over Eastern Africa.” Climate Dynamics (2013): 1–16. DOI: 10.1007/
s00382-013-1991-6; Liebmann, B., M. Hoerling, C. Funk, R. M. Dole,
A. Allured, et al. “Understanding Eastern Africa rainfall variability and change.”
Journal of Climate (2014); Shukla, S., A. McNally, G. Husak, and C. Funk.
“A seasonal agricultural drought forecast system for food-insecure regions of
East Africa.” Hydrology and Earth System Sciences Discussions 11.3 (2014):
3049–3081; Shukla S., C. Funk, and A. Hoell. “Using constructed analogs to
improve the skill of March-April-May precipitation forecasts in equatorial East
Africa.” Environmental Research Letters 9.9 (2014): 094009; Funk, C., A. Hoell,
S. Shukla, I. Bladé, B. Liebmann, J. B. Roberts, and G. Husak. “Predicting East
African spring droughts using Pacific and Indian Ocean sea surface temperature
indices.” Hydrology and Earth System Sciences Discussions 11.3 (2014):
3111–3136; Funk, C., and Hoell A. “The leading mode of observed and CMIP5
ENSO-residual sea surface temperatures and associated changes in Indo-Pacific
climate.” Journal of Climate 28 (2015): 4309–4329.
Figure 10.2 Pictures from Somaliland, taken by James Firebrace

and forwarded by James Magrath on November 15, 2016.
By October 2016, the La Niña had arrived. A cold

eastern Indian Ocean, a warm Indo-Pacific, and a cool eastern
Pacific created perfect conditions for a terrible East African
drought. Following the end of the 2016 El Niño, west Pacific
temperatures jumped to extremely high levels. On October 19,
the Climate Hazards Center issued a pessimistic forecast for the
October-to-December 2016 season via our group’s just-created
blog.10 On November 8 and November 16, we added blogs

updating this forecast and documenting the high likelihood of
poor pasture and crop conditions.11
Then on November 15, 2016, my friend John Magrath,
who worked with Oxfam, forwarded me the following from
James Firebrace, who had been bravely engaged in on-the-
ground field assessments of conditions in northern Somalia.
Figure 10.2 shows images from James Firebrace’s original e-
mail. John’s message went as follows:
From John Magrath

Fw: Drought in Somaliland - Hi Chris: we’ve just
received some disturbing photos þ information on the
humanitarian situation - thought you ought to know. . .
You getting any similar info? If not, can you raise con-
cerns with people you know? You’ll see these pics are not
sourced but they were sent to James Firebrace, a reliable
contact. The info about population movements is pretty
disturbing.
Best,
John
From James Firebrace
Richard, Beccy, Emma and Debbie
This is the latest I have picked up from Somaliland: The
Deyr rains completely failed in the eastern regions of
Somaliland. Many of those residing in the areas most hit
by the drought have migrated elsewhere. Those that were
able have migrated to an area called Oodale (HAWD),
where there were rains recently, its near Buuhodle. While
some have migrated all the way to Garawe (Puntland).
These two groups are easy to reach as they have congre-
gated in these two areas, and include roughly 300,000
people. These people are at a high risk, due to shortages
of food, water and shelter from a harsher than usual winter.
Even worse are those that did not have the means to move
out of the drought stricken regions, because they are much
10 11
blog.chc.ucsb.edu/?m=201610. blog.chc.ucsb.edu/?m=201611.
harder to reach. Please do your best to raise awareness

about the current situation in eastern Somaliland.
Also attached are the images sent to me . . .
Best wishes
James
As you can imagine, this really got my attention. So in

early December I worked very hard with my colleagues to put
together a statistical forecast for the March-to-June East African
long rains, which we released on December 5.12
On December 9, James Firebrace wrote to me from
Somalia:
Attn Chris Funk, Climate Hazards Group

Dear Chris
(cc relevant Oxfam staff )
You may recognize my name – John Magrath of Oxfam
sent on the earlier photos I had of the effect of the current
drought in Somaliland.
I have now been in Somaliland for close to 3 weeks (on
behalf of the President and National Drought Committee)
and have made a major tour of the most drought affected
areas to the east. There have been massive loss of livestock
in the areas I visited (over 70% of herd size across all the
interviews I made), and increasing dependence on (as yet
inadequate) water trucking and food aid. We are now
entering traditional dry season (Jilaal, Dec to March) and
looking ahead and one can foresee much greater loss of
livestock, as well as human hardship, to the point where
restocking herds is not going to be easy with serious impli-
cations for Somaliland’s economy.
Older people are insisting that this is the worst they’ve
seen – worse than what they call the Long Tail (Daba
Dheer) drought of 1974-5 which led to a massive resettle-
ment programme to the South, and worse than the ‘Red
12
blog.chc.ucsb.edu/?m=201612.
Dust’ (Sig’aas) drought of the mid 1950s. What makes it

different this time is the sheer extent of the drought and lack
of significant areas of accessible pasture to turn to (even
across the border in the Somali Region of Ethiopia). Those
limited areas that did obtain a little rain (and hence pasture)
were rapidly overgrazed by pastoralists seeking to save their
herds, with the result that their animals ended up in even
worse condition.
Forgive me for asking you a few questions. If you have
time to respond, even a short comment would help make
judgements for the best course of action looking forward.
1. How do you see the outlook for the next Gu rains (April to
June 2017) for Somaliland (and the Somali Region of Ethiopia
to the south much of which is largely the same region in
nomadic terms)? Can one at least get a feel for whether the
recent run of poor rains might now break?
2. Are you able to comment on the likely link between the
severity of this current drought and human-induced climate
change? I appreciate this is a complex topic. How far can one
justifiably go in saying that climate change is making the
situation worse?
...
I am back in UK on Monday and would happily set up a
Skype chat if you would like further briefing on my find-
ings. In any event I can copy you into my final report if that
is useful.
Kind regards
James Firebrace
Director, JFA
In answer to this first question, I was able to write back

and send James a link to our extensive December 5 blog post.
This post and several more that followed all called for a below-
normal outlook for the 2017 long rainy season. On January 16,
2017, FEWS NET and Somalia Food Security and Nutrition
Analysis Unit issued a joint alert calling out the high combined
risk generated by the poor 2016 and anticipated poor 2017 rainy
seasons.13 During this period, we were also networking and

sharing information with our partners in Europe at the Joint
Research Center, and in February the Food and Agriculture
Organization of the United Nations and the United Nations
World Food Programme issued a joint alert citing our outlook
for below normal rains in 201714:
Humanitarian partners should urgently prepare them-

selves to scale up their interventions in response to food
insecurity levels and food insecure population numbers
in Somalia and neighbouring regions, which are likely
to be at their highest levels since the 2010-2011 disaster.
Based on the drought impact evidence included in this
statement and on the more detailed food security and
nutrition information included in the multi-agency
assessment released on 2 February 2017, the following
actions are of highest priority:
provision of urgent and substantial food assistance to
help currently food-insecure households and support
the most affected livelihoods. Response activities to
be planned on the detailed IPC analysis outcomes
that were issued on 2 February 2017;
updating of emergency response for agro-pastoral
communities, intensification of advocacy and
resource mobilisation to address the impact of an
extended post-2016 “Deyr” harvest lean season. For
pastoral communities, provide updates of contin-
gency plans to face continued severe shortages in
livestock forage and drinking water;
continued close monitoring of the dry January to March
“Jilaal” season and the next “Gu” rainfall season to
inform decision-making on programming and targeting
increase awareness of the need for a regional
approach to address the effects of drought that are
13
www.fsnau.org/downloads/FEWS-NET-FSNAU-Somalia-Alert-2017-1-16.pdf.
14
fews.net/east-africa/somalia/special-report/february-21-2017.
becoming more frequent and intense, and to ensure

adequate access to populations in need of assistance.
In 2016 and early 2017, our understanding of the role played by

climate change – how extremely warm west Pacific sea surface
temperatures have triggered numerous frequent and predictable
droughts – helped motivate a timely and effective response in
late 2016 and early 2017. Additional humanitarian assistance
for about half a million people was already reaching Somalia in
December 2016, with the amount doubling again in January,
and again in February–March. The March-to-May rain season
was very poor, and the humanitarian crisis extensive, disruptive,
and painful. But we did not see the terrible outbreak of famine
that occurred in 2011: a disaster on that scale was averted.
We can see the influence of the improved 2016–2017
interventions in time series of sorghum prices in Baidoa, a major
food-insecure city in Somalia (Figure 10.3). Between July
2010 and June 2011, food prices in Baidoa climbed by more
than 300 percent, rising from about 5,000 shillings per kilogram
to 18,000 shillings per kilogram. In early 2017, humanitarian
assistance was helping stave off such a meteoric increase, so that
between June 2016 and July 2017, prices only increased from
Figure 10.3 Sorghum prices in Baidoa, Somalia.
around 6,000 to 10,000 shillings per kilogram. While our fore-

casts were just one piece of a complex early warning tableau,
understanding the dangerous relationship between human-
induced warming in the West Pacific and drought in East
Africa helped us provide valuable predictions.
Height and Wind Fields

This section will be more interesting if you know a little bit about
atmospheric height and wind fields. The broad strokes of their
relationships are pretty easy to understand. Scientists like to ana-
lyze atmospheric motions on “pressure fields.” These are the paths
of motion an airplane or hot weather balloon would follow. At the
surface of the Earth a barometer measures about 1,000 hecto-
Pascals (hPa) – this is the weight of air above you. The atmos-
pheric maps shown in Figure 10.4 are taken at 200 hPa – when
20 percent of the atmosphere is above you and 80 percent below.
As you glide along in a weather balloon, your actual height, in
meters, will increase or decrease depending on the temperature of
the air below you. This measurement gives us maps of atmospheric
highs and lows that we see on weather maps.
In the tropics, near the equator, winds blow from highs to lows.
In the extra-tropics, however, the spinning of the Earth works a
strange magic, turning winds to the left in both the Northern and
Southern hemispheres, so that they flow parallel to the height
fields.
A west–east equatorial sea surface temperature gradient induces
the first effect. A north–south extra-tropical gradient induces the
second. These work together to create extreme El Niños and
La Niñas.
Did Climate Change Make the 2017 East Africa

Drought More Intense?
A key aspect of our successful forecasts of the 2017
East African drought was an increased (and increasing) under-
standing of how climate change is interacting with natural La
Niña variations in a “positive” but dangerous way. Positive in

this sense means “amplifying,” i.e., interacting in ways that
increase the dynamic consequences of cool tropical East Pacific
sea surface temperatures. While the link between El Niño and
climate change is pretty straightforward – climate change makes
tropical East Pacific sea surface temperatures warmer – the La
Niña interaction is subtler. When cool east Pacific sea surface
conditions are surrounded by very warm west Pacific waters, the
influence of the cool east Pacific waters is increased. And this
can explain why recent La Niñas are more likely to produce East
African droughts.15
As shown in our 2018 paper in the Quarterly Journal of
the Royal Meteorological Society, warming of the western Pacific
has produced changes that increase the atmosphere’s response to
La Niña–like cool east Pacific sea surface temperatures.
Figure 10.4 shows the change in top-of-the-troposphere height
fields between “new” 1981–2017 La Niñas and “old”
1921–1980 La Niñas. The dark shading areas correspond with
higher atmospheric heights. These correspond to the “high-pres-
sure” patterns you might see on a weather map on your TV
evening news. The top of the troposphere is about 12 kilometers
up, at about the level where you see a thunderstorm flatten out
into an inverted anvil. As the atmosphere warms, it expands,
producing more high-pressure patterns. So almost everywhere in
Figure 10.4 we see dark shading indicating these increases. Note
the important exception – the subtropical eastern Pacific. Here,
where cool La Niña sea surface temperatures do their deeds, we
find a characteristic upper-level response. Twin upper-level cyc-
lonic circulation anomalies (marked with black ovals and L’s for
“lows”) describe the classic La Niña upper-level response to cool
tropical ocean conditions. Wind anomalies blow counterclock-
wise around these upper-level lows.
15
Funk, C., L. Harrison, S. Shukla, C. Pomposi, G. Galu, et al. “Examining the role
of unusually warm Indo-Pacific sea surface temperatures in recent African
org/10.1002/qj.3266.
Figure 10.4 Upper-level height changes in new versus old La

Niña events. From Funk et al. 2018. The ovals marked with L’s
indicate the upper-level cyclonic low-pressure anomalies
characteristic of La Niña-like climate conditions. The dark
regions over the northern and southern Pacific Ocean represent
areas where warming in the western Pacific has resulted in
increasing upper-level high-pressure anomalies. The interaction
of these low and high-pressure anomalies drives rapid upper-level
winds (arrows) that converge along the equator between 170E
and 150W. This enhances the Walker circulation and La La
Niña-like climate variations.
Now comes the interesting part. Around these lows, and

especially across the northern and southern Pacific, we find very
big (50 meter) increases in upper-level height fields. The intensi-
fication of these height fields is associated with long-term
warming influences in the western and extra-tropical central
Pacific. The warming ocean warms the air, which becomes less
dense and stretches, increasing upper-level height fields. This
region of high geopotential heights (shown with dark shading
in Figure 10.4) wraps hand-in-glove around the La Niña upper
cyclones, increasing atmospheric disruptions associated with La
Niña events. Winds (shown with arrows in Figure 10.4) respond
to the gradient (difference) between the low- and high-pressure
cells. Just to the north of the northern La Niña low and just to
the south of the southern La Niña low we see super-strong
height gradients and very rapid westward wind anomalies at
about 30N and 30S.
Such anomalies do not bring glad tidings to California

and the southern United States because they blow right into the
teeth of North America’s storm tracks. Traveling farther east,
these anomalies get turned toward the equator by rising height
fields to the east of China and Australia. This is caused by the
warming of the west Pacific. Converging and turned to the east
along the equator, we find a very strong enhancement of the
Walker Circulation. This enhancement is not just due to the
local influence of cool or neutral east Pacific Ocean tempera-
tures, but rather relies on the interaction of the tropical East
Pacific lows with the surrounding high pressures, and the overall
counterclockwise turning circulation patterns.
One way to think about Figure 10.4 is that the dark
shaded areas are doing what we expect a warming atmosphere
to do – expand, which makes the upper-level height field get
higher. When there is a La Niña, there is a natural “warming
hole” associated with La Niña’s characteristically cool east
Pacific waters. And over these naturally cool waters we find
natural upper-level lows (marked with ovals and L’s in
Figure 10.4). This interaction can explain a greater sensitivity
to naturally occurring La Niña conditions.
This probably seems like mega-geek-a-nomics to you.
But these insights helped save people’s lives in 2017. There
are many climate scientists who think that the East African
March–May rains are very hard to predict. They don’t appreci-
ate that climate change has upped the ante, making the region
much more susceptible, in predictable ways, to west Pacific
warming and East Pacific cooling. Our appreciation of this
sensitivity allowed us to start making successful predictions
of the 2017 March-to-June East African drought in early
December 2016,16 based on November 2016 data. These fore-
casts helped motivate effective and early intervention. When the
March and April rains in Southern Somalia failed again, assist-
ance was already arriving.
16
blog.chc.ucsb.edu/?m=201612.
Did Climate Change Make the 2017 March-to-May

Drought More Intense?
Did human-induced warming in the west Pacific
increase the probability or severity of the 2017 East African
drought?
Insights into recent East African droughts can be gained
by looking at simple “composites” of sea surface temperature
like those shown in Figure 10.5. Drought composites are pro-
duced by averaging conditions during drought years.
Figure 10.5 shows the typical state of the world’s oceans during
recent East African droughts. Areas with dark grey shading are
exceptionally warm. The black boxes over the western Pacific
define a region I refer to as the Western V region. Warm waters
in this region can produce a La Niña–like response associated
with dangerous dry conditions over Somalia, Kenya, and south-
eastern Ethiopia. March-to-June East African droughts are asso-
ciated with very warm Western V sea surface temperatures. The
Western V region is delineated with black boxes in Figure 10.5.
When the Western V region becomes very warm, you see circu-
lation anomalies like those plotted in Figure 10.4.
The time series shown in Figure 10.6 display the magni-
tude of Western V sea surface temperatures in 2017. The top
Figure 10.5 Standardized sea surface temperature anomalies

associated with recent March–June East African droughts.
Figure 10.6 Time series of observed and modeled March-to-June

sea surface temperature anomalies from regions of the Western V
area. The top panel shows values for the equatorial West Pacific.
The bottom panel shows values for the Northwest Pacific. Solid
lines correspond to NOAA Extended Reconstruction version 5
observations. The dashed lines show the average warming
predicted by the large Phase 5 Coupled Model Intercomparison
Project (CMIP5) climate change ensemble.
3.High* 4.Low
4
5.Low* c c
d d
b
e a 1.High
6.High* 2.Low*
Dry/Hot* Hot* Cold
Figure 10.7 La Niña modifications to the Walker Circulation

along with climate change enhancements.
panel shows conditions in the West Pacific. The bottom panel

shows conditions in the Northwest Pacific. Both regions were
incredibly warm, with either the warmest or second-warmest sea
surface temperatures on record. Estimates of the influence of climate
change, shown with dashed lines in Figure 10.6, indicate that
human-induced warming contributed about 0.6C of warming.
Warm ocean conditions in the equatorial western and
northwestern Pacific act together to create more extreme La
Niña–like climate conditions and droughts over East Africa
and Southwestern North America. To describe this dangerous
partnership, I have once again adapted a Walker Circulation
schematic developed by NOAA17 to explain how La Niña
enhances the global east–west Walker circulation, producing
drier conditions over East Africa (Figure 10.7). While describing
these responses will require a few paragraphs, remember that
the lives and livelihoods of people like Eregae Lokeno Nakali
and Aita Eregae Nakali (Figures 1.1 and 1.2) may depend on
these insights.
The numerals (1–6) in Figure 10.7 indicate characteris-
tic environmental conditions during La Niña events. Several of
these annotations have been marked with asterisks to indicate
17
https://www.climate.gov/sites/default/files/Walker_LaNina_2colorSSTA_large
.jpg.
climate change enhancement (discussed below). The associated

atmospheric responses are noted with letters (a–f ).
La Niña events begin with cooler-than-normal East
Pacific sea surface temperatures (1) and warmer-than-normal
West Pacific sea surface temperatures (2). Cool conditions over
the East Pacific produce higher atmospheric pressures because the
air there is denser. Over the West Pacific the air is less dense.
During La Niña events the low-level winds (a) flow from the
denser East Pacific toward the less-dense region around Indonesia.
These warm moist winds (a) and the warm sea surface
temperature conditions (2) warm the air above Indonesia. The
warmer atmosphere becomes less dense and stretches upward
(b), producing higher than normal pressures in the upper atmos-
phere (3). This anomalous upper-level high (3) rises above the
East Pacific (4) and East Africa (5). Like a skier on the top of a
hill, air parcels at this location (3) have extra potential energy.
The release of this energy sets up east–west upper-level wind
anomalies blowing away from Indonesia toward both East
Africa and the East Pacific (c). Over East Africa, in the upper
atmosphere (5), these wind anomalies converge, slow, and sink
(d), producing hot, dry high-pressure conditions at the surface
(6). This surface high-pressure cell can also slow the westward
flow of moisture into East Africa (e). So La Niña conditions can
produce dry conditions in East Africa by both producing stable
sinking atmospheric conditions (d) and disrupting the onshore
flow of moist air from the Indian Ocean (e).
Climate change enters into our story by enhancing West
Pacific sea surface temperatures (Figure 10.6) and the associated
West Pacific upper-level high pressure cells. These correspond to
2 and 3 in Figure 10.7. Climate change has produced strong
upward trends in these fields. When naturally occurring cool
East Pacific sea surface temperatures (1) or low East Pacific
upper-level pressures arise during “new” La Niñas, these
normal conditions lie alongside anthropogenically enhanced
conditions in the West Pacific. This combination increases all
of the La Niña atmospheric responses (a, b, c, d, e), leading to
hotter and drier conditions over East Africa.
Did climate change contribute substantially to the

2017 Western V sea surface temperatures and associated
extreme East African drought?18 To answer this question we
can contrast two large collections of climate change simulations.
One collection has been forced with observed greenhouse gas
and aerosol concentrations. This represents a world with cli-
mate change. Another collection has been forced with condi-
tions representative of preindustrial conditions. This represents
a world without climate change. In both cases, the simulations
were developed using version 1 of the Community Earth
Systems Model.
This collection of simulation results can be expressed as
probability distribution functions (PDFs). Figure 10.7 shows
these results. One PDF, shown with a dashed line, represents
the distribution of sea surface temperatures in a world without
climate change. Another PDF displays the probability distribu-
tions in a world with climate change.
The difference in equatorial West Pacific temperatures
between the world with and the world without is fairly modest:
+0.5C (Figure 10.8, top) but statistically significant and
dynamically important. This region is so warm that a half-a-
degree Celcius shift can enhance the West Pacific precipitation
substantially, increasing rising atmospheric motions near
Indonesia (b in Figure 10.7). Interestingly, the shift for the
Northwestern Pacific is substantially larger: +1.1C
(Figure 10.8, bottom). There is a tendency for some scientists
to discount warming in this region as natural decadal variabil-
ity, but both the observations and the climate change simula-
tions indicate substantial human-induced warming (bottom
panel Figure 10.6, bottom panel Figure 10.8).
Would you like to know how likely the observed West
Pacific sea surface temperatures would have been in a world
18
This analysis is based on the 2018 study for the annual BAMS Explaining
Extreme Events special issue, Funk C. et al. (2018) Examining the potential
contributions of extreme ‘Western V’ sea surface temperatures to the 2017
March-June East Africa Drought (2018), Bull. of Am. Met. Soc., S1–S6,
DOI:10.1175/BAMS-D-18-0108.
Figure 10.8 Western V SST attribution analyses. The left

panel shows probability distributions for Equatorial West Pacific
sea surface temperatures for a preindustrial world without
climate change and current world with climate change. The
observed sea surface temperature value is shown with a vertical
black line. The right panel shows a similar plot for the
Northwestern Pacific.
without climate change? One way to answer this question is to

calculate the Fraction of Attributable Risk (FAR).19 FAR is
defined as 1 – [Probability in a world without climate change]/
[Probability in a world with climate change]. We can calculate
FAR values using the observed 2017 conditions in the western
and northwestern Pacific (vertical black lines in Figure 10.8).
Thus the FAR values were 1. In plain language, this means that
such an event would have been impossible in a world without
human-induced climate change.
But did the exceptionally warm Western V sea surface
temperatures contribute to a 2017 drought over East Africa?
Our 2018 attribution study explored this question by contrast-
ing rainfall simulations from a world with and world without
climate change. Using PDFs like those shown in Figure 10.8, we
performed a formal attribution study by assessing the probabil-
ity of drought in East Africa. In the world with climate change,
Western V warming produced a La Niña–like response that
dried East Africa. This means that human-induced warming in
the Western V region enhanced the probability of drought over
East Africa. This formal attribution study contrasted a world
with and world without climate change. These PDFs indicate
that climate change doubled both the magnitude and probability
of the drought. So: warm Western V conditions doubled the
probability of the 2017 East African drought.
Putting the Pieces Together

These results may seem complicated, and might even
seem somewhat pedantic to you, but please let me explain why
they are anything but. Food crises in Kenya, Ethiopia, and
Somalia are strongly related to the rainfall deficits, made more
intense by climate change. FEWS NET estimates indicated that
between 5 and 10 million Somalians, 2.5–5 million Ethiopians,
and 0.5–1 million Kenyans faced acute food insecurity in early
19
Allen Myles, “Liability for Climate Change,” Nature 421.6926 (2003): 891–892.
2018.20 The results presented here demonstrate that climate

change contributed substantially to this severe food insecurity.
Good science can go beyond the explanation of past events: it
can support prediction and remediation. The science presented
here informed accurate climate forecasts that helped inform and
motivate effective early humanitarian responses. The 2016
October-to-December East African drought was predicted in
mid-October.21 The following March–May 2017 East African
drought was predicted on December 5, 2016.22 This climate
information, along with field reports and satellite observations,
was used collectively by many international partners who
worked together to provide effective early warning
(Figure 10.3).
By warming the tropical Indo-Pacific, climate change is
having dangerous impacts, increasing the frequency of East
Africa‘s droughts. But by understanding how these exception-
ally warm sea surface temperatures induce droughts23 we can
improve our ability to anticipate these impacts, and do some-
thing to alleviate at least some of the human suffering. By
combining appropriate conceptual models of climate change
with the tremendous power of the latest generation of climate
models, we can provide predictions of La Niña-related droughts
at very long leads. As another La Niña formed in 2020, we
pushed our forecast window back to June, warning of likely
back-to-back October-to-December and March-May eastern
East African rainfall deficits.24
20
fews.net/sites/default/files/documents/reports/July%202017_FAOB_final.pdf.
21 22
blog.chc.ucsb.edu/?m=201610. blog.chc.ucsb.edu/?m=201612.
23
rmets.onlinelibrary.wiley.com/doi/full/10.1002/qj.3266
24
https://blog.chc.ucsb.edu/?p = 757 https://blog.chc.ucsb.edu/?p = 937 Funk,
Chris. “Ethiopia, Somalia and Kenya face devastating drought.” Nature
586.7831 (2020): 645–645.
FIRE AND DROUGHT IN THE
11 WESTERN UNITED STATES
November 2018, Saturday afternoon, deep in the
brushy chaparral in the woods behind my house.
Temperatures sizzle. Sweat beads and drops from by body.
Bone-dry, steel-strong, the thin manzanita branches resist and
rebound, sending the blade leaping back with deadly intent.
Rebuked, violently returned and twisted, the flat of the machete
bounces off my forehead.
Some things will really make you stop and think, and a
machete bouncing off your forehead is one of those things. It
was pretty amazing. There were very few areas on that machete
that were not razor sharp or lined with serrated saw teeth. I was
totally isolated and alone, clearing a trail in the woods behind
my house at the top of San Marcos Pass, which sits above Santa
Barbara. I had been incredibly lucky.
The week prior had been surreal. Undergraduates on
the University of California, Santa Barbara campus strolled
about barely clad in 90F mid-winter weather. The vegetation
around my house was very dry. Though the weather was hot,
rapid winter winds whipped our California coast. Drought
combined with blazing temperatures had sucked all the water
from the ground and vegetation. When air temperatures
increase under dry conditions, relative humidity values decline,
235 / Fire and Drought in the Western United States
Figure 11.1 El Machete peligroso.
leading to drier vegetation. Plants’ photosynthesis transforms

carbon dioxide and sunlight into life-sustaining sugars. Drawing
in the CO2, however, means that plants must open tiny little
doors in their leaves. These doors, called stomata, also let out
water vapor as they draw in carbon dioxide. So, plants in arid
and semi-arid regions face an existential question. How can they
eat (photosynthesize) without dying of thirst, becoming weak
and susceptible to the increases in disease and pest invasion that
often accompany severe water stress? For California, and much
of the southwestern United States, the answer to this question is
“not very well.” Since 2010, the U.S. Forest Service estimates
that some 129 million trees have died in California’s national
forests “due to conditions caused by climate change, unpreced-

ented drought, bark beetle infestation, and high tree densities.”1
Concerned by the dry warm conditions, on November
6, 2018, I participated in a radio interview on a local radio
show, “Community Alert – Not If, but When – Positive
Preparation for Disaster”2 – hosted by my friends and neighbors
Mike Williams and Ted Adams. Mike and Ted are very active
within the volunteer firefighter community. I periodically talk on
their show, providing a climate perspective to current events.
The Southern California Geographic Coordination
Center tracks climate and fire risk conditions, and in their early
November Outlook3 they described California’s current situ-
ation. California had been exceptionally dry over the past
twelve months, especially Southern California, with 30 percent
less rain over much of the state, and 50 percent in the
southernmost areas.
Fire risk, however, is also a function of temperatures,
which act to increase the general level of aridity by increasing
the atmosphere’s water-holding capacity. Warmer air tempera-
tures create more atmospheric “pull” for water vapor, drawing
proportionally more moisture from plant stomata and bare soil.
For example, my friend Park Williams and his coauthor John
Abatzoglou in their paper, “Impact of anthropogenic climate
change on wildfire across western US forests,”4 show that
“human-caused climate change caused over half of the docu-
mented increases in fuel aridity since the 1970s and doubled the
cumulative forest fire area since 1984.” Figure 11.2 is taken
from their paper. There is a very strong and clear relationship
between increasing aridity and forest fire area in the Western
United States.
Being the data geek that I am, I wanted to include an
updated time series of US wildfire extent in this chapter.
1
https://www.fs.fed.us/psw/topics/tree_mortality/california/documents/
DroughtFactSheet_R5_2017.pdf.
2
https://www.wildlandresidents.org/community-alert/.
3
https://gacc.nifc.gov/oscc/predictive/weather/index.htm.
4
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5081637/.
Figure 11.2 Abatzoglou and Williams’s Figure 1 from “Impact of

anthropogenic climate change on wildfire across western US
forests.” The x-axis shows a measure of aridity, taking into
account both precipitation and temperature. The vertical y-axis
shows the total area burned in each year since 1984.
National wildfire extent totals for 1980–2019, obtained from

National Interagency Fire Center, are shown in Figure 11.3.
The extent of US wildfires is increasing rapidly. I have
included on this plot a simple regression-based trend estimate
for 2030. If the current trend in wildfire extent continues,
typical wildfire extent values in 2030 will be around 9.9
million acres – similar to the worst years to date (10 million
in 2017 and 10.1 million in 2015). This is a very crude
approximation, but the general point is valid – we should
expect wildfires to get bigger as the western US warms and
atmospheric water demand increases.
Figure 11.3 Annual US fire extent, in acres. Trend shown with a

dashed line (R2 = 0.37), along with a regression estimate for
2030 wildfire extent (dark square at upper-right).
Figure 11.4 shows some data that I discussed with Mike

and Ted on November 6, 2018. I was very worried about the
potential for yet another very bad California fire winter. This
was based on both my direct experience of the dry vegetation
conditions and my examination of the current climate condi-
tions. Figure 11.4 shows ten-year averages of annual October–
September rainfall, air temperatures, and standardized Palmer
Drought Severity Index (PDSI) values. The top panel shows
rainfall, which exhibits strong inter-decadal variations and per-
haps some of the driest conditions on record. The middle panel,
however, shows an unequivocal shift toward much warmer
conditions. Since the start of the twentieth century, we see an
annual increase in annual temperatures of about 4 degrees
Fahrenheit. Note that between approximately 2005 and 2018,
California experienced a very rapid 1-degree increase. These
rising temperatures are associated with increased moisture
losses due to increases in evaporation.
Figure 11.4 Ten-year averaged annual rainfall, air temperatures,

and standardized Palmer Drought Severity Index values for the
South Coast Drainage (shown in the map on the bottom left).
One measure of the balance between precipitation

supply and loss through evaporation is the Palmer Drought
Severity Index (PDSI). Negative PDSI values indicate an increase
in evaporation losses relative to precipitation. According to the

PDSI (bottom panel Figure 11.4), we have seen a major shift
toward drier conditions in my neighborhood, the South Coast
Drainage area. This large decrease in relative moisture supplies
creates conditions conducive to fires. As these conditions persist
into the early part of California’s rainy season, the risk of fire
can actually increase as more windy weather creates a potential
for red-flag conditions (red-flag conditions indicate extreme
fire risk).
For California, conditions conducive to recent droughts
and fire are often accompanied by a Ridiculously Resilient
Ridge – a moniker coined by Daniel Swain5 to describe a per-
sistent high-pressure cell over the northeast Pacific region. This
high-pressure ridge can block storms and produce dry hot sub-
siding air over California. As pointed out by the work of Swain
and others, this ridiculously resilient ridge can help explain
California’s recent demise into drier-hotter conditions. As con-
text, Figure 11.5 shows a long time series of annual November–
October upper-level heights near the Ridiculously Resilient
Ridge. These data extend from 1948/49 through 2018/2019.
Since 2012, six out of seven years (2013, 2014, 2015, 2016,
2018, 2019) have exhibited above-normal “ridging” that have
helped produce dry warm weather in California. Such were
conditions in early November 2018, when I experimented with
deep-forest auto-lobotomy techniques (Figure 11.1) and then sat
down to talk with Mike and Ted. On air, we shared our collect-
ive concern – that winds would whip up before the winter rains
came. These winds combined with extremely dry fuel condi-
tions, set the perfect stage for fire.
Unfortunately, these fears were verified. Over the next
few days much of California experienced wind gusts of up to
55 miles per hour accompanied by extremely low relative
humidity (humidity less than 10 percent). At sunrise on the
morning of November 8, the Camp Fire, California’s deadliest
5
Dan has a fantastic California weather blog: https://weatherwest.com/.
Figure 11.5 Annual November–October upper- level height

anomalies over the northeast Pacific.
and costliest conflagration, broke out in Butte County6

(Figure 11.6). At eight o’clock in the morning the fire entered
the town of Paradise. Wind speeds that day reached 50 miles per
hour, and the fire spread very rapidly. By November 10, more
than 6,700 structures were destroyed; by the 20th, almost
20,000 structures had been destroyed.7 Ultimately, eighty-six
people died. The fire was the costliest disaster of 20188 ($16.5
billion), according to the international insurance agency Munich
Re, and super-deadly. Two more fires broke out on November
8 – the Woolsey and Hill Fires – in Los Angeles and Ventura
counties, just south of Santa Barbara, ultimately destroying an
additional 1,647 structures.
6
https://en.wikipedia.org/wiki/Camp_Fire_(2018).
7
http://cdfdata.fire.ca.gov/ .
8
https://www.usatoday.com/story/news/2019/01/08/natural-disasters-camp-fire-
worlds-costliest-catastrophe-2018/2504865002/.
Figure 11.6 Landsat 8 image of the Camp Fire in northern

California, taken on November 8, 2018. By NASA (Joshua
Stevens) – NASA Landsat 8 Operational Land Imager.
In California, the 2017 and 2018 wildfire seasons were

massive, deadly, and costly. In both years, the fire season
stretched from April to December, and across much of the state.
The overall size of these wildfires has risen dramatically from
the late 1990s9 (Figure 11.7). The US National Oceanic and
Atmospheric Administration tracks disasters with costs over a
billion dollars.10 Here is their synopsis of the 2017 and 2018
California fire seasons.
2017
A historic firestorm damages or destroys over 15,000
homes, businesses, and other structures across California in
October. The combined destruction of the Tubbs, Atlas, Nuns,
and Redwood Valley wildfires represents the costliest wildfire
event on record, also causing forty-four deaths. Extreme wildfire
9
Data from https://www.fire.ca.gov/stats-events/.
10
https://www.ncdc.noaa.gov/billions/events/US/1980-2018.
Figure 11.7 Total size of California wildfires. Values for

2019 were not available in early 2020.
conditions in early December also burned hundreds of homes in

Los Angeles. Numerous other wildfires across many western
and northwestern states burn over 9.8 million acres exceeding
the ten-year annual average of 6.5 million acres. Montana in
particular was affected by wildfires that burned in excess of
1 million acres. These wildfire conditions were enhanced by
the preceding drought conditions in several states.
2018
In 2018, California has experienced its costliest, deadli-
est, and largest wildfires to date, with records back to 1933. The
Camp Fire is the costliest and deadliest wildfire, destroying more
than 18,500 buildings. California also endured its largest wild-
fire on record: the Medincino Complex Fire, burning over
450,000 acres. Additionally, California was impacted by other
destructive wildfires: the Carr Fire in Northern California and
the Woolsey Fire in Southern California. The total 2018 wildfire
costs in California (with minor costs in other Western states)
approach $24 billion – a new US record. In total, over 8.7
million acres has burned across the US during 2018, which is
Figure 11.8 California’s seven-year average Palmer Drougth

Severity Index values. Data from March-to-February data used to
allow for an update through January 2020.
well above the ten-year average (2009–2018) of 6.8 million

acres. The last two years of US wildfire damage has been unpre-
cedented, with losses exceeding $40 billion.
As with all natural disasters, human systems and human
decisions play an important role in setting the stage for crises.
In recent years, Californians have built more homes in high fire
risk regions. These homes have been connected by a fragile
system of electric wires, prone to sparking fires in high winds.
Teasing out the complex contributions of human choice, natural
climate variability, and climate change is beyond our scope and
my expertise. But we can look at low-frequency statewide aver-
age Palmer Drought Severity Index values (Figure 11.8) to get a
sense of how the balance between moisture supply and atmos-
pheric moisture demand may be changing. These values are
based on a March–February annual average to allow for an
early 2020 update. A seven-year averaging period was chosen
to emphasize the persistent recent dry and warm conditions –
the dry conditions leading up to the exceptional 2017 and
2018 fire years. According to the PDSI estimates, the statewide
seven-year drought was the most severe on record, and
these dry conditions persist. California precipitation pattern is
notoriously complex, but the progression toward warmer tem-

peratures and greater aridity seems very likely.
My friend Gary Eilerts mother lived in Paradise, California. Gary

kindly provided the following testimonial. The Camp Fire was
incredibly destructive, destroying more than 18,000 buildings
and killing 86 people.
“It took two angels to get me out of Paradise”
. . . said Laura Eilerts.
She was speaking late in the day on November 8, 2018, the day
of her evacuation from the Paradise due to the Camp Fire. On that
day, Grandma Laura, as we know her, lost her house, all her
belongings, a lifetime of her own paintings, some of her neighbors,
her community of friends, her church, and the entire town she’d
lived in for 18 years.
Laura is 91 years-old. She has a lung condition which requires
her to be on heavy oxygen 24/7. She can’t feel her feet, uses a
walker to move about very slowly, and loses her breath very easily.
She’s been told recently that she doesn’t have too long left to live,
so she’s been meticulously planning every step on that road. The
family was told to come and get anything they want before
January, and she had planned out everything that needed to
happen next, and even after that, so that she could remain where
she is happy, in her home in Paradise.
She’s very practiced with fire evacuations from her Paradise
home, having experienced more than a few. She has good
friends, the Lash family, twenty minutes away, down the ridge
into the Central Valley of California, in Chico, who have shel-
tered her several times when fires have menaced. At her home,
she is constantly alert at a very visceral level to the dangers that
brushfires present in her normally-dry city in the foothills of the
Sierra Nevada mountains, and all it takes is a simple outdoor
barbecue in her neighborhood to raise her antennae, focus her
attention, and take her away from whatever she was doing at
the time.
So, on a warm and dry Wednesday, November 7th, when the
news gave warnings about winds coming the next day, she began
mentally preparing for just another evacuation. And on a bright
Thursday morning, November 8th, somewhere between seven and
eight in the morning, when the first news came of a brushfire

north-east of Paradise, she began acting out a well-practiced
routine.
Over a very compressed period of a few minutes, she received a
call from the Lashes in Chico, who told her they were immediately
sending their son to evacuate her and would drive her back home
when it was all over. Shortly afterwards, paramedics arrived at the
door ready to evacuate her. But she turned them away, saying “I
have friends coming up to take me away; go help others who need
it more”. Not too long after that, she found out that her friends’
son was unable to come and get her, as there were some roads
already blocked due to the fires, and others were swamped by
evacuating residents. She decided to call the paramedics back, but
they told her that it would take a while to come and get her; there
were now just too many on the list for pick-ups before her. Still
assuming this was still just another evacuation, just like all the
others beforehand, her concern rose immediately when she got the
clipped official reverse emergency call that basically and starkly
told her to: “go now, as quickly as possible”.
Standing in her bedroom, still in her pajamas, she remembers
briefly considering what she should take with her, aware that each
additional step would take time and energy she had little of. She
noted her wedding rings in a small heart-shaped ceramic bowl on a
bedside table. “I’ve left them here before”, she thought, “and
they’ll be alright”. She grabbed sweat pants and a top as she
moved as fast as she could toward the garage, where she kept a
car for friends who would drive her to doctor visits. It was now the
only sure way to get out, and she had no reservations about doing
what she needed to do.
As she opened the garage door, she saw her neighbor, Dan, who
lives across the street. He was just getting into his truck with what
belongings he could carry, ready to bolt. When he saw the garage
door come up, he ran across the street to see if Laura needed help.
She told him “no, I’m going to drive myself to Chico”, but did ask
him to put her heavy oxygen concentrator in the car so she
wouldn’t have to use oxygen tanks while she waited out the fire
in Chico. He did, and then left shortly after her.
Laura got into the car she hadn’t driven - because of her feet -
and started driving for the first time in many months. “I hope
I don’t kill anyone with my driving”, she remembers thinking as

she concentrated on getting her feet to touch the right pedals.
Her car held her wheelchair, her walker, and the sweat pants
and top she grabbed as she left the house. In the back seat, and in
the trunk, were several more oxygen bottles, enough to make the
car into a pretty substantial moving bomb.
Turning right out of the estate she lived in, onto Sawmill Road,
she noted heavy smoke was quickly building up, and already
making it difficult to see. A few hundred feet later, she turned left
onto what she thought was the main road going West, Pearson
Road, but then found it was just a neighborhood road, and
followed it back to Sawmill. She got back on Sawmill and went
only a few feet more before seeing that flames had just reached the
house on the corner of Sawmill and Pearson, and it started becom-
ing clear to her for the first time that this might be more than just
an evacuation.
She turned left onto Pearson, merging into hundreds of other
cars doing what she was doing, heading towards the main road,
the Skyway, out of Paradise. She became a small part of a surging,
slow-moving, un-organized caravan of evacuees. But the quick-
moving fire was already beginning to out-pace them, and they
found they were already having to drive around, under and even
through blowing embers, and fires burning in houses and trees
along Pearson Road. She began seeing and hearing the sound of
what eventually were thousands of propane tanks exploding in
houses alongside the caravan, which continued throughout the
entire morning and afternoon, giving the evacuation from
Paradise a war-like feel.
Not too much farther ahead on Pearson, the caravan had to
stop, as the road had become engulfed in fire. From that point on,
Laura does not remember which way they turned, where they
went, nor how many routes they probed, and does not know at
all where she was. She was just “behind the car in front of me”,
with the fire, and embers, and smoke being much more visible than
the road. She does remember that at least three times the snaking
caravan ran upon a road entirely blocked by fire and had to stop
and turn completely around on a two-lane road. This was a huge
logistical maneuver for hundreds of drivers under great duress. But
she recalls the unnatural politeness and calm that every driver
showed during those moments, slowing and pausing to let others

in, giving way to others when needed, and gradually re-starting the
desperate caravan’s search for a way out.
After several hours of crawling towards a possible exit, only to
be blocked by fire, the oxygen tank that Laura was using in the
passenger-side front-seat began to run out. She realized she was
going to have to hook up a new bottle in the back seat, but it
wasn’t easily accessible from the front seat. She started struggling
to get out of her seat enough to reach it, but it was quickly
apparent to her that she wasn’t going to be able to reach it, and
wouldn’t be able to get out of the car.
But soon, the scope of the town’s devastation became apparent.
The Camp Fire burned more than 18,793 structures, among which
13,972 residences belonging to at least 25,000 people, and killed
86 people in their cars, on the road, and in their homes, the single
most devastating fire, ever, in California.
At that very moment, a woman with curly hair, an evacuee like
her, in a car moving right alongside of hers must have noticed
something, because she decided to stop in the middle of the road
and ran over to ask Laura if she needed help. The woman was
obviously practiced in transferring breathing devices from one
tank to another, and did it quickly, with little comment. Saying
Figure 11.9 Homes destroyed in the Sawmill Road area of

Paradise (Grandma’s home circled; Red = 100% destroyed).
Image credit: Gary Eilerts
nothing more, she finished the job, got back into her car and drove
off into the smoke. Laura is positive that unknown woman saved
her life right there - a small act of love and concern in the middle of
chaos, from a stranger.
It took several more hours before the evacuees that Laura was
moving with were able to find a backroad they could use to get
through the fire. She finally arrived in Chico - which is a 20-minute
drive on a normal day - more than five hours after she left her
home. Driving up in front of the store operated by her friends, the
Lashes, she was completely exhausted, and completely unable to
get out of the car. She honked once, and then waited until they
came out to investigate, found her, and helped her out of the car.
She ended up staying several days at their home, warmly cared-for,
by, as she calls them, “saints”.
Grandma Laura, like thousands of her neighbors who began the
day as evacuees, ended it as homeless refugees.
A week later, Grandma moved temporarily to her son Gregg’s
house in San Jose. In the weeks since, she has found a new place to
live, not too far from his home, and is ready to create a new life.
Just a few days ago, Gregg went up to Paradise, now accessible
after a month of emergency operations, to see if anything is
retrievable from the ashes of Grandma Laura’s house. Rains that
came soon after the fire compacted the ashes and left a sometimes
Figure 11.10 What’s left of 5430 Sawmill Road, #14.

Figure 11.11 Dan’s house was across the street, upper right.
The truck in the driveway, upper left, has an “X” to indicate
it was checked for bodies. Image credit: Gary Eilerts
toxic mire to dig through. The only things that he found intact
were ceramic in nature, born from heat and fire, and able to resist
it, like dishes and a few other pieces of pottery. Among these,
Laura was surprised to discover that there were two intact ceramic
angels, which she doesn’t remember ever owning, but which she
immediately accepted as being related somehow to the two angels
who helped her get out of Paradise.
Gregg also found the heart-shaped bowl. It had only a crumbly
residue in it, but nothing that resembled the rings that were last
sitting there. Doing a little research, Gregg found that diamonds,
unlike many other substances, don’t exactly burn – but they do
evaporate in high heat, kind of like the entire town of Paradise.
We’re all very thankful that Grandma Laura got out of Paradise
relatively safely. Many of her friends and former neighbors have
been left in very precarious conditions, with no clear way forward
from here. But it is still very sad to all of us that the simple pleasure
she wanted most - to live out the rest of her time where she
belonged, and was known and loved - will not happen.
Figure 11.12 Grandma Laura’s two angels.

Life is sometimes what happens after you think you have it

figured out. But may we all have the grace and courage that
Grandma Laura has already shown, while she, and we, continue
to figure it out.
FIRE AND AUSTRALIA’S
12 BLACK SUMMER
Filled with kangaroos, cockatoos, echidnas, and koalas,
Kangaroo Island rests across the Investigator Strait, about
100 miles (155 km), from Adelaide, Australia. Ecologists refer
to the Island as “Noah’s Ark” due to its rich biodiversity.1 But in
late 2019, half the island burned as a series of vicious fires carved
a path of destruction with unprecedented speed and ferocity. The
impact on the island’s ranchers, farmers, and wildlife was cata-
strophic.2 One-third of its 50,000 koalas (approximately 17,000
animals) perished. Thirty to forty percent of its kangaroos died.
Many other endangered or rare species experienced severe losses.
Species such as the black cockatoo, Rosenberg’s goanna, the
dunnart, and the short-beak echidna suffered terribly. Some
44,000 animals in total are believed to have perished, and non-
profit groups like the Royal Society for the Prevention of Cruelty
to Animals (RSPCA) worked around the clock to save thousands
of fire-damaged animals3 (Figure 12.1).
Tragically, Kangaroo Island’s losses were representative
of conditions across much of the entire continent of Australia.
1
www.bbc.com/news/world-australia-51102658.
2
www.nytimes.com/2020/02/04/world/australia/kangaroo-island-fire.html.
3
www.upi.com/Top_News/World-News/2020/01/17/Almost-44000-animals-on-
Australias-Kangaroo-Island-have-died-from-fire/5351579283755/.
253 / Fire and Australia’s Black Summer
Figure 12.1 A koala in the midst of a devastated landscape on

Kangaroo Island. Photo courtesy of the Royal Society for the
Prevention of Cruelty to Animals. www.rspcasa.org.au/ki-
bushfires-plan/. The RSPCA South Australia provided veterinary
support to Kangaroo Island Wildlife Park, and also ran special
feeding programs to prevent native animals from starving due to
the widespread destruction of their natural habitats and food
sources.
Extremely hot, dry conditions – combined with several years of

reduced rainfall and a severe 2019 drought (which was related
to the Indian Ocean Dipole event discussed in Chapter 8) – set
the scene for Australia’s unprecedented 2019–2020 fire season.4
While many factors contributed to Australia’s “Black
Summer,” in this chapter we will focus on how extremely warm
temperatures and associated increases in atmospheric water
demand helped set the stage for such a catastrophe. During
and before the fire-filled Black Summer, month after month of
very warm air temperatures helped increase the atmosphere’s
ability to draw moisture away from dead and live vegetation,
setting the stage for potentially explosive fire expansion. We can
start our exploration of these conditions by plotting a time series
of annual averages of monthly maximum air termpatures
(Figure 12.2). The final bar in this bar plot identifies an excep-
tionally warm year – the warmest year in the observational
record, with an anomaly of +2.4C when compared to the
1910–1939 mean. The positive anomaly in 2019 is much larger
than any previous value back to 1910, when the historical
record begins.
Looking at Figure 12.2, we should note a phenomenon
we have seen time and time again throughout this book. The
past five or so years have been much warmer than conditions
during the 1990s, indicating rapid climate change. In this case,
recent summers have been about 1C warmer than the 1990s.
Exceptionally warm conditions in 2019 were also preceded by
two very warm years in 2017 and 2018. These very warm air
temperatures combined with exceptionally low rainfall in 2018.
In many areas, the twenty-four-month 2018–2019 rainfall totals
were the driest on record. This tendency toward warmer, drier
conditions resulted in very dry fuel conditions and extreme fire-
risk weather, especially in eastern and southern Australia.5
4
Hughes, Lesley, Will Steffen, Greg Mullins, Annika Dean, Ella Weisbrot, Martin
Rice; Australia Climate Council. “Summer of Crisis,” March 11, 2020. www
.climatecouncil.org.au/resources/summer-of-crisis/.
5
www.bom.gov.au/state-of-the-climate/State-of-the-Climate-2018.pdf.
Figure 12.2 Annual maximum air temperature anomalies for

Australia. Data obtained from the Australian Bureau of
Meteorology. Between 1990–1999 and 2015–2019, maximum
air temperatures have increased by approximately 1C.
These conditions set the stage for massive conflagration.

Between September 2019 and early 2020, a stunning 5.8
million hectares of broadleaf forest burned in two of Australia’s
southeastern provinces, New South Wales and Victoria.6 Across
the continent,7 Australia’s fires burned 72,000 square miles
(186,000 square kilometers), destroying some 5,900 buildings
and killing more than one billion animals.8
An early 2020 Nature Climate Change article set these
fires in historic context. Satellite-based estimates of burned area9
and weather data10 were used to ask and answer two questions:
6
Boer, Matthias M., Víctor Resco de Dios, and Ross A. Bradstock. “Unprecedented burn
area of Australian mega forest fires.” Nature Climate Change 10.3 (2020): 171–172.
7
en.wikipedia.org/wiki/2019%E2%80%9320_Australian_bushfire_season.
8
www.sydney.edu.au/news-opinion/news/2020/01/08/australian-bushfires-more-
than-one-billion-animals-impacted.html.
9
Giglio, Louis, et al. “The Collection 6 MODIS burned area mapping algorithm
and product.” Remote Sensing of Environment 217 (2018): 72–85.
10
SILO – Australian Climate Data from 1889 to Yesterday, www.longpaddock.qld
.gov.au/silo/.
Are the 2019–2020 forest fires unprecedented in scale, and are

they the result of unparalleled fuel conditions?
To address the first question, the authors compared the
frequency of global fire extent in temperate broadleaf forests
over the 2000–2019 time period. This was the period in which
satellite-based fire extent estimates were available. The study’s
results were stunning: “[T]he 2019/20 forest fires have burned a
globally unprecedented percentage of any continental forest
biome: 21% of the Australian forests.” The typical burned area
for Australia in previous years was only about 5 percent.
To answer the second question, the authors analyzed
the frequency and extent of forested areas in New South Wales
and Victoria that were experiencing critically dry “dead fuel
moisture” conditions. Dead fuel moisture refers to the fraction
of the fraction of the mass of dead forest branches and litter that
is made up of water. Forests with low fuel moisture fractions are
much more flammable, and under the right weather conditions
this allows fires to spread much further and faster. Dead dry fuel
conditions can be estimated using daily vapor pressure deficits
(VPD).11 Vapor pressure deficits are based on the difference of
two terms, the saturation vapor pressure and vapor pressure.
“Pressure,” as it is used here, refers to the pressure exerted by an
amount of water vapor. In other words, it represents a quantity
of water. Saturation vapor pressure defines the theoretical upper
limit on the amount of water vapor the atmosphere can hold. It
represents conditions when the relative humidity is at 100 per-
cent. As discussed in Chapters 6 and 11, a warming atmosphere
can hold more moisture (Figure 6.3). This increased water-
holding capacity will lead to increasing vapor pressure deficits,
which are tied to increases in forest fire extent (Figure 11.2).
Saturation vapor pressure increases with increasing air
temperatures, since there is more room in warmer air to hold
water (Figure 6.3). Vapor pressure is simply the actual amount
11
Nolan, Rachael H., et al. “Large-scale, dynamic transformations in fuel moisture
drive wildfire activity across southeastern Australia.” Geophysical Research
Letters 43.9 (2016): 4229–4238.
of water vapor in the air. Vapor pressures increase after it has

rained, or when moist air blows in from an ocean. The combin-
ation of high saturation vapor pressures and low actual vapor
pressures produces high vapor pressure deficits. High vapor
pressure deficits help desiccate living and dead plants, producing
conditions conducive to extensive mega-fires.
The occurrence of large wildfires can be conceptual-
ized as the simultaneous triggering of three switches: (1) the
availability of dry biomass to burn, (2) weather conditions
conducive to spreading fire, and (3) a source of ignition(s).12
In southeastern Australia, scientists have documented a two-
stage sequence relating vapor pressure deficits to fire. First,
atmospheric conditions: vapor pressure deficits are linked to
dead vegetation fuel moisture. Dead fuel moisture content
refers to the fraction of the mass of the dead vegetation made
up by water. Second, the amount of dead fuel moisture is
related to the observed extent of wild fires. Figure 12.3 shows
the empirical results from a study in southeastern Australia.
The gray dots in the background correspond to observed fire
extents. The vertical gray bars denote break points at about
10 and 15 percent. When dead fuel moisture levels reach
these critical thresholds, the magnitudes of forest fires jump
upward.
It is very important that we all understand these results.
When dead fuel moisture levels drop below 15 or 10 percent, the
risk of a large forest fire increases dramatically. This relation-
ship provides extremely useful information for firefighters and
first responders. It also helps us understand how high tempera-
tures can set the stage for potentially catastrophic fires. In
Australia’s eucalyptus forests, fires primarily spread through
low-lying forest litter13; moist litter acts to limit the spread of
fires. Typically, naturally occurring gullies, south-facing slopes,
12
Bradstock, R. A. “A biogeographic model of fire regimes in Australia: Current and
future implications.” Global Ecology and Biogeography 19.2 (2010): 145–158.
13
Murphy, Brett P., et al. “Fire regimes of Australia: A pyrogeographic model
system.” Journal of Biogeography 40.6 (2013): 1048–1058.
Figure 12.3 Observed dead fuel moisture content from Nolan

et al. (2016). Used with permission. The annotation “Note that
fires with fuel moisture <10% are very large” was added to
facilitate interpretation.
and swamps provide fire-breaks, but under very dry conditions

fire extent can grow in a highly nonlinear fashion.
With this background, we can now look at how Boer,
Resco de Dios, and Bradstock answered their second question:
Are the 2019–2020 forest fires the result of unparalleled fuel
conditions? Their solution was based on calculations of daily
dry fuel moisture values using weather data. In an approach
quite similar to our analysis of heat waves (Chapter 5), the
authors identified very dry fuel days as whenever a location
had a daily dead fuel moisture value of less than 10 percent.
This corresponds to the 10 percent break point in Figure 12.3
that is strongly associated with large fires. Calculating the
number of days multiplied by the area in forested regions in
New South Wales and Victoria produced a metric that helps
summarize annual fuel moisture conditions (Figure 12.4). This
figure demonstrates that the 2019 fuel moisture levels were truly
exceptional, far higher than any on record. We can also see a
Figure 12.4 Forest area in critically dry fuel state, eastern

Australia (1990–2019). Annual variation in the duration and
cumulative area of large forest patches (>100,000 ha) in a
critically dry fuel state. The horizontal black line indicates the
thirty-year mean value; light and dark gray bands indicate mean
value 1 and 2 standard deviations, respectively. Figure and
caption taken with permission from Boer, Resco de Dios, and
Bradstock (2020)
strong upward trend in the data, a propensity for more forest

area to be in a critically dry moisture state. So, yes, unparalleled
fuel moisture conditions helped produce the exceptional
2019–2020 forest fires.
According to an early 2020 Australian Climate
Council Summer of Crisis report, these fires impacted 80 per-
cent of Australians14 and released between 650 million and 1.2
billion tons of CO2 into the atmosphere, far more than the
annual emissions of Australia or Germany.15 In New South
Wales, the hardest-hit province, more than 11,000 fires burned
5.4 million hectares (13.3 million acres). The Summer of Crisis
14
Biddle, N., B. Edwards, D. Herz, and T. Makkai. “Exposure and the impact on
attitudes of the 2019-20 Australian Bushfires.” ANU Centre for Social Research
Methods (2020). Accessed at: https://csrm.cass.anu.edu.au/research/publications/
exposure-and-impact-attitudes-2019-20-australian-bushfires-0.
15
Bloomberg. “Australia’s fires likely emitted as much carbon as all planes” (2020).
https://www.bloomberg.com/news/articles/2020-01-21/australia-wildfires-cause-
greenhouse-gas-emissions-to-double.
Figure 12.5 The Gospers Mountain Fire rages (December 21,

2019), devastating areas such as Bilpin. From Summer of Crisis
by Lesley Hughes, Will Steffen, Greg Mullins, Annika Dean,
Ella Weisbrot, and Martin Rice.
report detailed the incredible ecological destruction associated

with the Black Summer. More than 510,000 hectares
(1,260,000 acres) burned in the Gospers Mountain fire near
Sydney (Figure 12.5), making it the largest forest fire recorded
in Australia. Fires also burned through 80 percent of the Blue
Mountains World Heritage Area, and through more than half
of the Gondwana Rainforests in New South Wales and
Queensland. These ancient forests include unique species that
date back to the time of the Gondwana supercontinent, some
180 million years ago. For millions of years, too damp to burn,
the Gondwanan rainforest evolved in a fire-free state. In 2020,
according to initial assessment, more than 300 species, many
of them endangered, have had at least 10 percent of their
habitat burned by the fires.
We can use climate change simulations to place the
warm New South Wales and Victoria air conditions in an
historic context. Figure 12.6 shows average annual southeastern
Australia air temperature anomalies for a large collection of
Figure 12.6 Average New South Wales/Victoria air temperature

anomalies from a collection of climate change simulations.
eighty-one simulations.16 Between the late nineteenth century

and early twenty-first century, the models estimate an average
increase of about 1.4C. If we project out into the future using a
rapid emission scenario,17 the models anticipate that this area of
southeastern Australia will continue to warm rapidly. By 2060,
we appear likely to see another approximately 1.6C increase in
average temperatures. Natural year-to-year variations in tem-
perature will result in years even warmer than this average.
We can use this large collection of climate change simu-
lations to perform a formal climate attribution analysis. This
allows us to assess how climate change contributed to the excep-
tionally warm conditions in southeast Australia. We use a large
set of simulated 1900–1929 air temperatures to describe natural
year-to-year variations in a world without climate change. The
16
Based on an eighty-one-member ensemble of climate change simulations from
phase 5 of the Climate Change Model Intercomparison Project (CMIP5).
Simulations obtained from the KNMI climate explorer.
17
The 8.5 W/m2 Representative Concentration Pathway scenario. Discussed further
in Chapter 13.
Figure 12.7 Formal attribution analysis of New South Wales/

Victoria air temperature anomalies from a collection of climate
change simulations.
distribution of these temperatures, expressed as differences from

the 1900–1929 average, is shown with the dashed line in
Figure 12.7. Typical values range from about 1C to +1C.
The observed 2019 temperature anomaly of +1.8C is way
warmer, allowing us to formally state that temperatures as
warm as those observed in southeastern Australia in 2019 would
have been essentially impossible in a world without climate
change.
We can also use this large collection of climate change
simulations to examine how likely a year like 2019 might be in a
world with climate change. The solid line in Figure 12.7 shows
the distribution for 2019 based on the climate change simula-
tions forced with observed changes in greenhouse gasses and
aerosols. While the +1.8C lies on the upper side of this distri-
bution, it is quite likely to have conditions this warm, according
to the models, in a world with climate change. So, according
these models, a year as warm as 2019 should happen quite
frequently, about once every nine years, in a world with
observed concentrations of greenhouse gasses and aerosols.
The strong link between vapor pressure deficits and fire

extent (Figure 12.3) allows us to connect these increases in air
temperatures directly to changes in dead fuel moisture condi-
tions (Figure 12.4). We can do this by performing an attribution
analysis similar to that performed in a 2019 climate attribution
study focused on the 2018 Four Corners drought in the United
States.18 Because there is a direct relationship between increases
in air temperatures, saturation vapor pressure, and dead fuel
moisture levels, we can translate changes in temperature to
changes in dead fuel moisture.
Using the same observed air temperature and vapor
pressure data as analyzed by Boer, Resco de Dios, and
Bradstock, we can use a little simple mathematics to calculate
time series of annual vapor pressure deficits and saturation
vapor pressure values. In Figure 12.8, these values are averaged
across New South Wales and Victoria, and expressed as anom-
alies from the 1990–2019 mean. The black bars in Figure 12.8
show the observed annual vapor pressure deficits (saturation
vapor pressure minus actual vapor pressure). There is a very
strong upward trend (shown with a diagonal line). Note that the
2020 vapor pressure deficit was by far the largest on record –
about twice as large as any value prior to 2018.
The straight diagonal line in Figure 12.8 displays
a regression-based vapor pressure deficit trend estimate.
Interannual vapor deficit variations correlate fairly well with
this trend, with a correlation value of 0.6. Using this trend line
to project out to 2040 presents a scary potential future. In just
twenty years, a persistence of the observed trend would mean
that average years in New South Wales might have vapor pres-
sure deficits similar to the worst year on record (2019).
The vertical gray bars in Figure 12.8 show year-to-year
variations in saturation vapor pressure. These variations are
18
Williams, Emily, et al. “Quantifying human-induced temperature impacts on the
2018 United States Four Corners hydrologic and agro-pastoral drought.”
Bulletin of the American Meteorological Society 101.1 (2020): S11–S16. journals
.ametsoc.org/doi/pdf/10.1175/BAMS-D-19-0187.1.
Figure 12.8 Time series of observed annual average vapor

pressure deficits (black bars) and saturation vapor pressure
values (gray bars) for New South Wales and Victoria. Also
shown is trend estimate for 2040.
driven purely by changes in air temperature. The magnitude of

the trend in this data is essentially identical to the trend in vapor
pressure deficits, but the correlation is higher (0.7). Warming air
temperatures are largely responsible for the upward trend
shown in Figure 12.8. It is also very noteworthy that in
2018 and 2019, saturation vapor pressures were exceptionally
high as well. These data suggest that in both 2018 and 2019, the
impact of warm temperatures was even greater than the impact
of below normal atmospheric water vapor. Two back-to-back
very warm years help set the stage for fire.
Finally, we can connect the dots, linking our attribution
analysis (Figures 12.6 and 12.7) with estimates of vapor pres-
sure deficits and areas with exceptionally dry fuel moisture
conditions (Figure 12.9). The x-axis of Figure 12.9 shows
annual 1990–2019 vapor pressure deficit values, averaged over
New South Wales and Victoria. The y-axis shows estimates of
the area-days with dead fuel moisture values of less than 10
percent. As in Figure 12.8, the year 2019 – and, to a lesser
Figure 12.9 Scatterplot of New South Wales/Victoria annual

vapor pressure deficit values and area in critically dry fuel state.
degree, 2018 – stand out as exceptional. With an estimated

17 million km2-day area-days in critically dry fuel state, the
observed exposure levels in 2019 were about 70 percent above
the previous maximum values of about 10 million.
Our climate attribution analysis (Figure 12.6) suggested
that about 1.5C of the 2019 warming can be associated with
long-term changes in the mean state due to climate change. This
allows us to produce a “counterfactual” experiment that
describes a world without this warming. To simulate a world
with vapor pressure deficits similar to 2019, but with cooler
conditions similar to a world without climate change, we can
cool our observed weather data by 1.5C and recalculate the
daily pressure deficits for each location. Then, for each location,
we can recalculate the dead fuel moisture levels. Then we can
calculate the associated area with critically dry fuel conditions.
These results are shown with the left-hand diamonds in
Figure 12.9. Even with 1.5C of warming removed, the year
2019 would still have had exceptionally dry fuel conditions.
Naturally occurring variations in rainfall and air temperatures
were likely the primary driver of these arid conditions.

Nevertheless, a substantial fraction of the critically dry area-
days (~30 million km2-days) appears related to the human-
induced warming of 1.5C.
We can also use the projected near-term warming
(Figure 12.6) to estimate that southeastern Australia may soon
be, on average, about 1.5C warmer than at present. To explore
what a year like 2019 might look like, we can add 1.5C to the
observed weather data and recalculate the 2019 statistics. These
results are shown with the right-hand diamonds in Figure 12.9)
An exceptionally warm/dry year like 2019 may soon appear
much worse, from a fire weather perspective. Our future climate
projection indicates a massive increase of about 5 million km2-
days in areas with critically dry fuel conditions. Future dry fuel
events worse than 2019 may be on the way soon.
Conclusion
Mega-fires are very complex, and arise through an
interplay of ignition, weather conducive to fire, and very dry
fuel conditions. The simplest aspect of this complex causal
network is the fairly straightforward relationship between
extremely dry fuel conditions and wildfire extent. When fuel
conditions are extremely dry, natural barriers to fire spread are
greatly diminished. Given that there is a straightforward rela-
tionship between climate change and increasing annual tem-
peratures (Figures 12.2, 12.6, and 12.7) and increases in
saturation vapor pressure-driven decreases in dead fuel
moisture, we have analyzed annual vapor pressure values to
gain insights into one of the key drivers of the 2019–2020 fire
season in New South Wales and Victoria.
One simple but important takeaway from this analysis
is that air temperature-related impacts are extremely important
for fire hazards. Rainfall totals have been very low in Australia,
and we might have found that vapor pressure deficits were the
main driver of the total vapor pressure deficit anomalies. If that
was the case, then we would have found large black bars and
small gray bars in Figure 12.8. Examples of such vapor-pressure

deficit-driven years can be seen in 2002 and 2006. By contrast,
in 2018 and 2019 we see that the gray bars themselves are large.
In fact, in 2019, the results presented here suggest that the very
warm temperatures, acting alone, would have produced vapor
pressure deficits greater than any previous year since 1990.
These very warm conditions would have been impossible in a
world without human-induced climate change, and are almost
certain to become more frequent in the near future.
So, while exceptionally warm temperatures were just
one enabling component of the 2019–2020 fire season, they
certainly did help produce the exceptionally dry 2019 fuel con-
ditions in New South Wales and Victoria. Climate change con-
tributed in a major way to Australia’s exceptionally dangerous
and destructive Black Summer by increasing vapor pressure
deficits and fuel moisture conditions, allowing fires to spread.
Going forward, both observations (Figure 12.8) and climate
change simulations (Figure 12.6) indicate that these exception-
ally dry conditions are likely to become more frequent, and
soon. As discussed in the next chapter, the carbon dioxide
emitted from these types of massive fires may act as a positive
feedback, contributing to more global warming and a destabil-
ization of our “Dixie® Cup” planet.
DRIVING TOWARD +4C ON
13 A DIXIE® CUP PLANET
Representative Concentration Pathways

To explore our collective future, scientists develop
plausible emission, pollution, and land use scenarios. Some of
these scenarios describe rapid economic and population
growth, and concomitant increases in fossil fuel emissions.
Other scenarios describe a rapid transition to clean energy. In
keeping with our standard desire to baffle and obfuscate, us
scientists call these “scenarios” representative concentration
pathways (RCPs). “Scenarios” just seems so transparent and
blasé. Four main RCPs1 are used to project out to 2100,
providing plausible scenarios (what-if stories) that we can use
to explore the potential greenhouse impacts of different human
development pathways. Do we face a cheery green future, a
dark rapid emissions outlook, or something in between? The
RCPs help us understand how future changes in population,
gross domestic product (GDP), energy use, land use, mitigation,
1
Vuuren et al. “The representative concentration pathways: an overview.” Climatic
Change 109 (2011): 5–31, DOI 10.1007/s10584–011-0148-z. The latest work
featured by the Intergovernmental Panel on Climate Change uses “shared
socioeconomic pathways,” which provide a more complete description of
potential future social configurations. Here we focus on the older, simpler
“representative concentration pathway” framework because it is easier to explain
and still widely relevant.
269 / Driving toward +4C on a Dixie® Cup Planet
and pollution might change the atmospheric forcing associated

with greenhouse warming.
“Concentration” refers to estimates of the amount of
different greenhouse gasses and aerosols in the atmosphere at
different times. These estimates are a function of emissions, land
use, and mitigation activities. The RCP2.6 scenario2 is a happy
story characterized by a near-term peak in emissions, followed
by a rapid decline in emissions. At the other end of the spectrum,
RCP8.53 is characterized by rapid population growth, limited
technological innovation, and rapidly increasing greenhouse
gasses. The slightly less grim RCP64 scenario is characterized
by increasing greenhouse gas emissions over time, but with
stabilization shortly after 2100 due to the application of a range
of technologies and strategies for reducing emissions. The
RCP4.55 scenario has relatively slow increases in greenhouse
gas emissions and the stabilization of greenhouse gas emissions
shortly after 2100. These scenarios also represent different out-
comes in terms of “energy intensity” and “carbon fraction.”
Energy intensity, the energy per unit of income, refers to how
fuel efficient an economy is. Technology, investment, energy
efficiency, and transitions toward higher-value products and
service industries make some economies (like Germany’s) less
energy intense. Carbon fraction refers to the emissions per unit
of energy used in an economy; moving toward renewable energy
reduces an economy’s carbon fraction. RCP6 and RCP8.5 have
2
Riahi, K., A. Grübler, and N. Nakicenovic. “Scenarios of long-term socio-
economic and environmental development under climate stabilization.”
Technological Forecasting and Societal Change 74 (2007): 887–935.
3
Van Vuuren, D. P., M. G. J. Den Elzen, P. L. Lucas, B. Eickhout, B. J. Strengers, B.
et al. “Stabilizing greenhouse gas concentrations at low levels: An assessment of
reduction strategies and costs.” Climate Change 81 (2007): 119–159.
4
Clarke, L. E., J. A. Edmonds, H. D. Jacoby, H. Pitcher, J. M. Reilly, and
R. Richels. “Scenarios of greenhouse gas emissions and atmospheric
concentrations. Sub-report 2.1a of Synthesis and Assessment Product 2.1.”
Climate Change Science Program and the Subcommittee on Global Change
Research, Washington, DC, 2007.
5
Fujino, J., R. Nair, M. Kainuma, T. Masui, and Y. Matsuoka “Multigas
mitigation analysis on stabilization scenarios using aim global model.” The
Energy Journal Special Issue 3 (2006): 343–354.
high carbon factors throughout the twenty-first century.

RCP8.5 maintains a high level of energy intensity.
RCP2.6, RCP4.5, RCP6, and RCP8.5 indicate the
anticipated 2100 radiative forcing value in Watts per meter
squared [Wm 2]. The “radiative forcing value” is the change
in the amount of long-wave radiation absorbed by Earth’s sur-
face due to greenhouse gas and aerosol emissions. We are now
at a radiative forcing of about ~2 Wm 2. The future radiative
forcing values range from small changes (2.6 Wm 2, or ~130%
of the current radiative forcing value of 2 Wm 2) to very high
(8.5 Wm 2, or about ~425% of the current value). Hence the
numbers after the RCPs.
In terms of human and environmental impacts, the
differences between these scenarios are profound. In RCP2.6,
radiative forcing peaks very soon, shortly after 2025, and
declines to levels near those of today. In RCP8.5, conversely,
rapid CO2 and methane emissions quickly lead to runaway
increases in radiative forcing. By 2050, the RCP8.5 radiative
forcing has already more than doubled over the 2000 value, to
about 5 Wm 2. In RCP4.5 and RCP6, on the other hand,
mitigation and reduced emissions eventually lead to a stabiliza-
tion in greenhouse gas levels. This occurs around 2070 for
RCP4.5 and after 2100 for RCP6. By 2050, these latter two
scenarios also predict substantial warming, with increases in
radiative forcing of about 3.7 Wm 2.
The differences between these forcing numbers make a
big difference in where we are heading (Figure 13.1), even on a
short time scale like the change between now and 2050. The
RCP4.5 scenario suggests an additional about 1-degree
warming between 2000 and 2050, and about 2-degree warming
by the end of twenty-first century. The RCP8.5 scenario sug-
gests an additional about 1.5-degree warming between
2000 and 2050, and a greater than 3-degree warming by the
end of twenty-first century. This latter scenario would be very,
very bad.
In RCP8.5, population approaches 10.5 billion by
2050 and peaks near 12 billion in 2100. In RCP4.5, population
Figure 13.1 Simulated (CMIP5) and observed global

surface temperatures.
totals of about 9 billion in 2050, then levels out. RCP4.5 projects

that our total 2100 energy consumption will be about twice our
current level. In the RCP8.5 scenario, however, rapid population
growth and a lower rate of technology growth results in a rapid
increase in energy consumption, resulting in 2100 consumption
rates of more than three times those at the beginning of the
twenty-first century. The mix within the total energy consumed
varies substantially, as well, with RCP8.5 accounting for most of
its increase in energy through oil consumption. RCP2.6 repre-
sents a scenario of aggressive energy substitution, and oil con-
sumption begins dropping before 2025. While RCP4.5, RCP6,
and RCP2.6 all predict similar levels of energy consumption, the
transition toward alternative energy sources limits rapid

increases in radiative forcing in RCP4.5 and RCP2.6.
Changes early in twenty-first century emissions will
have important compounding impacts on the total amount of
greenhouse gas in the atmosphere. It is important to note that
greenhouse gasses have long residence times in the atmosphere.
Carbon dioxide is thought to have a residence time of between
30 and 95 years6 while nitrous oxide has a mean residence time
of 114 years.7 Methane (CH4), on the other hand, has a
relatively short residence time (14 years). Because greenhouse
gasses persist in the atmosphere, they build up over time. This
can, and will, if business-as-usual continues, have dire conse-
quences for our planet. Linear 2000–2060 increases in RCP8.5
CO2 emissions, for example, translate into exponential
increases in 2000–2100 CO2 concentrations. The converse
also holds. Were we to adopt a rapid near-term transition
away from fossil fuels, the RCP2.6 scenario suggests that we
would quickly see the increase in atmospheric CO2 stop, sta-
bilizing the climate system.
And what if we wait? It depends for how long. Unless
we quickly transition to an RCP2.6-like path very soon, the
early twenty-first-century accumulation of greenhouse gasses
will likely propel us beyond the dangerous 2C warming thresh-
old by 2050. Besides having devastating impacts on agriculture,
such warming would also heavily disturb eco-climatic zones,
endangering many species of animals, and heavily impact coral
reefs and associated fisheries.8 Remember that this would be
warming at a rate that is unprecedented – as far as we know – in
Earth’s history: animals and plants simply cannot migrate or
adapt – and certainly not evolve – fast enough to keep up with
6
Jacobson, M. Z. “Correction to ‘control of fossil-fuel particulate black carbon and
organic matter, possibly the most effective method of slowing global warming.’”
Journal of Geophysical Researth 110 (2005): D14105. doi:10.1029/
2005JD005888.
7
IPCC Fourth Assessment Report, Table 2.14, Chap. 2, p. 212.
8
“Turn down the heat,” Worlbank report. https://www.worldbank.org/en/topic/
climatechange/publication/turn-down-the-heat
the rate of change. We are already seeing this expressed as

catastrophic coral bleaching,9 massive marine fisheries
impacts,10 and fire-induced animal genocide.11 But this affects
humans too. As we saw in Chapter 5, increased temperatures
will expose billions of people to temperatures beyond our
body’s adaptive capacity. In Africa,12 a 2C warming by
2040–2050 is likely to produce substantial declines in crop
yields, a 30–40 percent increase in the rate of crop failure (from
about once in five years to once in four years), more frequent
extreme warm events, 30–50 centimeters of sea level rise around
the coasts, and put an additional 10–15 percent (above now) of
sub-Saharan African species at risk of extinction.
In Southeast Asia (Pakistan, India, Nepal, Bhutan,
Bangladesh, Sri Lanka) by 2050, some fisheries may experience
a 15–50 percent decline in productivity, and 98–100 percent of
coral reefs are likely to become “thermally marginal” – which
means that the water has gotten so warm, it will be difficult for
them to survive. During summer, 60–70 percent of the region
will experience unusual heat extremes, and 30–40 percent of the
region will likely experience unprecedented heat extremes. In
Southern Asia, the World Bank finds that a 2C warming will
likely be associated with a 20 percent increase in unusual sum-
mertime heat extremes, while a sea level rise might expose
millions more people to flood risk, especially in Bangladesh.
Melting glaciers may also reduce water availability in the
Ganges, Brahmaputra, and Indus watersheds. Without climate
change, the number of moderately stunted children is antici-
pated to decline to 11 percent of the total number of children
by 2050; with climate change, the +2C scenario, the percentage
9
Hughes, Terry P., et al. “Coral reefs in the Anthropocene.” Nature 546.7656
(2017): 82–90.www.nature.com/articles/nature22901.
10
Webb, Robert S., and Francisco E. Werner. “Explaining extreme ocean
conditions impacting living marine resources.” Bulletin of the American
Meteorological Society 99.1 (2018): S7–S10.
https://journals.ametsoc.org/doi/pdf/10.1175/BAMS-D-17-0265.1.
11
www.sydney.edu.au/news-opinion/news/2020/01/08/australian-bushfires-more-
than-one-billion-animals-impacted.html.
12
“Turn down the heat,” Table 3.4.
of moderately stunted children only declines to 14 percent. With

a 2050 estimated population of 2.2 billion for Southeast Asia,
this 3 percent difference might represent some 70 million chil-
dren more moderately stunted children. Pause and think about
that for a second.
Looking further out, however, we can see a huge differ-
ence between the responsible and plausible RCP4.5 scenario and
the extremely dangerous RCP8.5 scenario (Figure 13.1). So far,
we have “only” experienced about 1C of warming – and in this
book we have already explored some of the mayhem related to
that warming. A consistent theme of this book has been that
warming will not happen in an even-handed way. Some regions,
like California, have warmed much faster than the global aver-
age (Chapter 11). Half a degree or more of extra warmth in the
tropical oceans can amplify the intensity of Indian Ocean
Dipole, El Niño and La Niña–related droughts (Chapters 8, 9
and 10). And a warming atmosphere will produce a world with
more extreme heat waves and rainfall events (Chapters 5–7) and
very dry fuel moisture conditions (Chapters 11 and 12).
Looking forward, we face a collective choice, probably
ranging from a total warming of about 2C to possibly more
than 4.5C. And as we have already seen, climate change has
already made weather and climate extremes more dangerous.
Most concerning, perhaps, might be the potential for
runaway warming. The Earth, like many complex systems, is a
delicate yet robust network of energy balances and feedbacks.
Yes – delicate and robust. The Earth system has negative feed-
backs that can help maintain thermal equilibrium. For example,
as the concentration of carbon dioxide increases, the oceans and
rainforests, up to a point, can absorb more CO2. Conversely,
there are positive feedbacks for climate change, and “positive”
in this context is not a good thing. For example, if global
warming melts our ice caps, they will reflect less energy, the
Earth will absorb more energy, and the rate of global warming
will increase. I like to think of these feedbacks as a Dixie® Cup
system (Figure 13.2), named after a small inverted cardboard
“Dixie®” cup. Dixie® cups are small water glasses that you find
Figure 13.2 Schematic of a Dixie® Cup Planet.
in dentist offices. My conceptual Dixie® Cup system has the

Earth rolling around on the top of this inverted cup. Typically, a
Dixie® Cup system maintains homeostasis. Small perturbations
in temperature are met with small contrary responses, maintain-
ing a temperature equilibrium. The Earth system maintains its
position within its basin of attraction – the top of the Dixie®
Cup. But huge perturbations might tip the system over the edge.
Yours truly (the author) is a Dixie® Cup system. If
Planet Funk is hot, Planet Funk sweats. If Planet Funk is cold,
Planet Funk shivers.
Yet Planet Funk might be pushed beyond operating
norms. A runaway fever might kill invading bacteria but bake
the Funk brain along the way. Funk too cold might lay down in
the snow and sleep forever.
So too Planet Earth sits within a similar “basin of
attraction” in system space; small perturbations are met with
negative feedbacks, leading to system stability.
On Planet Earth, many negative feedbacks act to nat-
urally sequester atmospheric carbon dioxide. If the Earth
warms very slowly, plants’ metabolic processes increase,
pulling more CO2 out of the atmosphere. Plants take in atmos-
pheric CO2 through small holes in their leaves called stomata.
Through the miracle of photosynthesis, this CO2 is combined
with water and solar energy to make sugar. Plants, however,
face a fundamental trade-off. They have to open their stomata
to get CO2 but in so doing they lose precious H20. When there
is more CO2 in the atmosphere, plant photosynthesis is more
efficient, and this acts as a negative feedback, assuming there is
sufficient water available. There can be lots of other negative
feedbacks, with the oceans absorbing more CO2 as the Earth
warms; molds, fungi, and insects breaking down waste prod-
ucts faster; and vegetation in northern regions growing faster
and pulling more CO2 out of the atmosphere. These and other
negative feedbacks have helped keep temperatures on our
“Goldilocks Planet” in a very narrow range for most of our
planet’s past.13
Within a narrow range of temperatures, these negative
feedbacks act to maintain homeostasis, and Planet Ours is fine.
Pushed beyond the rim of the cup, however, these negative
feedbacks may fail, and destructive positive feedbacks may
ensue. We know that things can go pear-shaped rather quickly.
At least once in our planet’s past, we experienced freezing
conditions that geologists refer to as “Snowball Earth.”
During the distant past, the Earth grew so cold that even the
tropical oceans froze, and this acted as a positive feedback, as
frozen waters reflected back more of the sun’s energy. The
opposite process may be already under way, as the polar ice
caps shrink precipitously. Disappearing ice and snow no longer
reflect downwelling solar radiation back to space. The modeling
of sea ice loss is tremendously difficult, and sea ice responses
represent one of the great uncertainties in climate change
research. For the Arctic, we do have compelling satellite data
showing very large declines in sea ice extent (and thickness)
since 1979 (Figure 13.3). Between 1979 and 2019, minimum
sea ice extents have decreased by almost 39 percent, or 2.73
million km2.
Warming may also impact the vast tracts of land
covered with permafrost. These thawing lands of ice and snow
may emit more carbon and methane, amplifying global warming
13
Zalasiewicz, Jan, and Mark Williams. The Goldilocks Planet: The 4 Billion Year
Story of Earth’s Climate. New York: Oxford University Press, 2012.
Figure 13.3 September sea ice extents from 1979 to 2020.

Data provided by the National Snow and Ice Data Center,
accessed February 28, 2020.
in unexpected and nonlinear ways. Increases in global fire-

related emissions or reductions in the respiration rates of trop-
ical forests or the sequestration of CO2 by the oceans could also
help push planet Earth into a rapid warming trajectory.
Given that even a 2C warming by mid-century will
have serious implications for food security, crop production,
fisheries, and corals, we need to pay close attention to the path
we are following. Based on recent data, which seems to be the
most representative concentration pathway?
Examining the Global Carbon Budget

Since I’m fantastically and fatalistically addicted to
data, I went to the source to answer this question – downloading
the newly released 2019 Global Carbon Budget data from the
Figure 13.4 A time series of 1990–2020 observed emissions is

shown with filled circles. Straight lines depict emission scenarios.
Note the observed drop in 2020 emissions was due to impacts
from the COVID19 pandemic, as discussed in https://www
.globalcarbonproject.org/carbonbudget/20/files/GCP_
CarbonBudget_2020.pptx. Reduced economic activity and travel
reduced emissions. How these impacts unfold in a post-pandemic
world remains unclear.
Global Carbon Project (www.globalcarbonproject.org).14 This

group closely monitors atmospheric greenhouse gas emissions
and concentrations and posts their analyses on their website.
Here we examine their 2019 carbon budget results. Figure 13.4
shows 1990–2019 observed global fossil fuel and cement
14
http://www.globalcarbonproject.org/carbonbudget/.
emissions15 in gigatons of CO2. At the Earth’s standard pres-

sure, this much CO2 would create a one-meter layer some 2,125
by 2,125 kilometers; an area of 5 million square kilometers
covering half the United States.
Figure 13.4 supports both alarm and hope. Between 1990
and 1999, emissions rose at about 1 percent per year. Between
2000 and 2009, emissions increased almost three times faster, by
about 3 percent per year, a rate that is slightly faster than the
IPCC’s dire RCP8.5 scenario. To visualize this rate of change, we
can imagine total atmospheric CO2 as one-meter-thick goo (like
John Carpenter’s The Fog), spread across the eastern half of the
United States. By 2035, a continued 3 percent per year increase
covers the rest of the country (and Hawaii and Alaska).
But Figure 13.4 also shows signs of hope. Note how
since 2015 the increase in global CO2 emissions has fallen below
the dangerous RCP8.5 increases. Reductions in the use of coal
and other mitigation efforts appear to be moving us toward the
more modest RCP6.0 or RCP4.5 scenarios, scenarios that may
ultimately produce 2.5C to 3.3C of warming.16
We should all note that the near-term differences
between RCP8.5 and RCP6.0 or RCP4.5 are huge. Following
RCP8.5, we arrive at a ~3C warming by 2060 (Figure 13.1),
enough warming to leave our children in a world of hurt,
literally, and well on our way to a dangerous more than 4C
warming by century’s end.
As we have already seen in previous chapters, this
warming won’t just manifest as a smooth, low-frequency
increase in ocean and atmospheric temperatures. We will see
this warming manifest as more intense weather and climate
extremes – stronger heat waves, droughts, and floods, as well
as more intense climate anomalies like El Niños and La Niñas.
15
It takes about 0.5–0.8 pounds of CO2 to produce 1 pound of Portland cement
(www.co2list.org). Besides water, concrete is the most commonly used material
on earth, and accounts for about 1% of U.S. carbon emissions; http://www
.concretethinker.com/technicalbrief/Concrete-Cement-CO2.aspx.
16
Based on estimates provided by the 2019 Global Carbon Budget, https://www
.globalcarbonproject.org/carbonbudget/.
How old will your children, nieces, or nephews be in

2060? My daughter Amelie and son Theo will be sixty-eight.
Will they remember us as the generation that failed? If we reach
3 degrees of warming, we can expect a much greater frequency
of unusual and unprecedented heat extremes, declines in crop
yields, increases in malnutrition and childhood stunting, ocean
acidification, and the destruction of many coral reefs. We may
also face a real risk of runaway warming, and a dramatic drop
of livability on our Dixie® Cup Planet (Figure 13.2) as melting
polar ice caps reflect less heat, the oceans acidify, and the carbon
uptake of vegetation diminishes.
Who is to blame? What caused the rapid increase in
emissions? Are these the same questions? If we look at total
cumulative emissions since 1959, North America and the
twenty-eight European Union nations have contributed about
55 percent of global emissions, with Asia contributing about 31
percent. From this perspective, the United States and Europe
have produced about half of the warming.
Looking just at the recent acceleration, however, we see
that the relative contributions to annual emissions have changed
dramatically. Between 2000 and 2019, global carbon emissions
increased by 50 percent. We are literally stepping on the gas
pedal (or coal pedal, actually), just as we race head-first for the
cliff. Most recent emissions come from Asia, and most of those
increases involve increasing coal use. Between 2000 and 2018,
global coal emissions increased by 62 percent. Since 2010,
however, coal emissions have stabilized, and were overtaken
by growth in natural gas emissions.
As the developing world develops further, boosting
gross domestic products and raising household incomes, the
bootstraps they pull themselves up with are coated with the
same cloying soot as nineteenth-century England. So, topping
our list of 2018 emission contributors is coal, coming in at
40 percent of total emissions. It is only since 2005 that coal
has exceeded oil as the biggest emission source. In 2017, the top
four emitters covered 60 percent of total emissions: China
(27%), United States (15%), the twenty-eight European Union
Figure 13.5 Per capita CO2 emissions from 1959 to 2017.

Figure from the Global Carbon Project‘s “Carbon Budget 2018.”
countries (10%), and India (7%). Emissions growth is by far the

greatest in Asia, but on a per capita basis, the United States
is far and away the largest contributor to global warming
(Figure 13.5). The global average per capita CO2 emissions rate
is about 5 tons per person. The United States has historically
been at about four times that (20 tons per person). In 2018, the
US emission rate was about 16 tons per person, still far more
than Europe or China, which come in at around 7.5 tons per
person. What if the United States we to cut per capita emissions
to European levels? If 335 million people emitted 9 tons of
carbon less per year, that would reduce global emissions by
about 3 gigatons. That would be a big, big help.
So how bad are we doing? Our current path appears
similar to the more rapid emission growth scenarios. Our brief
analysis of the change in emissions, placing us on this path,
underscored the important role played by emissions increases by
China and other emerging economies. As industrialization has
swept the developing world, the global distribution of emissions
has shifted dramatically as well. In 1990, developed countries
accounted for 65 percent of all emissions. In 2011, they
accounted for 42 percent. Since 2000, developing countries have
Figure 13.6 Global CO2 emissions from fossil fuel use and
industry since 1959 and emissions intensity CO2/GDP. From the
2018 Global Carbon Budget analysis. Economic activity is
measured in purchasing power parity (PPP) terms in
2010 US dollars.
exhibited a consistent annual increase in emissions of 0.24 PgC

per year, while the developed countries emissions have remained
steady. While slower growth or reductions in the United States
and Europe could help reduce emissions substantially, another
key determinant of twenty-first-century global health will be the
energy economies of Asia and the southern hemisphere.
The Global Carbon Project has published a recent
assessment of emission projections.17 In this opinion piece they
focus on our need to reduce our Emissions Intensity – the
amount of CO2 emissions necessary to produce a certain
amount of productivity, i.e., emissions divided by GDP. Global
emissions are increasing as emissions intensities decrease
(Figure 13.6). Like many global trends, we find opposing ten-
dencies. Our economies are growing, leading to increased emis-
sions, but our economies are also growing more efficient. To
reduce emissions while feeding, housing, and employing a
17
Jackson, R. B., et al. “Reaching peak emissions.” Nature Climate Change (2015),
doi:10.1038/nclimate2892.
rapidly growing global population, we need to rapidly transition

away from coal and oil, adopting cleaner and more renewable
sources energy and industries.
A recent Nature Commentary by some of the Global
Carbon Project scientists18 stresses the need to convert renew-
able energy sources. Their message is directly in line with the
story of this book:
Every year of rising emissions puts economies and the
homes, lives and livelihoods of billions of people at risk.
It commits us to the effects of climate change for centur-
ies to come. Already, the terrible impacts of 1C of
warming above pre-industrial levels are evident.
Disasters triggered by weather and climate in 2017 cost
the global economy US$320 billion, and around 10,000
lives were lost.19 The full costs of 2018’s disasters have
yet to be tallied – including Typhoon Mangkhut, hurri-
canes Florence and Michael, and the heatwaves and
wildfires that have ravaged swathes of Europe and the
United States. These events are likely to contribute to an
exponential rise in damages, amounting to some $2.2
trillion over the past two decades.20
When it comes to rises in global average temperature,
every fraction of a degree matters. A report published in
October by the Intergovernmental Panel on Climate
Change (IPCC) projected devastating impacts at 2C.
These include the loss of almost all the world’s coral
reefs, and extreme, life-threatening heatwaves that
could affect more than one-third of the world’s popula-
tion.21 Limiting warming to 1.5C will significantly
lessen those impacts.
Striking, strident, true. We have experienced a warming
of 1C already, and we are already seeing dangerous increases in
climate extremes. Increases in climate extremes that are hurting
18
https://www.nature.com/articles/d41586-018-07585-6.
19
https://www.munichre.com/topics-online/en/climate-change-and-natural-
disasters/natural-disasters/2017-year-in-figures.html.
20 21
https://www.unisdr.org/archive/61121. https://www.ipcc.ch/sr15/.
Figure 13.7 Cost for solar power in US dollars per Watt.

From Figueres et al. (2018)
people. But this Nature Commentary also stresses reasons for

hope. Alternative energy is exactly that – a viable alternative.
Solar power has dropped from a cost of US$100 per Watt in
1975 to about 40 cents per Watt now (Figure 13.7). Four
countries – Morocco, Chile, Egypt, and Mexico – are producing
solar power at a cost of less than 3 cents per kilowatt hour, a
price that is cheaper than natural gas. Coal is becoming obso-
lete, yet still poses an existential threat to the planet, especially
in developing nations where it is often the easiest and cheapest
solution. Yet more than half the new energy production capacity
for electricity is renewable, and wind and solar capacity is
doubling every four years.22
22
https://exponentialroadmap.org/.
Summarizing the Known Unknowns

What we know is that there are a lot of known
unknowns about where Terra’s life support system is heading.
Let’s conclude by considering estimated changes in the
2009–2018 global carbon budget, as recently analyzed by The
Global Carbon Budget Project (Figure 13.8). The big thick
arrows show estimates of the anthropogenic carbon dioxide
fluxes. Greenhouse gas emissions and land use changes (deforest-
ation) increase CO2 fluxes into the atmosphere (+35 and +6
gigatons per year, respectively). These are partially offset by
increased absorption by the land and ocean (about 12 and 9
gigatons per year, respectively). The difference between these
inputs and extractions is the increase in atmospheric CO2, a big
+20 gigatons per year.
This is all plenty scary. But we are still rolling around
on top of the Dixie® Cup. Notice the nice thin arrows in
Figure 13.8 – the ones marked “carbon cycling GtCO2 per
year.” These are natural exchanges between the atmosphere
and land surface. And they are really big – about 440 and 330
gigatons per year for the land and ocean, respectively. These big
numbers mean there is a big natural exchange in CO2, an
exchange we are deeply involved with every time we breathe or
eat a carrot. On land, soils, permafrost, and plants store carbon.
Fossil CO2 Land-use change Biosphere Atmospheric CO2 Ocean

Anthropogenic fluxes
2009-2018 average
+18 GtCO2 per year
6 Carbon cycling
(3 - 8) 9 GtCO2 per year
0.5 12 (7 - 11) Stocks
(9 - 14)
440
35
(33 - 37)
330
Vegetation Dissolved
440
Gas reserves inorganic carbon
330
Rivers Organic carbon Marine
Permafrost
Oil reserves Soils and lakes biota
Coasts Surface
sediments
Coal reserves
Budget imbalance +2
Figure 13.8 Human-induced changes in the 2010–2019 Global

Carbon Budget, from the Global Carbon Budget.
In the ocean, sediments, plants, and animals store some carbon,

but a whole lot of carbon also ends up dissolved in water. This
dissolved carbon makes the ocean much more acidic.
There is a very real risk that global warming may muck
up the basic balances that keep our Dixie® Cup planet centered
on the up and up. On land, rising temperatures might kill soil
microbes, or melt permafrost, or kill trees. Perhaps all three
could happen together. These perils occur at scales and with
complexities not captured well by our current models. So too we
face very serious challenges when modeling glaciers and Arctic
and Antarctic sea ice. These are highly nonlinear systems that
we don’t understand very well. And then there are temperature
and ocean acidification impacts on the ocean, harming coral
reefs, oysters, mussels, sea urchins, starfish, plants algae, and
fish. Ocean acidification, for example, makes it harder for coral
reefs to rebuild themselves, so that “coral reefs will transition to
net dissolving before end of century.”23
Each of these negative impacts are concerning taken
alone. And each of these, and other unknown time bombs,
might seriously shift the subtle balance of carbon dioxide and
the Earth’s energy balance in a potentially explosive manner.
There could be catastrophic losses. Trees and reefs die. We
know that during the very large 2015/16 El Niño, coral reefs
faced unprecedented bleaching (i.e, death), while the Amazon
rainforest suffered and the absorption of C02 diminished sub-
stantially. If such strong El Niños become more common or
more intense, and occur in a world with a warmer ocean and
atmosphere and a more acidic ocean, it is very plausible that we
will both see rapid and lasting CO2 increases. Pushed beyond a
limit, ocean and tropical ecosystems fail. And they are likely to
fail just when we need them most.
23
http://science.sciencemag.org/content/359/6378/908.
WE CAN AFFORD TO WEAR
14 A WHITE HAT
It is better to light a candle than to curse the darkness.
W. L. Watkinson1
Everybody likes Rick Blaine in Casablanca, Rooster Cogburn in

True Grit, Han Solo in Star Wars: bad people who can’t help
but end up being good. Against their better judgment they do
the right thing. We love their drama because it is our daily
struggle. We are just regular girls and guys, not pretentious
Jedi or suave Czechoslovakian resistance fighters. We want to
keep our heads down, drive our cars, fly on planes, make and
save money for our children.
But we also feel. And feel compassion. In an increasingly
connected world, we are learning that climate change is making
weather and climate extremes more intense and frequent,
placing people in harm’s way (Figure 14.1).
And we also have faith in our religion, our philosophy,
our culture, our science. As the historian Yuval Harari has
1
Watkinson, W. L., and H. Fleming. “The invincible strategy,” in The Supreme
Conquest and Other Sermons Preached in America. New York: Revell Company,
1907.
Figure 14.1 Lives in the balance. A malnourished child looks up

at a scale as he is weighed and treated at Banadir hospital
in Mogadishu. UN Photo/Tobin Jones, March 9, 2017,
Mogadishu, Somalia, Photo # 716500
289 / We Can Afford to Wear a White Hat
pointed out,2 these “stories” that we tell ourselves permeate our

lives, structuring our body politic. Stories inform how and if our
civilization self-organizes and works, or doesn’t, for good or for
evil. Faith is the engine of history. We all believe in something.
Some of us believe that there is little to believe in: best perhaps to
just pursue wealth while promulgating doubt. Others assume
that we are powerless to stop environmental destruction, best
perhaps to just pursue wealth while promulgating liberal lamen-
tations. Both of these perspectives are wrong. We don’t need to
believe in climate change, we can see and understand it. We
don’t need to accept climate change, we can radically transform
our energy economies while improving our early warning and
forecasting systems.
If you look back over human history, you will certainly
find tragedy on catastrophic scales, but you also can’t deny that
modernity has worked out pretty well for many people. Yuval
Harari tells us, “History is something that very few people have
been doing while everyone else was plowing fields and carrying
water buckets.” Today most of us are free from lives of grueling
labor, and we enjoy tremendous opportunities to express our-
selves, learn, and recreate. For most of history many of our
children died young, one of the most tragic of all losses. Just
since 1991, the number of deaths for children under five has
dropped by more than 50 percent3 (Figure 14.2). In 1991, a
stunning number of children in the world died before they were
five: almost one out of every ten. By 2019, that number was still
too high – one out of about twenty – but we have made sub-
stantial progress. Since 2000, this improvement has saved the
lives of more than 50 million children. That is about the same
number of people who died fighting in World War II. If we act
with compassion, guided by science, together we can achieve
great things. We can share faith in our civilization, science, and
humanitarian ideals. Hope leads us to work together to effect
2
Harari, Yuval Noah. Sapiens: A Brief History of Humankind, New York:
HarperCollins, 2015.
3
https://data.unicef.org/topic/child-survival/under-five-mortality/.
Figure 14.2 UNICEF child mortality data. The number of deaths

per year per 1,000 children under 5. https://data.unicef.org/topic/
child-survival/under-five-mortality/
positive change. Compassion for the poor, the hungry, and our
beautiful fragile planet gives us courage – true grit sufficient
enough to wear the White Hat. Rick Blaine, Rooster Cogburn,
and Han Solo teach us that being blameless is not good enough.
Maybe it’s better to drink whiskey, fight, and yet be actively
good. Action links faith and change, producing progress.
Famine, or the lack thereof, provides a sterling example
of what we can achieve. At the close of the nineteenth century,
famines wracked Europe and Africa. In India and China, more
than 30 million people lost their lives.4 In the 1940s, 1950s, and
1960s, more than 43 million people perished of famine.5 Today,
despite the existence of a massive number of extremely hungry
people – more than 88 million in 2020 – famine-related
4
Third World. Verso Books, 2002.
5
Hasell, Joe, and Max Roser. “Famines.” Published online at OurWorldInData.org
(2020). Retrieved from: https://ourworldindata.org/famines.
Figure 14.3 German carbon dioxide emissions. Data based on

the Global Carbon Project archive.
mortality rates are very low. This reduction, achieved through

effective international collaborations that connect the power of
Earth science with humanitarian relief agencies is “one of the
great unacknowledged triumphs of our lifetime.”6
Just as humanity has tackled famine and infant mortal-
ity, so too fighting climate change is very much in reach. We can
be both wealthy and wise. For example, Germany has invested
in a highly efficient, high-tech, highly renewable economy.
Figure 14.3 shows a time series of annual German carbon diox-
ide emissions. Since a peak in 1979, emissions have dropped by
one-third to about 207 megatons of CO2 in 2018. In that same
time period, Germany’s gross domestic product increased by
more than 400 percent, going from US$878 billion to US$4
trillion in 2018. In 2019, renewable energy sources accounted
for more than 34 percent of Germany’s electricity production.
6
de Waal, A. “The end of famine? Prospects for the elimination of mass starvation
by political action.” Political Geography (2017).
Figure 14.4 California Emissions and Per Capita Emissions.
Investments in education, infrastructure, and worker education

created a sustainable, lucrative, and equitable economy in which
many people have good paying jobs.
In the United States, the state of California provides
another relevant example.7 Despite a rapidly growing economy,
Californian investments in alternative energy and commitments
on minimizing emissions have resulted in substantial emission
declines since 2000 (Figure 14.4). Per capita emissions declined
by about 24 percent from peak levels, and actual emissions by
about 14 percent.
We can also look at California’s emissions as a function
of gross domestic product (GDP) (Figure 14.5). “Energy
Intensity” measures the amount of CO2 emissions needed to
7
https://ww3.arb.ca.gov/cc/inventory/pubs/reports/2000_2017/ghg_inventory_
trends_00-17.pdf
Figure 14.5 California GDP and energy intensities.
produce a certain amount of GDP. Since 2001, California’s

energy intensity has declined by 38 percent. Over the same time
period, the economy experienced a massive growth of 41 per-
cent. Investments in green technologies and increased efficiency
can be completely compatible with rapid economic growth.
Why Our Human “Time Bomb” Can Be a Good Thing

We should build history together.
We should realize that we face an exciting “time bomb,”
rife with potential. The twenty-first-century “population explo-
sion” carries within it the seeds of our own resurrection.
Each life is franticly unique. Like the Little Prince in
Antoine De Saint-Exupéry’s story, who despairs when he real-
izes that there are thousands of roses identical to the one he
loves, we often forget how magically distinctive, utterly
irreplaceable, and existentially exceptional each human life is.

We all have our own unique past, our worldview, our own
thoughts, emotions, and insights. Within our diverse and
increasingly educated billions, we hold a tremendous ability to
innovate and self-organize, to tell stories and create.
But being alive is only a necessary, but not sufficient,
cause of consciousness. Having a life is not the same as owning
your life. Living, by itself, does not make you conscious. For that,
we need to pay careful attention to both the external world and
the world within. Of course, that requires effort, but that effort
brings rewards. With luck, each of us may have the gift of about
three billion precious seconds – 90 years 365 days 24 hours
60 minutes 60 seconds. How many of these seconds will we
claim as ours? This may make all the difference. Are we doing or
being done to? Are we paying attention to the world around us, to
Nature and our fellow travelers on Spaceship Earth? Or do we
fritter away our moments consumed by consumption, competi-
tion, and distraction? As the climate crisis draws us all together,
inattention may truly pose an existential threat.
As we have seen in the previous thirteen chapters, the
rate of climate change is alarming. We are already experiencing
extreme impacts. But our potential for effective intervention is
also immense.
I like to think of this potential as a “time bomb,” with
“bomb” used in a positive sense of explosive potential. As
Figure 14.6 Global population and equivalent quantities of

human people-years.
Figure 14.6 illustrates, between 1961 and 2050, humanity’s time

bomb will execute humanity’s greatest experiment in parallel
processing, as billions of individuals grow, think, discover, and
consume. We will do far more, learn far more, and consume far
more than any previous generation. Between 1961 and 2050, the
Earth’s population will triple, going from approximately 3 billion
in 1961 to more than 9 billion in 2050.8 As our population
rapidly expands, however, our industry and technology will also
enlarge rapidly. So we are currently in the middle of a very
exciting, very fraught, explosion of human activity. Many things
will be happening at once as we accelerate our consumption,
creativity, and interconnectedness. Paying attention right now,
seeing the role of climate change in the world today, is critical.
We can let this mess run unfettered or, like Germany and
California, we can grow wise but continue growing.
We can appreciate this urgency by breaking all of
human history into four equivalent blocks of people-years, or
“human capacity” (Figure 14.6). The y-axis represents global
population estimates from 8000 BC through 2050 AD. The
“hockey stick” structure clearly shows the rapid population
growth of the modern era. That rapid growth is certainly
placing strains on our planet. What is less appreciated, however,
is how greatly we are expanding our human capacity (for good
or ill). In analogy with the idea of “horsepower,” we can define
human capacity as the total number of people-years over a given
span of time. Human capacity is calculated by adding up all the
people living on the planet each year. Adding up all the people-
years between 1960 and 2050, we get an estimated total of
631 billion person-years. This staggering amount of time is
about forty-nine times the age of our universe (~13
billion years).
Another way to contemplate this temporal mass of
humanity is to see it as roughly equivalent to the total number
of global person-years between 1000 and 1960. The 90 years
between 1961 and 2050 will see as much human capacity as
8
Based on the United Nations medium variant projections.
approximately the 960 years prior (1000 AD to 1959). If we go

back even further, in the two and a half millennia between
1500 BC and 1000 AD we find about 631 billion person-years.
Finally, in the 6,500 years between 1500 BC and 8000 BC,
approximately 631 billion person-years brings us back to time
of the Neolithic revolution that saw the first continuous human
settlements, the advent of farming, and the first domestication
of animals.
Thus in just 90 years, 1961–2050, the Earth will witness
an explosion of human activity broadly equivalent to the previ-
ous 960, 2,500, and 6,500 years before that. Furthermore, we
are now mostly industrialized, far wealthier, far better educated,
linked electronically, computer-enhanced, and sensor-enabled.
We can also build on the knowledge of all the previous gener-
ations. A lot is likely to happen very quickly.
We will be able to wear the White Hat, if we want to
pay attention. Over the past fifty years, our creative capacities
have improved as education expanded, farmers became more
productive, and research, development, and entrepreneurialism
spread into Asia, South America, and Africa. There is good news
on many fronts. Rates of extreme poverty are falling. The
number of people without access to water has dropped by
50 percent since 1990. Overall global per capita production (a
measure of our total economic production divided by the total
number of people) has gone from $450 (in 2017 dollars) per
person in 1960 to $11,312 in 2017. That is a stunning 21-fold
increase in global per capita production – a measure of the
power of our three billion seconds in action. The increase in
total global gross domestic product has been incredible, and
now stands at nearly $90 trillion a year (Figure 14.7), doubling
just since 2004.
The world’s rapid economic growth, however, is also a
measure of our two-sided “time bomb.” Our increased wealth
(Figure 14.7), combined with our increased population
(Figure 14.6), has led to massive increases in greenhouse gas
emissions. Yet on the other side of the coin we could be ready to
cut away inefficiency, develop new green technologies, and build
Figure 14.7 World Bank estimates of global gross domestic

product in 2017 US dollars.
a more sustainable future, as Germany and California are

attempting to do. And as our creativity and interconnections
increase, magnified by rapid increases in global education and
scientific research, and augmented by ever-faster computers,
models, and sensors (satellites and various forms of microscopic
scanners), we are already well under way to an “imaging revo-
lution.” Never have so many seen so much so well. We use
satellites and sensors to probe the edges of time and the limits
of space. We also use satellites and sensors to map and monitor
our living world. We are also, as never before, teaching our-
selves to learn.
Since 1970, we have seen the proportion of children
receiving primary education rise from 72 percent to 89 percent
(Figure 14.8). Not only has education expanded in general, but
the gap between primary education for girls and boys has disap-
peared (at least globally, on average). Yet at more advanced
Figure 14.8 Percent of children receiving primary

school education.
Source: World Bank
education levels, women in poorer countries still face substantial

gaps, and women from the poorest families still face the highest
barriers to education.
And as we have grown better educated, we have become
more innovative. As wealth and education spread, the world has
produced many more scientists, engineers, and inventors. Just
since 1985, the number of global patent applications has quad-
rupled (Figure 14.9), reaching more than two million in 2016, a
figure twice that of a decade before. Together, we are product-
ive, educated, and innovative in ways our early ancestors could
not even begin to imagine.
But now we have come to an era in which our very
success may be our undoing. There is a lot going on, very
rapidly. This incredible increase in productivity may come at a
very steep price – because there has been a very strong coupling
Figure 14.9 Global patent applications.

Source: World Bank
between our increase in production and the increase in our

greenhouse gas emissions. But hope for the cure may be embed-
ded in the cause. Lifted out of abject poverty in mere decades,
our global billions certainly have the capacity to effectively
tackle climate change. Given the will.
The need for this effort is profound. Right now we
appear to be headed for 3C of warming or more. This level of
warming would almost certainly have catastrophic and poten-
tially irreversible impacts on our planet’s life support system.
This book has documented the dangerous and expensive
increase in extremes associated with just the 1C warming we
have already experienced to date. At +3C or +4C, the health
impacts of heat waves would become terrible killers for billions
of people in hot, humid regions. Crop production in the tropics,
and especially Africa, would be dramatically impacted. A +3C
or +4C world will bring much more frequent superstorms,
riding across a rising tide of higher sea levels. And then there
might be very real risks associated with positive feedbacks:
melting ice caps, dead coral reefs, acid deoxygenated seas, and
reduced carbon uptake from tropical forests could lead (very
easily) to runaway warming.
Even the difference between 1.5C and 2C of warming
will make an incredible difference. According to the
Intergovernmental Panel on Climate Change‘s Special Report
on the Impacts of Global Warming of +1.5C,9 having a total
2C of warming versus 1.5C would:
Double the number of people exposed to extreme heat waves

(limiting the total to 15% as opposed to 30%);
Increase the area of melting permafrost by about two million
km2. This is an area four times the size of France;
Double or triple the number of species experiencing severe
habitat loss;
Greatly increase the risk of irreversible loss of sea ice;
Greatly increase the damage to coral reefs and fisheries via the
impacts of very warm ocean temperatures, acidification, and
deoxygenation;
Increase the chance of an ice free Arctic ocean tenfold;
Increase the number of people exposed to water stress by 50
percent;
Increase by hundreds of millions the number of poor, food-
insecure people exposed to climate risks.
The data we have explored together in this book strongly sup-

ports these concerns. As we saw in Chapter 4, the 2015–2019
time period has been a lot warmer than even the 1990s, about
0.7C degrees warmer on land (on average). But this warming is
not spread evenly, but rather acts in a more dangerous way. At
any given time we find areas of exceptionally warm land and
ocean temperatures (Figure 5.3); the fraction of the Earth with
exceptional temperatures has tripled from the early 2000s,
9
The International Panel on Climate Change Special Report: “Global Warming of
1.5C,” www.ipcc.ch/sr15/.
going from 6 percent to 18 percent. This warming, combined

with population growth, has led to a stunning increase in the
number of people exposed to heat waves each year (Figure 5.6).
In Chapters 8, 9, and 10 we have explored the dangerous
implications of a 0.5C to 0.7C warming in the tropical oceans.
This “extra” warming can exacerbate the intensity of naturally
occurring climate phenomena like the Indian Ocean Dipole, El
Niños, and La Niñas. Fore-armed with this conceptual model of
climate change, however, we can translate these superwarm anom-
alies into opportunities for prediction. “Storyline” approaches to
climate attribution have linked climate change to deadly droughts
that have already helped push tens of millions more people into
food insecurity, risking a reversal of decades of progress.10 With
our own eyes, we have seen how climate change made the
2015–2016 El Niño and the 2016–2017 La Niña more dangerous.
And with our own eyes we have seen how the most
intense rainfall events are increasing in intensity (Figure 6.5),
and how these extremes have affected hundreds of millions of
people just since 2015 (Table 6.1). Focusing on the United
States, where there is adequate data to track disasters quite
precisely over time, we find huge increases in hurricane damages
(Figure 7.4) and wildfire extent (Figure 11.3).
But seeing alone is not enough. We need to understand.
To this end we have explored and explained how the Earth’s
radiation balance (and imbalance) works. Absorbing and re-
emitting the sun’s energy, the Earth provides a “fragile flame”
that translates solar energy into complex patterns, miraculous
life, and billions of astounding human stories. We are messing
heavily with the life support system of our living planet, the only
planet we know of that supports life. There is no planet B.
One recurring theme has been how incredibly thin the
Earth’s atmosphere is. While the sky seems vast when we look
up, it is rather empty; there just isn’t that much there up there.
10
The 2018 United Nations Food and Agriculture Organizations annual “The State
of Food Security and Nutrition Report.” www.fao.org/3/I9553EN/i9553en.pdf
Figure 14.10 The Earth as seen from the International

Space Station.
Source: NASA astronaut Scott Kelly
For me, pictures taken by humans from the

International Space Station evoke this understanding. Knowing
that a human heart and mind reside behind the lens evokes a
sense of intimacy. Taken from the edge of space, these photos
illustrate the beauty of Earth but also the shallowness of our
atmosphere. NASA astronaut Scott Kelly captured a beautiful
example11 on September 22, 2015 (Figure 14.10). On the left of
this image the Nile River stretches across Egypt and Sudan. The
band across the top shows the “airglow” at the very topmost
layer of our atmosphere – the ionosphere.12
Seen from this perspective, we can begin to understand
why dumping 37 gigatons of carbon dioxide a year into our
atmosphere is like five people chain-smoking in a small car with
the windows rolled up. That is not going to end well, and it is
not going well now. Increases in human exposure to weather-
related dangers are combining with increased climate volatility,
11
www.nasa.gov/image-feature/the-nile-at-night.
12
https://ourplnt.com/secrets-earth-multi-colored-airglow/#axzz6FTaOH9eP
costing us billions of dollars and yuan and euros and rupees

each year, and impacting millions of people each year.
So what might it cost us to seriously tackle climate
change? The IPCC Impacts of Global Warming of +1.5C
report looked at this issue. While a detailed treatment is beyond
the scope of this book, we can conclude Drought, Flood, Fire
with a brief synopsis of the report’s suggestions:
Reducing carbon and methane emissions by 45 percent by

2030, reaching zero emissions by 2050;
Switching very rapidly to renewable energy sources;
Rapidly improving the efficiency with which we use energy;
Stopping the burning of coal;
Switching to biofuels;
Planting billions of trees to absorb more CO2;
Increasing investments in low-carbon energy sources and
increased energy efficiency;
Taking an “all of the above” approach – simultaneously
pursuing reductions in emissions, increases in efficiencies,
and increases in terrestrial carbon uptake – is essential.
Can we afford to make these changes? Absolutely. It

won’t be cheap, but it is plausible, and our kids are worth it.
According to the +1.5C Special Report, “additional annual
average energy-related investments for the period 2016 to
2050 . . . are estimated to be around US$830 billion.” This
number is associated with changes that transition us to renew-
ables while increasing our efficiency. This a big number, but is
only about 1 percent of the total 2018 global gross domestic
product (US$85 trillion). For reference, we can note that
between 2001 and 2018, the United States is estimated to have
spent $5.9 trillion on wars and military action in Afghanistan,
Iraq, Syria, and Pakistan.13,14 That works out to about $328
13
https://watson.brown.edu/costsofwar/files/cow/imce/papers/2018/Crawford_
Costs%20of%20War%20Estimates%20Through%20FY2019%20.pdf.
14
www.cnbc.com/2018/11/14/us-has-spent-5point9-trillion-on-middle-east-asia-
wars-since-2001-study.html.
billion per year, or about 1.5 percent of our estimated US

2018 GDP of US$20.5 trillion. The global costs of weather-
related disasters in 2017 and 2018 have been estimated to be
similar in magnitude – about $325 billion in damages each
year.15 The point here is not to advocate specific mitigation
approaches, but rather to point out that the data tell us that
we are already essentially “at war” with the climate. The mag-
nitude of the economic and humanitarian impacts of this war
are already large, and are almost certainly going to increase. But
the cost of acting is relatively modest. What’s $830 billion
between friends in a $90-trillion global economy? We will end
up paying one way or another, and the right thing to do, from a
moral, economic, and environmental perspective, is to pay now
for an inevitable energy transition.
Here is where the stories we tell each other matter,
coordinating our intercontinental “time bomb” into explosive
positive change. We need to act coherently and quickly to avoid
rapid warming. We need to see and understand climate change
as something that is hurting people and our planet now. We
have been blessed with a miraculous planet, and the opportunity
to live through an exciting and momentous time, filled with
increasing prosperity and innovation. We can wear the White
Hat. Compassion and logic demand that we do so. The great
spiritual leaders taught us that “peace is a verb,” telling us
stories that eventually led to unprecedented peace and prosper-
ity.16 We need to make peace with our planet. Climate change is
hurting people and our planet; we need to act to prevent that,
and we can afford to do so.
15
http://thoughtleadership.aonbenfield.com/Documents/20190122-ab-if-annual-
weather-climate-report-2018.pdf.
16
Pinker, Steven. The better angels of our nature: Why violence has declined.
Penguin Group USA, 2012.
APPENDIX
A Few Resources for Further Reading
and Research
Please note that this book is intended as an accessible
starting point to help readers learn about recent climate
extremes. While it is not my intent to review or rigorously
represent the current scientific literature, below I provide brief
sections that provide pointers for you to delve deeper into the
various topics covered in this book.
Chapter 1
The Intergovernmental Panel on Climate Change
(IPCC) Reports1 provide extensive and authoritative climate
change analyses compiled by panels of international experts.
Recent reports on the dangerous impacts of +1.5C and +2C
of warming, climate change and land,2 and the ocean and cryo-
sphere in a changing climate3 provide valuable assessments of
climate change threats. The forthcoming 6th IPCC Assessment
Report4 will contain an exhaustive and detailed analysis with
contributions from hundreds of the best climate scientists. The
1
Intergovernmental Panel on Climate Change (IPCC), www.ipcc.ch/
2
IPCC Special Report on Climate Change and Land, www.ipcc.ch/report/srccl/
3
IPCC Special Report on the Ocean and Cryosphere in a Changing Climate, www
.ipcc.ch/report/srocc/
4
IPCC 6th Assessment Report - www.ipcc.ch/report/sixth-assessment-report-cycle/
306 / Appendix
annual issues of the Bulletin of the American Meteorological

Society on Explaining Extreme Events from a Climate
Perspective5 provide timely analyses intended for the general
public. Each year’s special issue provides a series of short articles
that analyze the potential role that climate change played in a
single extreme climate event. The National Academy of
Sciences’ Attribution of Extreme Weather Events in the
Context of Climate Change provides an excellent overview of
extreme event attribution. The International Detection and
Attribution Group,6 the World Climate Research Program’s
Grand Challenge on Weather and Climate Extremes,7 the dis-
tributed ClimatePrediction.net,8 the Climate Central World
Weather Attribution Group,9 the World Weather Attribution
Group,10 and many other excellent research teams constantly
publish research on extreme attribution.
For those interested in a good introduction to climate
change, Jeffrey Bennett’s A Global Warming Primer11 is a good
place to start, as is Joseph Romm’s Climate Change: What
Everyone Needs to Know.12 The edited volume Climate
Extremes: Patterns and Mechanisms13 has articles on many
recent topics.
5
Bulletin of the American Meteorological Society Annual Explaining Extreme
Event Special Issues, www.ametsoc.org/ams/index.cfm/publications/bulletin-of-
the-american-meteorological-society-bams/explaining-extreme-events-from-a-
climate-perspective/
6
Website of the International climate change Detection and Attribution Working
Group, www.clivar.org/clivar-panels/etccdi/idag/international-detection-
attribution-group-idag
7
World Climate Research Program Extremes Events program, www.wcrp-climate
.org/gc-extreme-events
8
Website for the climateprediction.net project, www.climateprediction.net/
9
Climate Central World Weather attribution page, www.climatecentral.org/go/
wwa
10
The World Weather Attribution Group website, www.worldweatherattribution
.org/
11
Bennett, Jeffrey O. A Global Warming Primer: Answering Your Questions about
the Science, the Consequences, and the Solutions. Big Kid Science, 2016.
12
Romm, Joseph. Climate Change: What Everyone Needs to Know. Oxford
University Press, 2018.
13
Wang, S-Y. Simon, et al., eds. Climate Extremes: Patterns and Mechanisms.
Vol. 226. John Wiley & Sons, 2017.
307 / A Few Resources for Further Reading and Research
Chapter 2
Chapter 2 draws heavily from Jan Zalasiewicz and
Mark Williams’s excellent book The Goldilocks Planet: The
4 Billion Year Story of Earth’s Climate.14 Fred Adam’s Origins
of Existence: How Life Emerged in the Universe15 also
informed this discussion. Several Scientific American articles
by Caleb Scharf, a British-born astronomer and the director
of the multidisciplinary Columbia Astrobiology Center at
Columbia University in New York, helped guide the discussion
of Blue, Green, and Red galaxies. You might be interested in
How Black Holes Shape the Galaxies, Stars and Planets around
Them16 or Is Earth’s Life Unique in the Universe?17 NASA
provides a great summary of Cosmology and the Big Bang
Theory at map.gsfc.nasa.gov/universe/. The El Niño-Southern
Oscillation blog posts (www.climate.gov/news-features/blogs/
enso/index-page-enso-blog-posts), written by Michelle
L’Heureux, Nat Johnson, and Tom Dilberto, featured (briefly)
in Figure 2.6, provide both good background material and
interesting up-to-date analyses. Global Physical Climatology18
by Dennis Hartmann provides a great overview of the General
Circulation. Meteorology Today: An Introduction to Weather,
Climate, and the Environment, 9th Edition19 by Donald
Ahrens is a deservedly popular treatment of these topics in
accessible format.
14
Zalasiewicz, Jan, and Mark Williams. The Goldilocks Planet: The 4 Billion Year
Story of Earth’s climate. Oxford University Press, 2012.
15
Adams, Fred C. Origins of Existence: How Life Emerged in the Universe. Simon
and Schuster, 2010.
16
Scharf, Caleb, Black Holes Shape the Galaxies, Stars and Planets around Them,
www.scientificamerican.com/article/how-black-holes-shape-galaxies-stars-
planets-around-them/.
17
Scharf, Caleb, Is Earth’s Life Unique in the Universe?, www.scientificamerican
.com/article/is-earth-s-life-unique-in-the-universe/.
18
Hartmann, Dennis L. Global Physical Climatology. Vol. 103. Newnes, 2015.
19
Ahrens, C. Donald. Meteorology Today: An Introduction to Weather, Climate,
and the Environment. Cengage Learning, 2012.
308 / Appendix
Chapter 3
This chapter (like Chapter 2) draws heavily from Jan
Zalasiewicz and Mark Williams’s excellent book The
Goldilocks Planet: The 4 Billion Year Story of Earth’s Climate
and Fred Adam’s Origins of Existence: How Life Emerged in
the Universe. Global Physical Climatology by Dennis Hartmann
and Meteorology Today: An Introduction to Weather, Climate,
and the Environment, 9th Edition, by Donald Ahrens provide
great introductions to climatology and meteorology. The book
Atmospheric Thermodynamics by Craig F. Bohren and Bruce
A. Albrecht,20 while very technical and not for the faint of heart,
is excellent, and has inspired certain aspects of this work, includ-
ing my attempts to mix humor and science.
Chapter 4
Following up on the DIY aspect of Chapter 4, I will
mention here some of the many excellent online resources for
examining long time series of weather and climate data. Please
note, however, that this is not intended to be an authoritative
review. I am simply listing some of the resources that I have used
in this book. The KNML climate explorer (climexp.knmi.nl)
provides access to a vast climate archive. This access portal
hosts a large number of observational data sets, as well as a
large number of climate change simulations, making it an excep-
tional resource for the community. NOAA’s Earth System
Research Laboratory has a rich set of online tools (www.esrl
.noaa.gov/psd/cgi-bin/data/getpage.pl), and the International
Research Institute’s Data Library (iridl.ldeo.columbia.edu/
index.html) is a powerful and freely accessible online data
repository and analysis tool that allows a user to view, analyze,
and download hundreds of terabytes of climate-related data
through a standard web browser.
20
Bohren, Craig F., and Bruce A. Albrecht. Atmospheric Thermodynamics. Oxford
University Press, 2000.
Chapter 5
The following resources are likely to be useful to inter-
ested readers. The World Climate Research Program group on
Weather and Climate Extremes website in general, and their
special issue in the journal Weather and Climate Extremes in
particular,21 provides a valuable collection of articles. A 2019
Lancet article22 provides a good overview of current impacts of
temperature extremes and the literature related to them. In gen-
eral, this chapter has broadly followed the approach used by the
Lancet article to estimate heat exposure events, though there are
many different ways heat exposure indices can be estimated. It
should also be noted that this chapter did not attempt to discuss
in detail many other important aspects of temperature extremes,
such as negative crop impacts, wildfire risk, reductions in worker
productivity, increases in conflict, and increased disease trans-
mission rates. The 2019 Intergovernmental Panel on Climate
Change reports on Climate Change and Land and Global
Warming of 1.5C23 provide relevant material discussing these
impacts, as will the 6th Assessment Report on Impacts,
Adaptation and Vulnerability.24
Chapter 6
Once again, The World Climate Research Program
group on Weather and Climate Extremes website in general,
and their special issue in the journal Weather and Climate
Extremes in particular, will be of value to interested readers.
In this chapter, we followed the general approach used in a
21
Weather and Climate Extremes, www.sciencedirect.com/journal/weather-and-
climate-extremes/.
22
Watts, Nick, et al. “The 2019 report of The Lancet Countdown on health and
climate change: ensuring that the health of a child born today is not defined by a
changing climate.” The Lancet 394.10211 (2019): 1836–1878.www.thelancet
.com/journals/lancet/article/PIIS0140-6736(19)32596-6/fulltext.
23
IPCC, Special Report: Global Warming of 1.5ºC, www.ipcc.ch/sr15/.
24
IPCC. 6th Assessment Report on Impacts, Adaptation and Vulnerability, www
.ipcc.ch/report/sixth-assessment-report-working-group-ii/.
310 / Appendix
seminal Nature Climate Change study by Markus Donat and

coauthors.25 A 2019 special issue of Environmental Research
Letters provides a useful collection of recent research,26 and a
2013 edited volume Extremes in a Changing Climate provides
an excellent set of technical articles on extremes. The annual
Global Assessment Reports on Disaster Risk Reduction
(GAR)27 are the flagship reports of the United Nations on
worldwide efforts to reduce disaster risk. The GAR is published
biennially by the UN Office for Disaster Risk Reduction
(UNDRR), and is the product of the contributions of sovereign,
public, and private disaster risk-related science and research,
among others. The 2019 GAR report discusses weather
extremes, climate change, and disaster risk management.
Disaster statistics can be obtained from the EM-DAT database,
the US Billion Dollars disasters database,28 and the Munich Re
natural catastrophe database.29 Brief news reports on flood
events are available at floodlist.com.
Chapter 7
Two excellent attribution studies by Wehner and
Risser and Patricola and Wehner31 were featured in this
30
chapter.
25
Donat, Markus G., et al. “More extreme precipitation in the world’s dry and wet
regions.” Nature Climate Change 6.5 (2016): 508.
26
Environmental Research Letters Focus on Extreme Precipitation, iopscience.iop
.org/journal/1748-9326/page/Focus_on_Extreme_Precipitation_Observations_
and_Process_Understanding.
27
United Nations Office for Disaster Risk Reduction, 2019 Global Assessment
Report on Disaster Risk Reduction, www.gar.undrr.org/sites/default/files/
reports/2019-05/full_gar_report.pdf.
28
NOAA Billion Dollar Disasters, www.ncdc.noaa.gov/billions/
29
Munich Re Database of Catastrophic Disasters, natcatservice.munichre.com/.
30
Risser, Mark D., and Michael F. Wehner. “Attributable human-induced changes
in the likelihood and magnitude of the observed extreme precipitation during
Hurricane Harvey.” Geophysical Research Letters (2017).
31
Patricola, Christina M., and Michael F. Wehner. “Anthropogenic influences on
major tropical cyclone events.” Nature 563.7731 (2018): 339. www.nature.com/
articles/s41586-018-0673-2.
The two multi-authored synthesis reports prepared by

the World Meteorological Organization Task Team on Tropical
Cyclones and Climate Change provide a useful synopsis of
current research. These reports focus on observations32 and
model-projected changes in tropical cyclone activity for a 2C
anthropogenic warming33. The National Academies report on
extreme event attribution34 describes the complexities surround-
ing the analysis of landfalling hurricanes and cyclones, and
provides a good introduction to general topic of climate extreme
attribution. Kevin Trenberth et al.’s 2015 article introduces the
idea of “storyline” attribution,35 and Lloyd and Oreskes36 pro-
vide a very thoughtful and accessible discussion contrasting this
approach to more traditional odds-based attribution
approaches, which builds on Shepard’s 2016 treatment of this
subject.37 How we frame our questions matters.
Chapter 8
Readers interested in finding out more about the
emerging “storyline” approach to climate attribution may be
interested in the accessible and interesting paper Climate change
attribution: When is it appropriate to accept new methods? By
32
Knutson, Thomas, et al. “Tropical Cyclones and Climate Change Assessment:
Part I. Detection and Attribution.” Bulletin of the American Meteorological Society
2019 (2019). journals.ametsoc.org/doi/pdf/10.1175/BAMS-D-18-0189.1.
33
Knutson, Thomas, et al. “Tropical cyclones and climate change assessment: Part
II. projected response to anthropogenic warming.” Bulletin of the American
Meteorological Society 2019 (2019). journals.ametsoc.org/doi/pdf/10.1175/
BAMS-D-18-0194.1.
34
National Academies of Sciences, Engineering, and Medicine. Attribution of
Extreme Weather Events in the Context of Climate Change. National Academies
Press, 2016.
35
Trenberth, K. E., J. T. Fasullo, and T. G. Shepherd. “Attribution of climate
extreme events.” Nature Climate Change, 5 (2015), 725–730. doi.org/10.1038/
nclimate2657.
36
appropriate to accept new methods?.” Earth’s Future 6.3 (2018): 311–325. doi.
org/10.1002/2017EF000665.
37
Shepherd, Theodore G. “A common framework for approaches to extreme event
attribution.” Current Climate Change Reports 2.1 (2016): 28–38. link.springer.
com/article/10.1007%2Fs40641–016-0033-y.
312 / Appendix
Elisabeth Lloyd and Naomi Orestes. This paper contrasts more

traditional “odds-based” attribution approaches with newer
“storyline” approaches, developed by scientists such as
Trenberth, Fasullo, and Sheperd. An accessible 2019 overview
of this topic, quoting Stephanie Herring and Elisabeth Lloyd, is
available at www.slate.com.38 Theodore Sheperd39 has also a led
a relevant 2018 study on this topic published in Climatic Change.
In Chapter 7 we have already seen a very relevant application and
contrast of odd-based and storyline approaches applied to hurri-
canes and tropical cyclones. In Chapter 8, consistent with the
energy-centric focus of this book, I have linked the increasing
energy content of the oceans to potentially dangerous natural
climate fluctuations like the Indian Ocean Dipole. NOAA scien-
tists LuAnn Dahlman and Rebecca Lindsey have written an
excellent blog describing the link between climate change and
ocean heat content.40 Figure 8.1 is an update of the results
presented in their report. The Australian Bureau of
Meteorology provides a good overview of the Indian Ocean
Dipole,41 as well as a video summary42. My concluding thoughts
on Galileo draw from The Case of Galileo: A Closed Question?
by Annibale Fantoli, translated by George V. Coyne.43
Chapter 9
These notes are intended to guide nonspecialists inter-
ested in learning more about El Niño, El Niño impacts, and
38
www.slate.com/technology/2019/12/attribution-science-field-explosion-2010s-
climate-change.html.
39
Shepherd, Theodore G., et al. “Storylines: An alternative approach to representing
uncertainty in physical aspects of climate change.” Climatic Change 151.3–4
(2018): 555–571.link.springer.com/article/10.1007/s10584–018-2317-9.
40
Dahlman, LuAnn, and Rebecca Lindsey. Climate Change: Ocean Heat Content,
www.climate.gov/news-features/understanding-climate/climate-change-ocean-
heat-content.
41
Australian Bureau of Meteorology. Introduction to the Indian Ocean Dipole,
www.bom.gov.au/climate/iod/.
42
www.youtube.com/watch?v=J6hOVatamYs.
43
Fantoli, Annibale. The Case of Galileo: A Closed Question? University of Notre
Dame Press, 2012.
climate change. Mickey Glantz’s Currents of Change: Impacts

of El Niño and La Niña on Climate and Society44 provides a
wonderful introduction. Mike Davis’s Late Victorian
Holocausts: El Niño Famines and the Making of the Third
World45 provides a magisterial melding of global political his-
tory and global environmental history. More about my climate
hero Sir Gilbert Walker can be found in Richard W. Katz’s
accessible historical analysis Sir Gilbert Walker and a
Connection between El Niño and Statistics46. Motivated by
the terrible late nineteenth-century El Niño–related famines that
devastated India, Walker set out on an encyclopedic exploration
of global weather that uncovered the global teleconnections that
set the stage for much of modern climate science. Readers
interested in finding out more about the recent impacts of the
2015–16 El Niño might be interested in the NOAA Bulletin of
the American Meteorological Society‘s (BAMS) Explaining
Extreme Events issues focused on 2015 and 2016. For those
interested in marine ecosystems, the 2016 BAMS attribution
issue has a number of articles focused on impacts of marine
heat waves and their impacts on tropical marine ecosystems.
Many of these results are summarized by Robert S. Webb and
Francisco E. Werner in their article “Explaining Extreme Ocean
Conditions Impacting Living Marine Resources. The annual
United Nations State of Food Security and Nutrition in the
World” (www.fao.org/publications/sofi/) documents the post-
2015 increase in food insecurity and the need to better manage
the effects of El Niño–related climate variability. The forthcom-
ing International Panel on Climate Change 6th Assessment
Report will provide a valuable resource (http://www.ipcc.ch/
assessment-report/ar6/). My personal research has explored the
idea that a substantial proportion of the global warming signal
44
Glantz, Michael H., Currents of Change: Impacts of El Niño and La Niña on
Climate and Society. Cambridge University Press, 2001.
45
Third World. Verso Books, 2002.
46
Katz, Richard W. “Sir Gilbert Walker and a connection between El Nino and
statistics.” Statistical Science (2002): 97–112.
314 / Appendix
can be broken into two components: a high-frequency compon-

ent linked to stronger, warmer El Niños in the equatorial East
Pacific; and a low-frequency “warming mode” linked to
warming trends in the western Pacific and Indian Oceans.47
A paper on this topic is available in the Journal of Climate
(2015). A 2018 paper in the Quarterly Journal of the Royal
Meteorological Society focuses explicitly on the 2015–2016
southern Africa drought (linked to El Niño) and the subsequent
2016–2017 East African droughts, which were tied to excep-
tionally warm West Pacific ocean conditions and La Niña.48 In
this chapter I also draw from two Bulletin of the American
Meteorological Society attribution papers focused on climate
change, the 2015–16 El Niño and extreme droughts and food
insecurity in Ethiopia49 and Southern Africa.50
Chapter 10
Please note: having spent twenty years studying East
African rainfall has led to a longer treatment for this chapter.
There have been too many articles on the East African
rainfall paradox and decline to list them all, but here is a historical
synopsis with some of the highlights. The rainfall decline was first
detected during routine efforts to monitor crop conditions in
Ethiopia in 2003, and documented in a FEWS NET report in
47
Funk, Chris C., and Andrew Hoell. “The leading mode of observed and CMIP5
ENSO-residual sea surface temperatures and associated changes in Indo-Pacific
climate.” Journal of Climate 28.11 (2015): 4309–4329.
48
Funk C., L. Harrison, S. Shukla, C. Pomposi, G. Galu, et al. “Examining the role
of unusually warm Indo-Pacific sea surface temperatures in recent African
org/10.1002/qj.3266.
49
Funk, C., L. Harrison, S. Shukla, A. Hoell, D. Korecha et al. “Assessing the
contributions of local and east Pacific warming to the 2015 droughts in Ethiopia
and Southern Africa,” Bulletin of the American Meteorological Society
(December 2016): S75–S77, doi:10.1175/BAMS-16-0167.1.
50
Funk C., F. Davenport, L. Harrison, T. Magadzire, G. Galu, et al.
“Anthropogenic enhancement of moderate-to-strong El Niños likely contributed
to drought and poor harvests in Southern Africa during 2016,” Bulletin of the
American Meteorological Society, 37 (2017): S1–S3, DOI. 10.1175/BAMS-D-17-
0112.2.
2005 (pdf.usaid.gov/pdf_docs/PNADH997.pdf). I led a paper in

200851 that argued that this decline was tied to warming in the
Indian Ocean. In 2010 and 2011, Park Williams and I published a
USGS report and then a Climate Dynamics52 paper in 2011 that
extended the warming to the broader Indo-Pacific region. These
studies documented the emergent dangers associated with recent
La Niñas, and helped inform a successful FEWS NET drought
forecast during 2010.53 In 2012, Brad Lyon and Dave Dewitt
published an important model-based study54 that argued that
Pacific sea surface temperatures played a dominant role in produ-
cing East African droughts. In 2013, Jessica Tierney and coauthors
used paleo-climate data to link drying in East Africa to warming in
the eastern Indian Ocean.55 A similar analysis by Tierney and
coauthors in 201556 also pointed out a potential partial solution
to the East Africa Climate Paradox – the climate models had a
tendency to dramatically underestimate the March–May East
African rainy season. Several papers led by Andrew Hoell and
Brant Liebmann in 201357 and 201458 emphasized the important
role played by the interaction of the western and eastern Pacific.
The year 2014 also saw important contributions by Wengchang
Yang59 emphasizing the important role of decadal variability in the
51
Funk, Chris, et al. “Warming of the Indian Ocean threatens eastern and southern
African food security but could be mitigated by agricultural development.”
Proceedings of the National Academy of Sciences 105.32 (2008): 11081–11086.
52
Williams, A. Park, and Chris Funk. “A westward extension of the warm pool
leads to a westward extension of the Walker circulation, drying eastern Africa.”
Climate Dynamics 37.11–12 (2011): 2417–2435.
53
Funk, Chris. “We thought trouble was coming.” Nature 476.7358 (2011): 7.
54
Lyon, Bradfield, and David G. DeWitt. “A recent and abrupt decline in the East
African long rains.” Geophysical Research Letters 39.2 (2012).
55
Tierney, Jessica E., et al. “Multidecadal variability in East African hydroclimate
controlled by the Indian Ocean.” Nature 493.7432 (2013): 389–392.
56
Tierney, Jessica E., Caroline C. Ummenhofer, and Peter B. deMenocal. “Past and
future rainfall in the Horn of Africa.” Science Advances 1.9 (2015): e1500682.
57
Hoell, Andrew, and Chris Funk. “The ENSO-related west Pacific sea surface
temperature gradient.” Journal of Climate 26.23 (2013): 9545–9562.
58
Hoell, Andrew, and Chris Funk. “Indo-Pacific sea surface temperature influences
on failed consecutive rainy seasons over eastern Africa.” Climate Dynamics
43.5–6 (2014): 1645–1660.
59
Yang, Wenchang, et al. “The East African long rains in observations and
models.” Journal of Climate 27.19 (2014): 7185–7202.
316 / Appendix
Pacific Ocean. Two 2014 papers led by Shraddhanand Shukla60

demonstrated the predictability of the March–May rainy season,
helping set the stage for the successful 2016 Climate Hazard Center
forecasts. Two 2014 papers by Sharon Nicholson61 documented
the dangerous increase in recent droughts and explored the pre-
dictability of the March–May rainy season.
In 2015, David Rowell and coauthors published
Reconciling Past and Future Rainfall Trends over East
Africa,62 the paper that termed the paradox phrase, concluding
that natural variability “is unlikely to have been the dominant
driver of recent droughts.” In 2015, Andrew Hoell and
I explored the idea that a substantial proportion of the global
warming signal can be broken into two components: a high-
frequency component linked to stronger, warmer El Niños in
the equatorial East Pacific, and a low-frequency “warming
mode” linked to warming trends in the western Pacific and
Indian Oceans. In 2018, I led two efforts that (1) examined both
the 2015–2016 southern African droughts and the 2016–2017
east African droughts; and (2) formally attributed the influence
of climate change on the 2017 East African drought.63 Finally,
in their 2019 paper, “‘Eastern African Paradox’ rainfall decline
due to shorter not less intense Long Rains,”64 Caroline
60
Shukla, Shraddhanand, et al. “A seasonal agricultural drought forecast system
for food-insecure regions of East Africa.” Hydrology and Earth System Sciences
18.10 (2014): 3907–3921; Shukla, Shraddhanand, Christopher Funk, and
Andrew Hoell. “Using constructed analogs to improve the skill of National
Multi-Model Ensemble March–April–May precipitation forecasts in equatorial
East Africa.” Environmental Research Letters 9.9 (2014): 094009.
61
Nicholson, Sharon E. “A detailed look at the recent drought situation in the
Greater Horn of Africa.” Journal of Arid Environments 103 (2014): 71–79;
Nicholson, Sharon E. “The predictability of rainfall over the Greater Horn of
Africa. Part I: Prediction of seasonal rainfall.” Journal of Hydrometeorology 15.3
(2014): 1011–1027.
62
Rowell, David P., et al. “Reconciling past and future rainfall trends over East
Africa.” Journal of Climate 28.24 (2015): 9768–9788.
63
Funk, Chris, et al. “Examining the potential contributions of extreme “western
V” sea surface temperatures to the 2017 March–June East African drought.”
Bulletin of the American Meteorological Society 100.1 (2019): S55–S60.
64
Wainwright, Caroline M., et al. “‘Eastern African Paradox’ rainfall decline due
to shorter not less intense Long Rains.” NPJ Climate and Atmospheric Science
2.1 (2019): 1–9.
M. Wainwright and coauthors have suggested that the recent

decline is strongly associated with a shorter rainy season.
During March, warmer waters to the south of East Africa delay
the onset of the rainy season. During May, a decrease in surface
pressure over Arabia supports an earlier cessation of the
rainy season.
As this brief summary illustrates, substantial uncertainty
continues, and the paradox remains only partially resolved. But
we have come a long way.
Chapter 11
Readers interested in fire in the United States and else-
where might be interested in Michael Kodas’s compelling book
Megafire: The Race to Extinguish a Deadly Epidemic of Flame.
This book combines compelling narrative and science to
describe both recent conflagrations as well as the complex issues
surrounding these terrifying environmental disasters. In the
United States, the Union of Concerned Scientists has assembled
a good synopsis,65 which was updated in March 2020. This
page also links to a fascinating podcast by a fire expert,
Professor John Bailey.66 The Fourth US Climate Assessment67
covers fires in its chapter on forests, and in the individual
chapters on US climate regions.
Chapter 12
Boer, Resco de Dios, and Bradstock’s accessible
Unprecedented Burn Area of Australian Mega Forest Fires com-
mentary68 provided the main input for Chapter 12, augmented
65
The Connection Between Climate Change and Wildfires, Union of Concerned
Scientists, www.ucsusa.org/resources/climate-change-and-wildfires.
66
Bailey, John.The Science of Forest Fires: Culture, Climate, and Combustion,
www.ucsusa.org/resources/science-forest-fires-culture-climate-and-combustion.
67
The Fourth United States Climate Assessment, nca2018.globalchange.gov/.
68
Boer, Matthias M., Víctor Resco de Dios, and Ross A. Bradstock.
“Unprecedented burn area of Australian mega forest fires.” Nature Climate
Change (2020): 1–2.
318 / Appendix
by two papers by Rachel Nolan and coauthors that describe and

document the link between vapor pressure deficits, dead fuel
moisture levels, and fire extent. Dr. Nolan’s paper in Remote
Sensing of Environment69 describes the link between vapor
pressure deficits and dead fuel moisture. Dr. Nolan’s paper in
Geophysical Research Letters describes the link between dead
fuel moisture and wildfire extent.
This chapter has also drawn heavily on the Summer of
Crisis report, by Lesley Hughes, Will Steffen, Greg Mullins,
Annika Dean, Ella Weisbrot, and Martin Rice, which aptly
describes the 2019–2020 fire season for New South Wales.
Both this chapter and the Summer of Crisis report refer to the
excellent reports produced by Australia’s Bureau of
Meteorology. Their site www.bom.gov.au/climate/ provides
access to these reports, as well as more information about
Australia’s climate, the Indian Ocean Dipole (www.bom.gov
.au/climate/iod/) and El Niño-La Niña impacts (www.bom.gov
.au/climate/enso/).
Readers will also likely be interested in a detailed climate
attribution study by Geert von Oldenburg and coauthors.70
A summary of this excellent study can be found at the World
Weather Attribution website www.worldweatherattribution
.org/bushfires-in-australia-2019-2020/. The study focused on
changes in fire weather over the area of the most intense bush-
fires, and concluded that climate did, and will, increase the
probability of extreme fire weather. Readers are encouraged to
access the full study at www.nat-hazards-earth-syst-sci-discuss
.net/nhess-2020-69/.
69
Nolan, R. H., V. Resco de Dios, M. M. Boer, G. Caccamo, M. L. Goulden, and
R. A. Bradstock. “Predicting dead fine fuel moisture at regional scales using
vapor pressure deficit from MODIS and gridded weather data.” Remote Sensory
Environment 174 (2016): 100–108.
70
van Oldenborgh, G. J., F. Krikken, S. Lewis, N. J. Leach, F. Lehner, et al.
“Attribution of the Australian bushfire risk to anthropogenic climate change,”
Natural Hazards and Earth Systems: Scientific Discussions (2020, in review), doi.
org/10.5194/nhess-2020-69.
Chapter 13
Each year the Global Carbon Budget project puts out a
fabulous assessment of the previous year’s carbon emissions
(www.globalcarbonproject.org/carbonbudget/). Observations
of annual atmospheric CO2 levels provided by Scripps
(scrippsco2.ucsd.edu/) and NOAA (www.esrl.noaa.gov/gmd/
ccgg/trends/global.html) provide a key input. The annual carbon
budget is summarized in a yearly article in Earth System Science
Data. The 2019 paper71 is available at www.earth-syst-sci-data
.net/11/1783/2019/. This paper describes the estimation process
and major results. Another 2019 Global Carbon Project report,
“Global Energy Growth is Outpacing Decarbonization,”72 pro-
vides a concerning summary.
Chapter 14
This chapter draws on Yuval Noah Harari’s excellent
book Sapiens: A Brief History of Humankind.73 A guiding
principle of Sapiens is that shared stories have acted as a central
innovation in human history, fostering coordinated activities
that greatly leveraged our individual capabilities. Whether
teaming up to take down a mastodon or build a corporation,
stories provide a common framework that supports coherent
action. This book has sought to reinforce such a framework by
providing an accessible description of recent climate extremes,
alongside accounts of associated humanitarian and economic
impacts. As we face the dangers of climate change, we are also
empowered to watch our evolving planet and humanity in ways
that were never possible before. To this end, this chapter draws
on the incredible data resources provided by the World Bank
71
Friedlingstein, Pierre, et al. “Global carbon budget 2019.” Earth System Science
Data 11.4 (2019): 1783–1838.
72
www.globalcarbonproject.org/global/pdf/GCP_2019_Global%20energy%
20growth%20outpace%20decarbonization_UN%20Climate%20Summit_HR
.pdf.
73
Harari, Yuval Noah. Sapiens: A Brief History of Humankind. Random House,
2014.
320 / Appendix
(data.worldbank.org/indicator/). Another very valuable

resource used here is provided by the annual United Nations
Food and Agriculture Organization’s The State of Food Security
and Nutrition reports.74 Since 2015, this data source indicates
an increase in global hunger, related in part to increasingly
extreme climate. Finally, the chapter also builds on natural
catastrophe data summarized by AON Benfield.75 AON
Benfield, Munich Re,76 and other re-insurance companies spe-
cialize in helping insurance companies manage risk. They have a
profit-driven motivation to pay very close attention to
climate extremes.
74
www.fao.org/state-of-food-security-nutrition.
75
thoughtleadership.aonbenfield.com/Pages/Home.aspx?ReportYear=2020.
76
Munich Re Natural Catastrophe Database, natcatservice.munichre.com/.
INDEX
acute malnutrition, 189–190, 197 Camp Fire, 11, 240, 243, 245, 248
adenosine triphosphate, 39 Carnot Cycle, 43
anthropic principle, 21 Carr Fire, 243
arrow of time, 44 Centre for Research on the
Arthur Eddington, 44 Epidemiology of Disasters, 99,
attribute, 112 198
attribution, 149 China, 10, 99, 135–136, 152, 160,
attribution analysis, 211, 261–262, 190, 225, 280–281, 290
264–265 Clausius-Clapeyron, 127–128, 130
attribution science, 107 climate attribution, 13, 117, 153,
attribution study, 117–118, 232, 165, 204, 261, 263, 265, 301,
263, 318 311, 318
Australian Bureau of Meteorology, climate attribution, 1, 12
254–255, 312 climate change attribution, 149,
available energy, 44, 47, 53 165, 311
climate change projections, 136
Bangladesh, 5, 9, 135, 151, 273 Climate Explorer, 63, 84, 137
bathtub warming, 168 climate hazard, 165
Berkeley Earth, 101 Climate Hazards Center, 101, 123,
Big Bang, 23 131, 196, 214, 216
black hole, 26–27 coal, 5, 35, 38, 279–280, 283, 303
Black Summer, 267 Community Alert, 236
Bulletin of the American complexity, 16, 22, 39, 46–49, 148,
Meteorological Society, 116, 189
127, 148, 171, 198, 204, 263, Convectively Available Potential
273, 306, 311, 313–314, 316 Energy, 53, 55
322 / Index
coral bleaching, 9, 211, 273 204, 210, 214, 219, 226, 232,
coral reefs, 17, 80, 94, 114, 171, 314
272–273, 280, 283, 286, 300 event attribution, 12–13, 108, 153,
Coriolis, 34 306, 311
counterfactual, 112, 118–119, 265 exposure, 7–8, 17, 37, 92–93, 101,
crop production, 4, 299 103, 105–106, 109, 113–114,
cumulonimbus, 56 116, 265, 302, 309
cyanobacteria, 58 external variations, 167
DanChurchAid, 1 famine, 4, 10, 19, 86, 171, 189–191,

dark energy, 23 196, 210, 213, 221, 290–291
dead fuel moisture, 256–258, famines, 14, 32, 190, 193, 290, 313
263–266, 318 fingerprints, 5
dead fuel moisture content, 258 fire extent, 257
detection, 149 flooding, 10, 132, 164, 171
food prices, 196, 198, 213, 221
Earth–Sun system, 16, 110 Food Security and Nutrition
East Africa, 2, 9–10, 12, 18, 164, Working Group, 4, 123
171, 173, 176, 179–180, 213, France, 99–100, 300
215, 222, 228–229, 232–233,
315–316 galaxies, 26
East African Climate Paradox, 18, Galileo, 26, 37, 49, 183, 312
212–213 general circulation, 33
Egypt, 103, 191, 284, 302 Global Carbon Budget, 123, 277,
Einstein, 69, 73 279, 282, 285, 319
El Niño, 7, 9, 13–14, 18, 32, 77, Global Carbon Project, 68, 278,
186–187, 190, 192–195, 281–282, 291, 319
197–198, 201–202, 204, Goldilocks, 36
206–210, 212–213, 216, 222, Goldilocks Planet, 16, 36, 74, 276,
274, 286, 290, 301, 307, 312, 307–308
318 gravity, 23, 26
El Niño–Southern Oscillation, 32, Great Barrier Reef, 9
193 greenhouse gas effect, 29, 65
Emergency Events Database, 99, 136 gross domestic product, 152, 268,
energetically closed, 47 291, 296
energetically open, 47 Guinea, 103, 133
energy balance model, 71
Energy Convergence Model, 170 Hadley Circulation, 31, 34, 47, 53–54
energy intensity, 292 heat death, 45, 47, 50–51
entropy, 43–49, 51, 57 heat engine, 33, 45
epidemiology, 117, 153 heat wave, 99–100, 103, 105–106,
Ethiopia, 15, 18, 103, 164, 174, 113, 115, 117, 198
177, 189, 191, 195–196, 202, height fields, 222–223, 225
323 / Index
Heraclitus, 39, 67 226, 228, 232, 274, 301, 313,

holy moly, 175 318
human capacity, 295 latent heat, 56
Hurricane Harvey, 9, 17, 131, 145, longwave radiation, 48
150, 154, 310
Hurricane Maria, 11, 18 Madagascar, 133, 160
Mali, 103
increased energy efficiency, 303 Mars, 30, 74
India, 11, 18, 32, 99–101, 103, methane, 38, 66, 70, 270, 276, 303
116–118, 134–135, 151, 160, Mexico, 33, 187, 203, 284
190–192, 195–197, 202, 273, Milky Way, 27, 35, 37
281, 290, 313 moral consequences, 182
Indian Monsoon, 33 moral culpability, 113
Indian Ocean Dipole, 165, 171, 176, Mozambique, 188, 199
178, 182, 301, 312, 318 Munich Re, 7–9, 151, 160–161,
Indonesia, 15, 33, 53, 123, 133, 241, 310, 320
173, 194, 198, 229–230
integrated phase classification, 189 National Academies report, 148, 311
Intergovernmental Panel on Climate Natural Catastrophe Database, 7–8
Change, 74–75, 81–82, 110, nebulae, 26
268, 283, 300, 305, 309 negentropic systems, 57
internal variations, 168 negentropy, 45–46, 58
Isaac Newton, 26, 44, 48, 73, 185 Nepal, 135, 151, 273
new normal, 79, 100
Jacob Bjerknes, 193 Nicolas Léonard Sadi Carnot, 43
Japan, 11, 99, 135, 137, 152, 160 Niger, 103
Jet Streams, 31 Nigeria, 103, 135
John Snow, 108
Judea Pearl, 154 ozone, 28, 56, 66
Kangaroo Island, 252 Pakistan, 99–100, 103, 116–117,

keeling curve, 30 135, 198, 273, 303
Kenya, 2, 4–5, 15, 123, 164, Palmer Drought Severity Index,
173–174, 177, 214, 226, 232 238–239, 244
Kepler, 26 Papua New Guinea, 123
Kerala, 11, 134 pastoralists, 2, 127, 219
koalas, 252 Pearl Causality, 155
Koninkliijk Nederlands Peru, 33, 135, 193–194
Meteorologisch Instituut, 137 Phase 5 Coupled Model
Korea, 160, 191 Intercomparison Project, 84,
138, 227
La Niña, 13–14, 18, 176, 180, 187, phase transition, 144
191, 193, 212–216, 223–224, Philippines, 133, 135, 191
324 / Index
photosynthesis, 35, 38, 58, 125, South Africa, 188, 191, 201
235, 275 South Sudan, 103, 164, 197
population growth, 75, 105, 114, Southern Oscillation, 32, 190, 193,
116, 268–269, 271, 295, 301 307
positive feedback, 55, 144, 267, 276 Spaceship Earth, 38, 294
poverty, 1–2, 4, 11, 19–20, 133, star generation, 27
191, 296, 299 stomata, 35, 235–236, 275
prediction, 11, 15, 18, 63, 85, 107, storyline, 150, 153, 311
165, 169, 180, 185, 193, 211, stratosphere, 28
215, 233, 301 strong force, 26, 28
preindustrial simulations, 207 Sudan, 103, 191, 202, 302
pressure fields, 222
pressure waves, 27 Tahiti, 33, 193
Tasmania, 99
radiative balance, 47, 110 temperature inversion, 55
radiative forcing, 110–111, 270, thermal energy, 107
272 thermal infrared longwave
radiative transfer, 29, 72, 75 radiation, 51
relative humidity, 61, 128, 130, 234, thermodynamic control, 130
240, 256 thermodynamic efficiency, 43
Representative Concentration Thomas Fire, 60, 87, 91
Pathway, 104, 110, 261, 268, thought experiment, 128, 170
278 trade winds, 31, 33–34
ribonucleic acid, 57 transpiration, 14, 26
Ridiculously Resilient Ridge, 240 troposphere, 31, 33, 47, 56, 144,
Royal Society for the Prevention of 223
Cruelty to Animals, 252 Turkey, 103
Rudolf Julius Emanuel Clausius, 44 Type I errors, 149
Type II errors, 149
saturation vapor pressure, 256
sea surface temperature gradient, Uganda, 103, 164, 174, 177
170, 172, 179–180, 222, 315 United Nations Office for
sea surface temperature gradients, Disaster Risk Reduction, 133,
167–168, 175 310
shortwave radiation, 48
Sierra Leone, 10 Van Allen belts, 28, 38
Sir Clive Granger, 154 vapor pressure deficits, 256, 263,
Sir Gilbert Walker, 32, 193–194, 265–267, 318
313 Venus, 30, 74
societal change, 183 vertical velocity, 128, 130
Somalia, 2, 15, 164, 173–174, 177, Vietnam, 160, 191
213–215, 218–221, 225–226, vulnerability, 7–8, 92–93, 101,
232, 288 113
325 / Index
Walker Circulation, 32–33, 53–54, World Meteorological

194–195, 225, 228 Organization, 148, 311
Warm Pool, 33, 214
Western V, 226, 230–231 Yuval Harari, 287
wildfire extent, 19, 237–238, 266,
301 Zimbabwe, 15, 188, 199, 201

Drought, Flood, Fire, How Climate Change Contributes to Catastrophes

Uploaded by

Copyright:

Available Formats

Drought, Flood, Fire, How Climate Change Contributes to Catastrophes

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Drought, Flood, Fire, How Climate Change Contributes to Catastrophes

Uploaded by

Copyright:

Available Formats

Drought, Flood, Fire

Every year, droughts, ﬂoods, and ﬁres impact hundreds of millions

Chris Funk is an internationally renowned climate hazard scientist

One Liberty Plaza, 20th Floor, New York, ny 10006, USA

477 Williamstown Road, Port Melbourne, vic 3207, Australia

314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,

79 Anson Road, #06–04/06, Singapore 079906

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of

© Chris Funk 2021

This publication is in copyright. Subject to statutory exception

First published 2021

Printed in the United Kingdom by TJ Books Limited, Padstow Cornwall

Library of Congress Cataloging-in-Publication Data

isbn 978-1-108-83987-7 Hardback

Cambridge University Press has no responsibility for the persistence or accuracy

Chris Funk is an internationally renowned climate hazard scientist

One Liberty Plaza, 20th Floor, New York, ny 10006, USA

477 Williamstown Road, Port Melbourne, vic 3207, Australia

314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,

79 Anson Road, #06–04/06, Singapore 079906

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of

© Chris Funk 2021

This publication is in copyright. Subject to statutory exception

First published 2021

Printed in the United Kingdom by TJ Books Limited, Padstow Cornwall

Library of Congress Cataloging-in-Publication Data

isbn 978-1-108-83987-7 Hardback

Cambridge University Press has no responsibility for the persistence or accuracy

Chris Funk is an internationally renowned climate hazard scientist

One Liberty Plaza, 20th Floor, New York, ny 10006, USA

477 Williamstown Road, Port Melbourne, vic 3207, Australia

314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,

79 Anson Road, #06–04/06, Singapore 079906

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of

© Chris Funk 2021

This publication is in copyright. Subject to statutory exception

First published 2021

Printed in the United Kingdom by TJ Books Limited, Padstow Cornwall

Library of Congress Cataloging-in-Publication Data

isbn 978-1-108-83987-7 Hardback

Cambridge University Press has no responsibility for the persistence or accuracy

1 Climate Extremes, Climate Attribution, Extreme

2 Welcome to an Awesome Planet: A Series of Delicate

3 The Earth Is a Negentropic System, or “the Bright Side

4 Do-It-Yourself Climate Change Science 59

5 Temperature Extremes – Impacts and Attribution:

6 Precipitation Extremes: Observations

7 Hurricanes, Cyclones, and Typhoons 140

8 Conceptual Models of Climate Change and Prediction,

9 Climate Change Made the 2015–2016 El Niño

10 Bigger La Niñas and the East African

11 Fire and Drought in the Western United States 234

12 Fire and Australia’s Black Summer 252

13 Driving toward +4C on a Dixie® Cup Planet 268

14 We Can Afford to Wear a White Hat 287

Appendix A Few Resources for Further Reading

First and foremost, I would like to thank the “ﬁrst

13 Driving toward +4C on a Dixie® Cup Planet 268