Noaa 52721 DS1

Article
Advancing Weather and Climate Forecasting

for Our Changing World
Gilbert Brunet, David B. Parsons, Dimitar Ivanov, Boram Lee, Peter Bauer,
Natacha B. Bernier, Veronique Bouchet, Andy Brown, Antonio Busalacchi,
Georgina Campbell Flatter, Rei Goffer, Paul Davies, Beth Ebert, Karl Gutbrod,
Songyou Hong, P. K. Kenabatho, Hans-Joachim Koppert, David Lesolle,
Amanda H. Lynch, Jean-François Mahfouf, Laban Ogallo, Tim Palmer, Kevin Petty,
Dennis Schulze, Theodore G. Shepherd, Thomas F. Stocker, Alan Thorpe, and Rucong Yu
ABSTRACT: Our world is rapidly changing. Societies are facing an increase in the frequency and
intensity of high-impact and extreme weather and climate events. These extremes together with
exponential population growth and demographic shifts (e.g., urbanization, increase in coastal
populations) are increasing the detrimental societal and economic impact of hazardous weather
and climate events. Urbanization and our changing global economy have also increased the need
for accurate projections of climate change and improved predictions of disruptive and potentially
beneficial weather events on kilometer scales. Technological innovations are also leading to an
evolving and growing role of the private sector in the weather and climate enterprise. This article
discusses the challenges faced in accelerating advances in weather and climate forecasting and
proposes a vision for key actions needed across the private, public, and academic sectors. Actions
span (i) utilizing the new observational and computing ecosystems; (ii) strategies to advance Earth
system models; (iii) ways to benefit from the growing role of artificial intelligence; (iv) practices
to improve the communication of forecast information and decision support in our age of internet
and social media; and (v) addressing the need to reduce the relatively large, detrimental impacts of
weather and climate on all nations and especially on low-income nations. These actions will be based
on a model of improved cooperation between the public, private, and academic sectors. This article
represents a concise summary of the white paper on the Future of Weather and Climate Forecasting
(2021) put together by the World Meteorological Organizations’ Open Consultative Platform.
KEYWORDS: Climate prediction; Nowcasting; Operational forecasting; Seasonal forecasting;

Numerical weather prediction/forecasting; Artificial intelligence
https://doi.org/10.1175/BAMS-D-21-0262.1
Corresponding author: David B. Parsons, [email protected]
In final form 28 June 2022
©2023 American Meteorological Society
For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy.
AMERICAN METEOROLOGICAL SOCIETY

Brought to you by NOAA Central Library | Unauthenticated |A PDownloaded E909 01:58 PM UTC
R I L 2 0 2 3 08/02/23
AFFILIATIONS: Brunet and Ebert— Bureau of Meteorology, Melbourne, Victoria, Australia;
Parsons—School of Meteorology, University of Oklahoma, Norman, Oklahoma; Ivanov—World
Meteorological Organization, Geneva, Switzerland; Lee—Global Facility for Disaster Risk Reduction,
World Bank, Washington, D.C.; Bauer and Brown—European Centre for Medium-Range Weather
Forecasts, Reading, United Kingdom; Bernier and Bouchet—Environment and Climate Change Canada,
Dorval, Quebec, Canada; Busalacchi—University Corporation for Atmospheric Research, Boulder,
Colorado; Campbell Flatter and Goffer—TomorrowNow.Org, Boston, Massachusetts; Davies—Met
Office, Exeter, United Kingdom; Gutbrod—Meteoblue, Basel, Switzerland; Hong—NOAA/ESRL/PSL, and
Cooperative Institute for Research in Environmental Sciences, University of Colorado Boulder, Boulder,
Colorado; Kenabatho and Lesolle—Department of Environmental Science, University of Botswana,
Gaborone, Botswana; Koppert—Deutscher Wetterdienst, Offenbach, Germany; Lynch—Brown
University, Providence, Rhode Island; Mahfouf—Université de Toulouse, Météo-France, CNRS, Toulouse,
France; Ogallo*—University of Nairobi, Nairobi, Kenya; Palmer—University of Oxford, Oxford, United
Kingdom; Petty—Spire Global, Boulder, Colorado; Schulze—MeteoIQ, Berlin, Germany; Shepherd and
Thorpe—Department of Meteorology, University of Reading, Reading, United Kingdom; Stocker—Physics
Institute, and Oeschger Centre for Climate Change Research, University of Bern, Bern, Switzerland;
Yu—China Meteorological Administration, Beijing, China
* Deceased
O
ur changing world dictates the need for substantial forecast improvements in numerical
Earth system and weather-to-climate prediction (NEWP). One challenge facing society
as evidenced in daily news headlines is the increase in extreme weather events with, for
example, in 2021, record-breaking cold in the U.S. Southern Great Plains, heat in the Pacific
Northwest followed by wildfires, devastating flooding in Europe, Canada, Australia, and China,
and rain-induced landslides in India. A statistical analysis from Munich Re’s NatCatService
2019 database reveals that 1,383 weather-related natural disasters took place across the globe
from 1980 to 1985 and the number grew to 4,020 registered events from 2013 to 2018
(Fig. 1). Human influence in extreme events was already shown in the Fifth Assessment
Report of the Intergovernmental Panel on Climate Change (AR5; IPCC 2013). Subsequently,
the AR6 report concludes that a warming climate plays a role in increasing extremes in heat,
drought, rainfall, wildfire, tropical cyclone intensity, and flooding (IPCC 2021). The need for
improved weather and climate forecasts is also driven by this increase in extremes occurring
against a backdrop of nonclimatic global trends that include biodiversity loss, unsustainable
consumption of natural resources, degradation of ecosystems, human demographic shifts
including rapid urbanization, social and economic inequalities, and a pandemic (IPCC 2022).
For example, the dramatic loss of Arctic sea ice associated with a warming climate has also
led to damaging shoreline erosion, changes in shipping routes, threats to national security,
and damage to the Arctic ecosystem.
Currently, 19% of the world’s population (1.47 billion people) face substantial risk of a
1-in-100-yr flooding event (Rentschler and Salhab 2020). While extremes in the climate
system have a devastating effect on populations across the
world, the impacts often fall disproportionately on low-income 1
The dollar amounts are defined in the cited
nations and marginalized populations as ~587 million people publications, which span from 2017 to 2020.
who face substantial risks from flooding are living on less Given this relatively short time period, we did not
attempt to standardize the amounts.
than $5.51 (U.S. dollars) per day (Rentschler and Salhab 2020).

R I L 2 0 2 3 08/02/23
Fig. 1. Major weather- and climate-related loss events from the Munich Re, NatCatService, 2019
database plotted utilizing their online data access tools. For the periods (a) 1980–85 and (b) 2013–18.
From keynote presentation by D. Parsons at InterMet Asia 2019.

R I L 2 0 2 3 08/02/23
Natural disasters also force a $520 billion (U.S. dollars) drop in consumption and
drive ~26 million people into poverty each year (Hallegatte et al. 2017).
The importance of weather and climate risk–based decision-making to society is also
increasing substantially with population growth and demographic shifts that include urban-
ization and movement to coastlines. For example, Majumdar et al. (2021) outlines the growing
societal need for accurate and timely forecasts of flooding, heatwaves, and air pollution events
in urban areas. The trends in urbanization are dramatic as the year 1800 only about 2%
of the world’s population lived in urban areas compared to ~55% in 2018 and projections of
68% in 2050 with 2.25 billion more urban residents added in Asia and Africa alone (United
Nations 2019). These trends mean more people, assets, and essential services (e.g., power, water,
transport, telecommunications, the internet) are exposed and vulnerable to weather, climate,
water, ocean, and even space hazards. The demographic trends have resulted in a growth in the
catastrophic events along coasts (Fig. 1) implying a clear need for highly localized, accurate,
and frequently updated information, as well as tailored services for informed decision-making
over multiple time scales. Meeting these needs will require a new level of collaboration between
the stakeholders and user communities across the weather and climate enterprise.
Another driving force for the NEWP enterprise results from our global “just in time” manu-
facturing process that increasingly requires accurate, timely, and high-resolution forecasts
of both detrimental events and routine day-to-day variations. These improved forecasts are
needed since the components used in developing the final product are created and delivered
as needed rather than created, shipped, and stored well in advance. A survey of 1,000 global
executives in 15 countries by the IBM Institute for Business Value (IBV) in cooperation with
Oxford Economics found that 93% of the executives felt that improved insight into weather
variations would positively impact their revenue with estimates of growth in annual revenue
exceeding $500 billion (U.S. dollars) (Orrell et al. 2018). This large economic impact is con-
sistent with the study by Lazo et al. (2011) that found 3.4% of the U.S. gross domestic product
is affected by routine variations in day-to-day weather.
These economic drivers have resulted in a growth in private sector companies providing
targeted forecasts of weather variations from the time scale of minutes to climate risk
and resiliency information spanning decades into the future. Policy debates around the
future of the planet and society are intense in a world with significant global technological
transformations and growing environmental risks. Such debates shape high demand and
need for better weather and climate information and for services addressing the risks and
socioeconomic impacts of the weather, climate, and water hazards.
Another factor in the debate on future directions of the NEWP enterprise is that we are living
in the “decade of digital transformation” with increased reliance on data and information
technology (IT). This reliance will profoundly change the organizations across the NEWP
enterprise. These data opportunities coupled with the previously mentioned factors will
drive increased competition between the public sectors and the growing private sector in
providing specialized, targeted forecasts. However, investigation of this issue across eight
diverse nations (World Bank 2020) suggests that defined cooperation between the public,
private, and academic sectors will help meet the growing demands and need for forecasts,
help sustain national meteorological and hydrological services (NMHSs), and foster economic
development. Historical background on the U.S. prospective on these partnerships can be
found in National Research Council (2003) report.
Objective and scope

The main objective of this paper is to provide a basis for informed decision-making by weather
and climate enterprise stakeholders in planning their activities and investments in the NEWP
enterprise during the coming decade. The weather and climate forecast enterprise involves

R I L 2 0 2 3 08/02/23
various stakeholders from public, private, and academic sectors as shown in Fig. 2. The
important drivers of the innovation cycle are highly qualified personnel, computational and
observational infrastructures (in the middle of the figure), and increasing stakeholder and cus-
tomer demand and need (on the circumference of the figure) for tailored and seamless weather
and climate forecasting (localized, timely, precise, and accurate). Stakeholders and customers
can push clockwise new initiatives (Fig. 2) at different positions in the innovation cycle: R&D,
operation, and services. This paper will cover these three components of the innovation cycle:
infrastructure, R&D, and operation. Stakeholders engaged in all three components will have to
make strategic choices in the coming years, and some will struggle to keep up as technologies
continue to combine and advance, and new ways of doing business appear quickly.
This paper summarizes the views, knowledge, and expertise of a group of scientists and
practitioners from the public, private, and academic sectors that was produced under the
leadership of WMO (Brunet et al. 2021). Our summary represents an update of the vision
described over a decade ago in Shapiro et al. (2010). In addition, our approach of including
input across these sectors allows us to go beyond the scientific vision described in Shapiro
et al. (2010) through increased knowledge of the needs of society and economy. We do not
attempt to dictate solutions to the many open questions on the future of weather and climate
forecasting. Instead, we strive to improve the understanding of the ongoing R&D and to iden-
tify technological trends, and any possible impediments to progress. In this way, risks and
Fig. 2. The innovation cycle: the public–private–academic engagement challenge taken from Brunet
et al. (2021). The operations component is the creation of forecast and other products, while service
is the delivery of these products to stakeholders and customers. Note that each sector can make
contributions to the various components (i.e., research and development, operations, and service).

R I L 2 0 2 3 08/02/23
opportunities for each stakeholder can be better assessed, and decisions on future organiza-
tional plans and investments can be better informed. Although our focus is on the future of
weather and climate forecasting, we note that many of our recommendations are consistent
with the priorities and targets of the Sendai Framework for Risk Reduction 2015–30 (UNISDR
2015) that had a greater emphasis on public health.
Weather and climate forecasting: Setting the scene

Brief history of NEWP. Significant progress has been achieved in NEWP over the past six
decades through collaborative efforts by many institutions across the public, private, and aca-
demic sectors at national and international levels. There are several papers on the history of the
highlights of NEWP developments (e.g., Pudykiewicz and Brunet 2008; Magnusson and Källén
2013; Bauer et al. 2015; Parsons et al. 2017; Benjamin et al. 2019; Bock et al. 2020; Ritchie
et al. 2022) with our summary of highlights provided in the “Major milestones in operational
weather and climate forecasting” sidebar. While weather prediction systems were initially de-
signed to provide useable information on time scales covering hours to days, NEWP forecasts
Major milestones in operational weather and climate forecasting

• 1861: Met Office weather forecast services using telegraphs established by FitzRoy
• 1873: Working toward global meteorological observatories and international data sharing with the
foundation of the International Meteorological Organization in Vienna
• 1900–22: Birth of numerical weather prediction (NWP) with the work of Abbe (1901), Bjerknes (1904),
and Richardson (1922)
• Early 1920s: Onset of statistical climate prediction and global atmospheric teleconnection insights
pioneered by Walker
• 1950: First computer NWP forecast on Electronic Numerical Integrator and Computer (ENIAC) by Charney
et al. (1950)
• 1960 onward: Satellite-based meteorological observations and telecommunications at the forefront of
technological innovations since the launch of TIROS-1 (first weather satellite)
• 1960s onward: Emergence of general circulation models for climate research and forecasting
• 1962: Establishment of the World Weather Watch (WWW) program with its three main components (Global
Observing System, Global Telecommunication System, and Global Data-Processing System)
• 1963: Lorenz’s seminal work on chaos initiated atmospheric predictability theory and paved the way to
numerical ensemble prediction in the 1980 and 1990s
• 1969: Launch of the Global Atmospheric Research Program (GARP) led by Charney
• 1969 onward: Global NWP innovations since the first global spectral NWP simulation by Robert
• 1975: Federation of global NWP R&D effort in Europe with the foundation of the European Centre for
Medium-Range Weather Forecasts (ECMWF)
• 1979: First GARP Global Experiment, to gather the most detailed observations ever of the global atmosphere
• 1980: World Climate Research Programme (WCRP) established under the joint sponsorship of the
World Meteorological Organization (WMO) and the International Council for Science (ICSU) currently the
International Science Council
• 1980s onward: Development of coupled ocean–atmosphere climate models
• 1992: Operational implementation of ensemble prediction systems at ECMWF and the National Centers
for Environmental Prediction (NCEP)
• 1993: The Intergovernmental Oceanographic Commission of UNESCO became a cosponsor of WCRP
• 1995–97: Global reanalyses developed by NASA, NOAA–NCAR, and ECMWF fostering the use of NWP
analyses in climate monitoring
• 1997: Ground-breaking numerical prediction advances in the use of multiple sources of Earth system
observations with the introduction at ECMWF of four-dimensional data assimilation
• 1998: World Weather Research Programme (WWRP) established by the WMO
• 2002: Earth Simulator, Japan—A landmark supercomputer investment for climate, weather, and
geophysical research
• 2002: Met Office began use of high-resolution, convective allowing model for operational prediction

R I L 2 0 2 3 08/02/23
• 2004: WMO establishes the 10-yr THORPEX program to accelerate improvements in the skill in global
NWP models
• 2005: First operational assimilation of cloud and precipitation affected satellite data at ECMWF
• 2007: A great step forward for weather and climate Earth system forecasting with 3,000 Argo oceanic
floats in global operation
• 2008 onward: Dealing with prediction uncertainty in data assimilation with ensemble data assimilation
techniques
• 2008: First successful launch and application of CubeSat technology by University of Toronto for
atmospheric profiling of column integrated CO2, atmospheric humidity, and electron density
• 2013: First global simulations with subkilometer-scale horizontal resolution conducted using the
Nonhydrostatic Icosahedral Atmospheric Model on the K supercomputer
• 2013: Cloud computing first used for regional NWP applied to user needs
• 2014: Agreement reached in the ECMWF system between the annual average of the ensemble spread
and error in increasing the reliability of ensemble forecasts
• 2015: Development of a 40-member ensemble for the atmospheric component of the Community Earth
System model
• 2018 onward: Introduction of coupled ocean/sea ice models to global NWP systems (e.g., leading to
improved medium-range prediction of Arctic cyclones)
• 2018 onward: Acceleration in the development of convolution neural networks and other machine
learning approaches for accelerating computing, advancing model physics, and weather and climate forecasts
• 2020 onward: Joint weather and climate community initiatives toward global convection resolving models
through the use of exascale computing
now extend from minutes and hours to seasons, years, and even decades on spatial scales rang-
ing from urban to planetary for different components of the Earth system. The achievements
are remarkable; for instance, the ability of NEWP models to predict variations in the 500 hPa
height field generally improves at a rate of a day per decade (e.g., Bauer et al. 2015) so that the
midlatitude 5-day weather forecasts today are as accurate as 1-day forecasts 40 years ago. The
advances in Earth system modeling are another outstanding achievement that encompasses
understanding and predicting the atmosphere and its chemical composition, the oceans, land/
sea ice, and other cryosphere components as well as the land surface, including surface hydrol-
ogy and wetlands, lakes, and human activities. On short time scales, Earth system forecasting
needs to include phenomena that result from the interaction between one or more components
(e.g., ocean waves, tides, sea ice, and storm surges). On longer time scales, the Earth system also
includes terrestrial and ocean ecosystems, the carbon and nitrogen cycles and slowly varying
solar and cryosphere components (e.g., large continental ice sheets and permafrost).
These advances in operational weather forecasting and climate predictions began well
before computers were utilized to produce forecasts (see “Major milestones in operational
weather and climate forecasting” sidebar). Subsequently the success of the first numerical
prediction by Charney et al. (1950) launched a spectacular trend of innovations in NEWP
over the following seven decades. Routine, real-time forecasting with NEWP started in the
mid-1950s and was introduced in operation in the 1960s. Improved observational coverage,
the advent of satellite observations, the steady growth of computer power, and breakthroughs
in the theory of Earth system coupled processes all underpinned a successful story of weather
forecasting in the NEWP era.
The high cost of NEWP, including the capital investment for computers and their running
and maintenance costs, as well as resources needed in R&D, meant that the most developed
nations had the highest concentration of major developments. Nonetheless, exemplary coop-
eration and knowledge sharing with scientists from many countries and institutes has nur-
tured advancement of NEWP. European countries undertook a strong collaborative move with
the establishment of the European Centre for Medium-Range Weather Forecasts (ECMWF)

R I L 2 0 2 3 08/02/23
in 1975 as an intergovernmental organization. International cooperation has also been
supported by our professional societies (e.g., American Meteorological Society, Royal
Meteorological Society, the American and European Geophysical Unions) and government
agencies funding field campaigns and focused research efforts.
Progress in NEWP is often illustrated by the improvement in the horizontal and vertical
resolution of operational models even in global modeling systems (e.g., Skamarock et al.
2012; Miyamoto et al. 2013; Stevens et al. 2019). There has been an almost 40-times increase
in the horizontal resolution of global numerical weather prediction (NWP) models (from
about 400 km in the early 1960s to less than 10 km in 2020); in addition, regional fine-mesh
models have reached 1 km in resolution. In the vertical direction, from the early one- and
three-layered quasigeostrophic models, today’s models utilize more than 130 levels, reaching
an altitude of about 80 km. The role of the private sector in NEWP has evolved significantly
from their initial role as a provider of infrastructure (observation instruments, telecom and
supercomputer ecosystems) and primarily producing value-added products based in part on
NMHS forecast data. In many countries, the private sector plays a significant, if not leading,
role in communicating and interpreting weather information to the public and providing
specific weather forecasts tailored for use by governments and businesses. The private sector
activities have also grown to include providing global satellite-based measurements, targeted
high-resolution modeling, and climate risk products utilized by corporations and governments.
These climate risk products are obtained by combining historical observations of extremes
with the output of ensemble-based climate models from research institutions, downscaling
with high-resolution models, and artificial intelligence approaches.
WMO role. Obtaining the observations to initialize NEWP models requires data sharing
across national boundaries. The WMO has played a critical role in fostering data sharing
and furthering the progress of weather and climate forecasting. In cooperation with partners
in the 1960s, the WMO established the World Weather Watch (WWW) program composed
of the Global Observing System, the Global Telecommunication System, and the Global
Data-Processing and Forecasting System (GDPFS), coupled with the Meteorological Applica-
tions Programme. The result was a global set of observational and forecast data that were
shared among WMO member states and territories serving as the foundation for develop-
ment of the spectrum of user-oriented applications and services.
The global observing system is changing, however, to include observations from CubeSat
satellite systems, drones, vehicle-borne systems, microwave links, airborne Mode-S systems,
social media hazard data, low-cost 3D printed weather stations, urban in situ and enhanced
air quality systems, and other systems (e.g., Majumdar et al. 2021). To meet the growing
demand and need for observations useful to NEWP within this changing observing sys-
tem, the WMO has launched a new Unified Data Policy (https://public.wmo.int/en/our-mandate/
what-we-do/observations/Unified-WMO-Data-Policy-Resolution). This new policy encompasses
WMO-relevant Earth system observations (e.g., weather, climate, hydrology, atmospheric
composition, cryosphere, and space weather) with urging for a clear commitment to free
and unrestricted exchange of relevant data. Despite the movement toward data sharing,
open access to regional measurements (e.g., mesonets, radar data), research datasets, and
archives of observations taken by the private sector often remains a challenge. Such limita-
tions can hamper both scientific research and efforts to improve NEWP forecasts.
Today, the WMO GDPFS includes an elaborate system of global and regional cen-
ters, including nine World Meteorological Centres (WMCs) and 11 Regional Specialized
Meteorological Centres (RSMCs) with geographical specialization (Fig. 3). Various centers are
tasked with production of analyses and other data assimilation products, global determinis-
tic and ensemble NWP; limited-area deterministic and ensemble NWP; nowcasting; various

R I L 2 0 2 3 08/02/23
Fig. 3. WMO-designated (a) GDPFS centers for nowcasting and weather forecasting (up to 30 days) and
(b) long-range and climate forecasting (over 30 days) (WMO 2019a,b,c).
specialized forecasting activities (e.g., tropical cyclones); atmospheric transport and disper-
sion modeling (nuclear and nonnuclear); atmospheric sandstorm and dust storm forecasting;
numerical ocean wave prediction; aviation forecasting; and so forth. In addition, 13 centers are
designated as Global Producing Centres for Long-Range Prediction (monthly to seasonal),and
four centers as Global Producing Centres for Annual to Decadal Climate Prediction.
The WMO technical commissions were instrumental in facilitating international collabo-
ration and knowledge sharing. The World Weather Research Programme (WWRP) and the

R I L 2 0 2 3 08/02/23
World Climate Research Programme (WCRP)2 have been at the
2
The WCRP includes cosponsorship of by the
forefront of scientific efforts underpinning progress in NEWP WMO, the International Science Council (ISC),
development and in research-to-operation transition. The WWRP and Intergovernmental Oceanographic Commis-
and WCRP contributions span nowcasting (e.g., Majumdar et al. sion (IOC) of UNESCO (IOC-UNESCO), while the
WWRP is an international program of the WMO.
2021), global NWP (e.g., Parsons et al. 2017), polar prediction
(e.g., Rabier et al. 2010), and climate projections (e.g., Brasseur and Carlson 2015; Wake
2019) including the Coupled Model Intercomparison Project (CMIP) (Meehl et al. 2000;
Eyring et al. 2016; Bock et al. 2020). WWRP and WCRP collaboration has also been valuable
particularly on subseasonal to seasonal time scales at the weather–climate interface
(e.g., Brunet et al. 2010; Vitart 2017).
Challenges and opportunities in the coming decade: Recommendations

Operational weather forecasting based on numerical prediction systems has continuously im-
proved over the past few decades. The usefulness of NEWP forecasts has been pushed to lead
times beyond 10 days for some weather phenomena including for high-impact events such as
midlatitude snowstorms in North America. However, the steady progress has been at a slower
pace for some forecasted elements, like quantitative precipitation, where more efforts are needed
that span from observational needs and model improvements to communication of forecasts
(e.g., Balmaseda et al. 2020; NOAA 2020). This section highlights some key messages and rec-
ommendations that were discussed in Brunet et al. (2021), although we note that the helpful
comments of the reviewers and additional discussions between the authors have provided us with
some new perspectives. As such guidance must account for uncertainties and society’s ability to
use weather and climate forecasts, we emphasize optional ways of development and additional
considerations (e.g., economic, operational) that may need to be applied in decision-making.
Policy changes. In our changing world, the ever-growing need for weather and climate services
has meant the value of information is becoming better understood and appreciated (e.g., Orrell
et al. 2018). The market growth for NEWP is stimulating a variety of new business m odels
that may fit diverse national and transnational frameworks, improve the quality and access
to information, and contribute to the robustness and efficiency of the requisite underpinning
activities like environmental monitoring and research. Improved coordination and collabora-
tion among the public and private sectors of the weather and climate enterprise, with strong
engagement of academia, will accelerate the achievement of the vision for improving skill over
the coming decade. While technological advances and open data policies have enhanced col-
laboration between these sectors, improved coordination and collaboration remains a funda-
mental need with the degree of cooperation varying substantially across different countries
(World Bank 2020). The expectation is that each sector will continue in its most relevant roles;
at the same time, new potential synergies need to be effectively promoted and implemented.
On the policy side, prevailing free and open data sharing among countries and institu-
tions led by the WMO has facilitated progress including the establishment of the Global
Basic Observing Network (GBON) in which the basic surface-based observing network is
designed, defined, and monitored at the global level for international exchange. The in-
creased availability and adoption of forecast-driven tools for weather- and climate-informed
decision-making, especially by the commercial sector, has facilitated major progress. The need
and demand for decision-support tools in our “just in time economy” is growing rapidly. These
needed decision-support tools are a challenge as the transport of materials for manufacturing
often requires both accurate high-resolution, short-range forecast for components arriving by
trucks and rail to long-range forecasts for overseas shipping. However, in some areas, policies
implying commercial or other barriers to accessing important datasets need to be removed to
accelerate progress for detailed forecasts across these time scales. In addition to commercial

R I L 2 0 2 3 08/02/23
restrictions, regional observations (e.g., mesonets, radar) are often not shared widely. This
challenge of balancing open data versus protecting commercial licensing has been with the
weather enterprise for many years (e.g., National Research Council 2003).
Research and development. The general framework for R&D in NEWP has long been to
better utilize NEWP forecast data to meet society’s needs while striving to reach the theo-
retical limits of predictability (i.e., Lorenz 1963). Defining this theoretical limit remains a
challenge since the Earth system contains both slowly varying components (i.e., sea surface
temperature, soil moisture, snow cover, sea ice) that can aid predictive skill (Krishnamurthy
2019) and highly nonlinear, rapidly changing phenomena such as convection that can pro-
duce both large local errors and significant errors downstream on the jet stream due to the
modification of Rossby wave packets (e.g., Rodwell et al. 2018). Efforts to advance predictive
skill include increasing high-performance computing (HPC) capacity; improving observa-
tional instrumentation providing more accurate data with higher temporal and spatial reso-
lutions; advanced dynamical cores and better representation of complex physical processes;
better model initialization through the utilization of expanding satellite observations (e.g.,
both from government agencies and occultation and other observations from commercial
vendors), taking advantage of novel new measurement strategies such as crowd-sourced
data (e.g., Hintz et al. 2019), rainfall from cell phone tower attenuation (e.g., Overeem et al.
2013), and the improved utilization of polarimetric radar measurements as models move to
kilometer scale (Zhang et al. 2019), and more effective data assimilation methods; and use
of ensembles to represent uncertainty in the initial state and model processes.
The Brunet et al. (2021) report describes the R&D challenges that need to be met to ac-
celerate weather and climate forecasting in response to the needs of our changing world.
These challenges include the need to advance predictability, dynamical and physical model-
ing (see “Some predictability, dynamical, and physical modeling challenges” sidebar) and
Some predictability, dynamical, and physical modeling challenges

Significant outstanding challenges remain in terms of weather and climate predictive skill:
• Midlatitude weather regime transitions at the week 3 to week 4 time range
• Teleconnections between tropical phenomena (e.g., the Madden–Julian oscillation) and midlatitude
synoptic-scale weather patterns at the monthly range
• Seasonal-scale sea ice cover and thickness predictions, and the different trends observed in the Arctic
and Antarctic
• Large-scale patterns of precipitation and midlatitude circulation changes in climate change projections
• Deep-ocean circulation and ice sheet sensitivity to change in climate models
• Convective precipitation beyond the short range
• Closure of the water and energy cycle
Model and data assimilation shortcomings include the following:
• Major uncertainties associated with parameterizations of key processes such as deep convection,
orographic drag, cloud microphysics (and aerosol interactions), thermodynamic and dynamic coupling with
Earth surfaces, including land, waves, snow, and sea ice, and their impact on global circulation
• Limited understanding of the possible feedbacks of atmospheric composition on NEWP predictive skill
and climate simulations
• Mismatching fluxes at the land–atmosphere, ocean–atmosphere, and ocean–sea ice–atmosphere
interfaces
• Insufficient and incomplete treatment of systematic and random errors in data assimilation and ensembles
• Insufficient use of diagnostic methods for tracing key sources of model error back to their roots in weather
and climate models
Source: WMO (2015)

R I L 2 0 2 3 08/02/23
catalyzing innovations in numerical prediction system design like artificial intelligence (AI)
and machine learning (ML) techniques to help advance observing and prediction strategies
with topics of current and future interest shown in the “Potential major breakthroughs from
ML” sidebar. The reader is also referred to summary articles on the current and future direc-
tions of relevant ML and AI research (e.g., McGovern et al. 2017; Wang et al. 2019; Mansfield
et al. 2020; NOAA 2021). Since these approaches could prove valuable to forecasting on time
scales from nowcasting to climate resiliency with benefits to both the public and private
sectors, collaboration across the private, public, and academic sectors would prove benefi-
cial. Addressing these challenges at the national level will require NMHSs to engage more
in community-based modeling, data initiatives, and the R&D consortia. The importance of
working closely with users and the opportunities for private–public partnerships should be
recognized and promoted. Confederating R&D activities and intercountry collaborations will
help scale core NMHS technology and capability development. The WMO has been effective
in involvement in broad collaborative R&D topics relevant to advancing NEWP (i.e., Future
Earth and Integrated Research on Disaster Risk, projects of the International Science Council,
Risk Kan) and could play a major role in bringing NEWP advances to assist in addressing
the sustainable development issues, disaster impacts–responses, and other relevant topics.
Reaching these goals must also include addressing the funding challenges often im-
peding R&D progress relevant to advancing weather and climate forecasting including
(i) developing and refining strategies to advance interdisciplinary collaboration relevant
to NEWP, (ii) sustained funding strategies (i.e., >5 years) and supporting research that lies
between fundamental process studies and efforts to transition applied research to operation,
Potential major breakthroughs from ML

The use of ML will lead to significant progress over the whole range of “classical” modeling, data assimilation
and postprocessing algorithms in the coming decade. It can be applied at each step of the weather forecast-
ing process. Examples of topics relevant to NEWP can be found in the ESA–ECMWF Workshop: Machine
Learning for Earth System Observation and Prediction (www.ml4esop.esa.int). Current and future research
topics and potential advantages of utilizing ML include the following:
• Improve the utilization of observing systems, allowing forward planning of maintenance activities, improved
quality control, and intelligent filling of data gaps
• Advance the ability to input information from complex observing systems into models’ data assimila-
tion by providing better compression rates, which will significantly increase the precision and speed of
forward operators
• Help better estimate model and observation errors in data assimilation through combining hybrid ML and
data assimilation techniques
• Allow incorporation of further complex subgrid physical parameterizations, or training subgrid parameter-
izations on data provided by higher-resolution models like large-eddy simulation models
• Advance postprocessing of numerical variables by allowing incorporation of nonlinear correlations
• Estimate impacts by combining ML, model output, and further observations or statistically relevant data
• Advance nowcasting and improve rainfall estimation by applying ML to radar observations
• Use of ML approaches to detect and forecast severe weather such as tornadoes, large hail, and damaging winds
• Detect regimes where forecast failures are more likely to occur by combining ML with observations, model
initial conditions, and forecast output
• Explore whether ML is a computationally efficient tool to replace the NWP forecast approach
A wide range of open-source AI and ML frameworks (e.g., TensorFlow and PyTorch), together with well-known
programming languages like Python or C++, can support the development of these applications. Therefore, it
becomes possible to tackle large-scale problems with the advent of dedicated accelerators. Cloud infrastruc-
tures provide easy access to accelerator hardware. If meteorological data are readily available, small- and
medium-sized enterprises can develop successful applications, especially if they bring together expertise
from different disciplines.

R I L 2 0 2 3 08/02/23
(iii) fostering collaborations between weather and climate researchers to advance forecasting
at weather–climate interface and other problems of mutual benefit (e.g., physical parameter-
izations), (iv) supporting international R&D collaborations considering that the NEWP is a
global enterprise, and (v) developing efforts aimed at improving the utility of weather and
climate data by society.
Toward improved systems for forecasting

Investment strategies. Improved forecasts and better utilization of forecast information
will require investments in the global observing system, especially for sparsely populated
parts of the world and oceans where the boundary layer and lower troposphere are often
poorly sampled. Public investments in an accessible observational network will benefit all
enterprise stakeholders, including the private sector and academia. Investments will also
be needed in the supercomputing capability that is necessary to create weather and climate
information of high accuracy, resolution, and reliability to address growing societal needs.
These investment strategies must recognize the critical role of the private sector in creat-
ing the instrumentation for observational networks and the computational infrastructure.
Hence, these investments should be accompanied by collaborations between the weather
and climate enterprise, computer scientists, and industry on what changes need to be under-
taken in the design of supercomputers, ML, and data sharing. Investments in observational
networks should be coordinated with NEWP system development (e.g., determination of the
likely forecast improvement from a proposed observing strategy, timely development of data
assimilation systems) to significantly and efficiently improve forecasts and maximize the
impact of observations on forecast skill.
An integrated system-of-systems approach should be used with considerations of scope,
complementarity, and cost–benefit analysis across all space and time scales. The role of the
WMO will continue to be important for improved coordination of investments and broaden-
ing the participation of the private and academic sectors. The private sector can help make
the case to funders and development agencies to meet these key computing, observational,
and R&D needs including advancing state-run long-term observations, reference networks,
and data sharing.
Changing roles. There will be some notable shifts in roles, functions, and performance
requirements for the NEWP enterprise stemming from the expected breakthroughs in tech-
nology and research developments. The development/improvement of climate models would
benefit from coordination with efforts to advance weather prediction with global weather
models evolving in the years to come to explicitly represent convection. R&D efforts to over-
come fundamental limitations in physics parameterization are accelerating and would benefit
from collaboration between the weather and climate communities. Efforts to develop a unified
single model system (e.g., Côté et al. 1998a,b) across a range of time scales (nowcasting to
centennial) and spatial scales (convective-scale to climate system Earth modeling) should be
accelerated since such systems are possible (i.e., Met Office strategy since 1990), desirable,
and perhaps even necessary for accelerating forecast advancements.
The ECMWF model of international cooperation may need to be expanded and replicated
to a certain extent to other regions, which will allow consolidation of human resources and
expertise, and optimization of running costs. Such a regional approach could be successful
only with strong political support for national-based regional institutions and scientists to
build capacity and consideration of existing regional climate centers. This approach is a
task and challenge for WMO in cooperation with other relevant international organizations,
including development agencies. Private–public engagement could be a guiding principle to
ensure partnerships, availability, and usability of skilled regional/local products.

R I L 2 0 2 3 08/02/23
The challenges of big data. Numerous segments of society are facing challenges in transport-
ing, storing, and utilizing large datasets and NEWP is at the forefront of this big data challenge.
The scope of this challenge is evident in Bauer et al. (2015), which noted that 107 observations
are incorporated each day into operational weather prediction systems that produce 10 TB
of model output per day. Increases in observational datasets (e.g., crowd sourcing, advanced
satellite observations), the movement of NEWP systems to ensembles operating at high resolu-
tion (1–3 km), and attempts to observe and predict more components of the Earth system will
only increase the difficulty humans and our technology face in handling these datasets. This
big data challenge will pose a key dilemma for the NWEP enterprise and for society.
High-resolution global ensemble modeling systems will continue to be run operationally
only by those organizations having access to very significant supercomputing capabilities.
Currently many NMHSs and private sector entities are operating their own regional modeling
systems. The quality and small-scale details of these new global models will likely exceed
locally run models operating with limited computational resources. Thus, local models may
appear to be a waste without adding value to freely available high-resolution global models.
One option is the downscaling of forecast data from the global ensembles into useful
products, including the use of AI, which would be highly valuable for the various national
impact sectors, especially for low-income countries. This approach will require using the
reforecast data from the global ensembles and training scientists from NMHSs in applying
these approaches to local scales in their own countries with national observations utilized
to optimize the calibrations. This shift might be one of the most disruptive changes of the
coming decade requiring advanced planning and change management. The dissemination
and utilization of ECMWF and North American Ensemble Forecast System forecasts by many
NMHSs and private organizations are showing the viability of such an option.
Implementing NEWP systems with postprocessing, production, and visualization in the
cloud may offer a more sustainable and cost-effective approach. Such an approach could
benefit nations and offer improved cooperation and burden/benefit sharing between public
and private partners. Certain competencies and financial resources will be needed with an
emphasis on maximizing scalability (e.g., cloud-based software solutions that can be custom-
ized and redeployed to assist countries with similar needs).
Progressing together with low-income nations. The importance of quality observations

to global prediction means that stabilizing the national observing networks in low-income
countries and ensuring their long-term maintenance to provide standard-quality observa-
tional data are of paramount importance. One striking element of addressing this situation
is that the massive investments (estimated as billions of dollars) through various develop-
ment assistance projects, engaging numerous development funding agencies, have often
provided disappointingly low impacts on local capacity. Some analysts pointed out that the
main reason for such a failure is that the development funds were spent mostly on equip-
ment such as automatic weather stations, communication systems, computers to run NEWP,
and weather radars (widely not utilized after delivery due to maintenance problems). Unfor-
tunately, less attention is often given to the longer-term success of these investments, which
requires planning and funding to maintain and sustain instrumentation, improve communi-
cation systems and service delivery, capacity building including training and retaining staff,
access to and utilization of observations and forecast products from the WMO and major
operational centers, engagement with the private and academic sectors, and advanced data
management systems (e.g., Snow et al. 2016; Rogers et al. 2019).
In addition, partnerships between the private, public, and academic sectors has recently
been shown to be useful in fostering development and building and sustaining the NEWP
enterprise (World Bank 2020). The WMO, in collaboration with development agencies, needs

R I L 2 0 2 3 08/02/23
to play an enlarged role in supporting low-income countries to contribute effectively and
sustainably to their national observational programs as components of the global observ-
ing system. Fortunately, many players in the weather enterprise well understand the need
for a paradigm change in the provision of development assistance. As an example, WMO
has recently established the Systematic Observations Financing Facility (SOFF) to provide
financial and technical support for the implementation and sustained operation of GBON for
low-income countries.
Regarding NEWP, as mentioned previously, a more efficient path might be to invest available
resources to serve national purposes through building sustainable expertise and computation
infrastructure for accessing and utilizing available data from international sources like
high-resolution global model products. Sharing observations and the best global and regional
prediction datasets with low-income countries is consistent with obligations of every WMO
member. There is currently a marked inequality in the availability of weather and climate
forecast information in low-income countries, and insufficient capacity to produce and utilize
such information to save lives and bring socioeconomic benefits.
Hence, it is vital that all countries have the expertise and technical capability to utilize the
highest-quality weather and climate information. NEWP model providers and HPC vendors
could work together to integrate multimodels (using multicenters and at multiscales) in HPC
clusters with preoptimization, then make those models available on the cloud as a service.
Rather than buying HPC and running a NEWP model, the low-income countries could buy
cloud service to run NEWP systems or postprocess global model output for their specific
region in an affordable way, without the burden of future heterogeneous HPC, optimization of
models, and downloading massive amounts of observational data. In this regard, we note the
progress and success of Copernicus, Earth on AWS, Google Earth Engine, and other ventures
to make data more accessible.
A closer strategic alignment among low-income countries and enhancing regional me-
teorological cooperation is essential. Scientists from low-income countries can contribute to
the development of global NEWP systems. When a NEWP system does not perform well for
a specific high-impact event, regional scientists need to be able to study the factors that the
forecast had issues with and make recommendations on how to improve the global forecast
system. These international collaborations will also act as a catalyst for technical capacity
building, engendering improvements in education and infrastructure. International col-
laborations have been shown to be a catalyst in furthering research productivity across the
low-income and developed nations of the world (e.g., Wagner et al. 2015) and will therefore
accelerate innovation across the entire NEWP enterprise. Thus, global cooperation would im-
prove regional capacity in meteorological disaster monitoring, prediction, and early warning
and help to achieve sustainable development of the global economy and society. An excellent
model to build on is three International Desks (i.e., African, South American, and Tropical) at
NOAA’s National Centers for Environmental Prediction. This program hosts ~20 visitors each
year with its in-residence training program and also includes on-site training via workshops,
and follow-up training through distance learning.
NEWP services also require investment in academic institutions in low-income countries
to carry out targeted research, scientific consulting, and training of future expert forecast-
ers and communicators. It is critical that efforts to grow a cross-national, multigenerational
community of scholars consider rethinking the way experts teach, research, publish, and
influence across the whole multidisciplinary domain of weather and climate forecasting.
Targeted funding opportunities would prove beneficial, and we note that for U.S. research-
ers efforts aimed at advancing education, training, and/or forecasting in low-income nations
would contribute to meeting the broader impact components of NSF grants. Collaborations
between educational institutions within low-income nations and between universities in

R I L 2 0 2 3 08/02/23
developed and low-income countries would also be advantageous in enhancing education
and training. Increasing partnerships between academic institutions and those working on
development activities relevant to the NEWP enterprise (e.g., WMO, World Bank, U.S. Agency
for International Development) are likely to be beneficial. Finally, as noted in World Bank
(2020), cooperation between the private, public, and academic sectors also greatly contributes
to long-term advances in the NEWP enterprise and training the next generation.
Finally, we note that a disproportionately large impact of weather and climate extremes
and other high-impact events fall on the poor and those living in the low-income nations of
the world (e.g., Hallegatte et al. 2017; Rentschler and Salhab 2020). Insights from scientists
around the world in the selection of research problems aimed at reducing these large detri-
mental impacts would advance observational, modeling, and postprocessing strategies and
reduce the large detrimental impacts. Success in this approach requires that the societal
needs and research findings be shared with and fed back to the primary forecast centers to
obtain improvements in new operational cycles. Established frameworks and forums, such
as the WMO Regional Climate Centres, Regional Climate Outlook Forums, and the WWRP
and WCRP should exploit their full potential in this endeavor including a focus on translat-
ing these advances with organizations focused on a more sustainable and equity future
(e.g., START, IRDR, Future Earth).
Conclusions
The decade 2021–30 will be the decisive period for realization of the 17 United Nations
Sustainable Development Goals and targets outlined in the Sandei Framework. Most of these
goals have links with the changing environment—climate change, water resources, and
high-impact weather and environmental events. The desired outcomes in all areas require
enhanced resilience, which is also the main call of the WMO Vision 2030 (WMO 2019b). The
advances expected in weather forecasting and climate prediction during this decade will
support those ambitious goals by enabling a next generation of weather and climate services
that help people, businesses, and governments to better mitigate risks, reduce losses, and
materialize opportunities from the new intelligence of highly accurate and reliable forecasts.
These advances will also help assess systemic risks arising from the long-term evolution of
weather and climate high-impact events, variability, and change across time and space.
This paper has shown the major challenges of and opportunities in each of the first three
components of the innovation cycle (see Fig. 2): infrastructure, R&D, and operation. Several
key factors will determine the success and effectiveness of future progress in NEWP, includ-
ing open data access, ability for near-real-time data processing and transmission, monitoring
data and forecast quality, quick and focused dissemination to users, and advances in societal
decision-making. Some of these factors are common to the three components of the innovation
cycle (infrastructure, R&D, and operation), but their implication and logic can be different, such
as sharing data policies or HPC requirements for R&D and operation. It has been emphasized
that the only way to address these challenges and enable the uptake of the technological and
scientific achievements is through inclusive public, private, and academic partnerships.
Acknowledgments. The contributors to this white paper express their great sadness on the tragic
demise of two of contributors to this article with major input to the “Challenges and opportunities
in the coming decade: Recommendations” section. Prof. Laban A. Ogallo passed away in November
2020. Prof. Ogallo was one of the pioneers of climate science in Africa whose research contributions
include playing a lead role in the identification of leading modes of climate variability, such as the
decadal variability mode. He was a recipient of the WMO Award for the Encouragement of Young
Scientists. The second colleague was Netatua Pelesikoti, the first Pacific Islands woman to become a
lead author of an IPCC report. She was the director of the Climate Change Division at the Secretariat

R I L 2 0 2 3 08/02/23
of the Pacific Regional Environment Programme for more than 7 years and a consultant to the World
Bank. In 2019, she became a member of the International Scientific Advisory Panel for the World
Meteorological Organization She sadly passed away in Tonga in late 2020.
The funding sources for participation in the grant include the following: David Parsons received
funding from the ONR Grant N00014-18-1-2163 and NSF Grant 1921587. Tim Palmer was funded by
the European Research Council Advanced Grant ITHACA 74112. Thomas F. Stocker received funding
from SNF Project 200020_200492.
Data availability statement. This paper is based on a WMO report, and thus, any information utilized
in this manuscript will be made readily available upon request due to the WMO’s open data policy.

R I L 2 0 2 3 08/02/23
References Majumdar, S. J., and Coauthors, 2021: Multiscale forecasting of high-impact
weather: Current status and future challenges. Bull. Amer. Meteor. Soc., 102,
Abbe, C., 1901: The physical basis of long-range weather forecasts. Mon. Wea. E635–E659, https://doi.org/10.1175/BAMS-D-20-0111.1.
Rev., 29, 551–561, https://doi.org/10.1175/1520-0493(1901)29[551c:TPBOLW] Mansfield, L. A., P. J. Nowack, M. Kasoar, R. G. Everitt, W. J. Collins, and
2.0.CO;2. A. Voulgarakis, 2020: Predicting global patterns of long-term climate change
Balmaseda, M., and Coauthors, 2020: NOAA-DOE Precipitation Processes and from short-term simulations using machine learning. Climate Atmos. Sci., 3,
Predictability Workshop report. NOAA Tech. Rep. OAR CPO-9, 48 pp. 44, https://doi.org/10.1038/s41612-020-00148-5.
Bauer, P., A. Thorpe, and G. Brunet, 2015: The quiet revolution of numerical weather McGovern, A., K. L. Elmore, D. J. Gagne II, S. E. Haupt, C. D. Karstens,
prediction. Nature, 525, 47–55, https://doi.org/10.1038/nature14956. R. Lagerquist, T. Smith, and J. K. Williams, 2017: Using artificial intelligence
Benjamin, S. G., J. M. Brown, G. Brunet, P. Lynch, K. Saito, and T. W. Schlatter, to improve real-time decision-making for high-impact weather. Bull. Amer.
2019: 100 years of progress in forecasting and NWP applications. A Century Meteor. Soc., 98, 2073–2090, https://doi.org/10.1175/BAMS-D-16-0123.1.
of Progress in Atmospheric and Related Sciences: Celebrating the American Meehl, G. A., G. J. Boer, C. Covey, M. Latif, and R. J. Stouffer, 2000: The Coupled
Meteorological Society Centennial, Meteor. Monogr., No. 59, Amer. Meteor. Model Intercomparison Project (CMIP). Bull. Amer. Meteor. Soc., 81, 313–318,
Soc., https://doi.org/10.1175/AMSMONOGRAPHS-D-18-0020.1. https://doi.org/10.1175/1520-0477(2000)081<0313:TCMIPC>2.3.CO;2.
Bjerknes, V., 1904: Das Problem der Wettervorhersage, betrachtet vom Stand- Miyamoto, Y., Y. Kajikawa, R. Yoshida, T. Yamaura, H. Yashiro, and H. Tomita,
punkte der Mechanik und der Physik. Meteor. Z., 21, 1–7. 2013: Deep moist atmospheric convection in a subkilometer global simula-
Bock, L., and Coauthors, 2020: Quantifying progress across different CMIP phases tion. Geophys. Res. Lett., 40, 4922–4926, https://doi.org/10.1002/grl.50944.
with the ESMValTool. J. Geophys. Res. Atmos., 125, e2019JD032321, https:// National Research Council, 2003: Fair Weather: Effective Partnership in Weather
doi.org/10.1029/2019JD032321. and Climate Services. National Academies Press, 238 pp.
Brasseur, G. P., and D. Carlson, 2015: Future directions for the World Climate NOAA, 2020: Precipitation prediction grand challenge strategy. NOAA Rep., 44 pp.,
Research Programme. Eos, 96, https://doi.org/10.1029/2015EO033577. www.noaa.gov/sites/default/files/2022-01/PPGC-Strategy_FINAL_2020-1030.pdf.
Brunet, G., and Coauthors, 2010: Collaboration of the weather and climate ——, 2021: NOAA Artificial intelligence strategy—Analytics for next generation
communities to advance subseasonal-to-seasonal prediction. Bull. Amer. Earth science. NOAA Rep., 8 pp., https://sciencecouncil.noaa.gov/Portals/0/
Meteor. Soc., 91, 1397–1406, https://doi.org/10.1175/2010BAMS3013.1. 2020%20AI%20Strategy.pdf.
——, and Coauthors, 2021: Future of weather and climate forecasting. WHO Rep. Orrell, G., L. Pelhemus, C. Reese, and D. Zaharchuk, 2018: Just add weather—How
WMO-1263, 44 pp., https://library.wmo.int/doc_num.php?explnum_id=10611. weather insights can grow your bottom line. IBM Institute for Business Value
Charney, J. G., 1950: Progress in dynamic meteorology. Bull. Amer. Meteor. Soc., Rep., 8 pp., www.weather.gov/media/wrn/calendar/JustAddWeatherIBVStudy
31, 231–236, https://doi.org/10.1175/1520-0477-31.7.231. Summary.pdf.
Côté, J., S. Gravel, A. Méthot, A. Patoine, M. Roch, and A. Staniforth, 1998a: The Overeem, A., H. Leijnse, and R. Uijlenhoet, 2013: Country-wide rainfall maps from
operational CMC-MRB Global Environmental Multiscale (GEM) model. Part I: cellular communication networks. Proc. Natl. Acad. Sci. USA, 110, 2741–2745,
Design considerations and formulation. Mon. Wea. Rev., 126, 1373–1395, https://doi.org/10.1073/pnas.1217961110.
https://doi.org/10.1175/1520-0493(1998)126<1373:TOCMGE>2.0.CO;2. Parsons, D. B., and Coauthors, 2017: THORPEX research and the science of
——, J.-G. Desmarais, S. Gravel, A. Méthot, A. Patoine, M. Roch, and A. Staniforth, prediction. Bull. Amer. Meteor. Soc., 98, 807–830, https://doi.org/10.1175/
1998b: The operational CMC-MRB Global Environmental Multiscale (GEM) BAMS-D-14-00025.1.
model. Part II: Results. Mon. Wea. Rev., 126, 1397–1418, https://doi.org/10. Pudykiewicz, J., and G. Brunet, 2008: The first hundred years of numerical
1175/1520-0493(1998)126<1397:TOCMGE>2.0.CO;2. weather prediction. Large-Scale Disasters: Prediction, Control, and Mitigation,
Eyring, V., and Coauthors, 2016: Towards improved and more routine Earth system M. Gad-El-Hak, Ed., Cambridge University Press, 427–446.
model evaluation in CMIP. Earth Syst. Dyn., 7, 813–830, https://doi.org/10. Rabier, F., and Coauthors, 2010: The Concordiasi project in Antarctica. Bull. Amer.
5194/esd-7-813-2016. Meteor. Soc., 91, 69–86, https://doi.org/10.1175/2009BAMS2764.1.
Hallegatte, S., A. Vogt-Schilb, M. Bangalore, and J. Rozenberg, 2017: Unbreakable: Rentschler, J., and M. Salhab, 2020: People in harm’s way: Flood exposure and pov-
Building the resilience of the poor in the face of natural disasters. World Bank erty in 189 countries. World Bank Policy Research Working Paper 9447, 26 pp.
Rep., 187 pp. Richardson, L. F., 1922: Weather Prediction by Numerical Process. Cambridge
Hintz, K. S., and Coauthors, 2019: Collecting and utilising crowdsourced data University Press, 236 pp.
for numerical weather prediction: Propositions from the meeting held in Ritchie, H., and Coauthors, 2022: Recherche en prévision numérique contributions
Copenhagen, 4–5 December 2018. Atmos. Sci. Lett., 20, e921, https://doi. to numerical weather prediction. Atmos.–Ocean, 60, 35–64, https://doi.org/
org/10.1002/asl.921. 10.1080/07055900.2022.2038071.
IPCC, 2013: Climate Change 2013: The Physical Science Basis. Cambridge Rodwell, M. J., D. S. Richardson, D. B. Parsons, and H. Wernli, 2018: Flow-dependent
University Press, 1535 pp., https://doi.org/10.1017/CBO9781107415324. reliability: A path to more skillful ensemble forecasts. Bull. Amer. Meteor. Soc.,
——, 2021: Climate Change 2021: The Physical Science Basis. V. Masson- 99, 1015–1026, https://doi.org/10.1175/BAMS-D-17-0027.1.
Delmotte et al., Eds., Cambridge University Press, 3949 pp. Rogers, D. P., V. V. Tsirkunov, H. Kootval, A. Soares, D. Kull, A. M. Bogdanova, and
——, 2022: Summary for policymakers. Climate Change 2022: Impacts, M. Suwa, 2019: Weathering the change: How to improve hydromet services
Adaptation, and Vulnerability, H.-O. Pörtner et al., Eds., Cambridge University in developing countries? World Bank Rep., 112 pp.
Press, 3–33. Skamarock, W. C., J. B. Klemp, M. G. Duda, L. D. Fowler, S.-H. Park, and T. D. Ringler,
Krishnamurthy, V., 2019: Predictability of weather and climate. Earth Space Sci., 2012: A multiscale nonhydrostatic atmospheric model using centroidal
6, 1043–1056, https://doi.org/10.1029/2019EA000586. Voronoi tesselations and C-grid staggering. Mon. Wea. Rev., 140, 3090–3105,
Lazo, J. K., M. Lawson, P. H. Larsen, and D. M. Waldman, 2011: U.S. economic https://doi.org/10.1175/MWR-D-11-00215.1.
sensitivity to weather variability. Bull. Amer. Meteor. Soc., 92, 709–720, Snow, J. T., B. Biagini, G. Benchwick, G. George, J. Hoedjes, A. Miller, and J. Usher,
https://doi.org/10.1175/2011BAMS2928.1. 2016: A new vision for weather and climate services in Africa. UNDP Rep.,
Lorenz, E. N., 1963: The predictability of hydrodynamic flow. Trans. N. Y. Acad. Sci., 148 pp., www.thegef.org/sites/default/files/publications/WeatherAndClimate
25, 409–432. ServicesAfrica.pdf.
Magnusson, L., and E. Källén, 2013: Factors influencing skill improvements in the Stevens, B., and Coauthors, 2019: DYAMOND: The Dynamics of the Atmospheric
ECMWF forecasting system. Mon. Wea. Rev., 141, 3142–3153, https://doi. general circulation Modeled On Non-hydrostatic Domains. Prog. Earth Planet.
org/10.1175/MWR-D-12-00318.1. Sci., 6, 61, https://doi.org/10.1186/s40645-019-0304-z

R I L 2 0 2 3 08/02/23
UNISDR, 2015: Sendai framework for disaster risk reduction 2015–2030. UNISDR World Bank, 2020: The power of partnership: Public and private engagement
Rep., 37 pp., www.wcdrr.org/uploads/Sendai_Framework_for_Disaster_Risk_ in hydromet services. World Bank Rep., 148 pp., https://alliancehydromet.
Reduction_2015-2030.pdf. org/wp-content/uploads/2021/07/The-Power-of-Partnership-Public-and-
United Nations, 2019: World urbanization prospects: The 2018 revision (ST/ESA/ Private-Engagement-in-Hydromet-Services.pdf.
SER.A/420). United Nations Rep., 126 pp., https://population.un.org/wup/ World Meteorological Organization, 2015: Seamless prediction of the Earth
publications/Files/WUP2018-Report.pdf. system: From minutes to months. WMO Rep. WMO-1156, 483 pp., https://
Vitart, F., 2017: Madden–Julian oscillation prediction and teleconnections in library.wmo.int/doc_num.php?explnum_id=3546.
the S2S database. Quart. J. Roy. Meteor. Soc., 143, 2210–2220, https://doi. ——, 2019a: Report of High-Level Round Table (5–6 June 2019, Palexpo,
org/10.1002/qj.3079. Geneva): Open consultative platform, partnership and innovation for the next
Wagner, C. S., H. W. Park, and L. Leydesdorff, 2015: The continuing growth of generation of weather and climate intelligence. WMO Rep., 37 pp.
global cooperation networks in research: A conundrum for national gov- ——, 2019b: WMO strategic plan 2020–2023. WMO Rep. 1225, 22 pp., https://
ernments. PLOS ONE, 10, e0131816, https://doi.org/10.1371/journal.pone. library.wmo.int/doc_num.php?explnum_id=9939.
0131816. ——, 2019c: Manual on the global data-processing and forecasting system.
Wake, B., 2019: Climate research for the twenty-first century. Nat. Climate Annex IV to the WMO Technical Regulations. WMO Rep. 485, 147 pp., www.
Change, 9, 183–185, https://doi.org/10.1038/s41558-019-0428-6. nmc.cn/userfiles/ueditor/userid/files/cms/article/2019/07/1562050831112.
Wang, B., J. Lu, Z. Yan, H. Luo, T. Li, Y. Zheng, and G. Zhang, 2019: Deep uncer- pdf?v=1562050831245.
tainty quantification: A machine learning approach for weather forecasting. Zhang, G., and Coauthors, 2019: Current status and future challenges of weather
Proc. 25th ACM SIGKDD Int. Conf. on Knowledge Discovery and Data Mining, radar polarimetry: Bridging the gap between radar meteorology/hydrology/
Anchorage, AK, Association for Computing Machinery, 2087–2095, https:// engineering and numerical weather prediction. Adv. Atmos. Sci., 36, 571–588,
doi.org/10.1145/3292500.3330704. https://doi.org/10.1007/s00376-019-8172-4.

R I L 2 0 2 3 08/02/23

Noaa 52721 DS1

Uploaded by

Copyright:

Available Formats

Noaa 52721 DS1

Uploaded by

Document Information

Original Title

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

Noaa 52721 DS1

Uploaded by

Copyright:

Available Formats

Article

Advancing Weather and Climate Forecasting

KEYWORDS: Climate prediction; Nowcasting; Operational forecasting; Seasonal forecasting;

AMERICAN METEOROLOGICAL SOCIETY

AMERICAN METEOROLOGICAL SOCIETY

AMERICAN METEOROLOGICAL SOCIETY

Objective and scope

AMERICAN METEOROLOGICAL SOCIETY

AMERICAN METEOROLOGICAL SOCIETY

Weather and climate forecasting: Setting the scene

Major milestones in operational weather and climate forecasting

AMERICAN METEOROLOGICAL SOCIETY

AMERICAN METEOROLOGICAL SOCIETY

AMERICAN METEOROLOGICAL SOCIETY

AMERICAN METEOROLOGICAL SOCIETY

Challenges and opportunities in the coming decade: Recommendations

AMERICAN METEOROLOGICAL SOCIETY

Some predictability, dynamical, and physical modeling challenges

AMERICAN METEOROLOGICAL SOCIETY

Potential major breakthroughs from ML

AMERICAN METEOROLOGICAL SOCIETY

Toward improved systems for forecasting

AMERICAN METEOROLOGICAL SOCIETY

Progressing together with low-income nations. The importance of quality observations

AMERICAN METEOROLOGICAL SOCIETY

AMERICAN METEOROLOGICAL SOCIETY

AMERICAN METEOROLOGICAL SOCIETY

AMERICAN METEOROLOGICAL SOCIETY

AMERICAN METEOROLOGICAL SOCIETY

AMERICAN METEOROLOGICAL SOCIETY

You might also like