10.1007@978 981 13 8363 2 PDF

India Studies in Business and Economics
K. Palanisami
Krishna Reddy Kakumanu
Udaya Sekhar Nagothu
C. R. Ranganathan
Climate Change
and Future Rice
Production in India
A Cross Country Study of Major Rice
Growing States of India
The Indian economy is considered to be one of the fastest growing economies of the
world with India amongst the most important G-20 economies. Ever since the
Indian economy made its presence felt on the global platform, the research
community is now even more interested in studying and analyzing what India has to
offer. This series aims to bring forth the latest studies and research about India from
the areas of economics, business, and management science. The titles featured in
this series will present rigorous empirical research, often accompanied by policy
recommendations, evoke and evaluate various aspects of the economy and the
business and management landscape in India, with a special focus on India’s
relationship with the world in terms of business and trade.
More information about this series at http://www.springer.com/series/11234

K. Palanisami Krishna Reddy Kakumanu
• •
Udaya Sekhar Nagothu C. R. Ranganathan

•
Climate Change and Future

Rice Production in India
A Cross Country Study of Major Rice
Growing States of India
123
K. Palanisami Krishna Reddy Kakumanu
Emeritus Scientist Centre for Natural Resource
International Water Management Institute Management (CNRM)
New Delhi, India National Institute of Rural Development &
Panchayati Raj (NIRDPR)
Udaya Sekhar Nagothu Hyderabad, India
Norwegian Institute of Bioeconomy
Research C. R. Ranganathan
Ås, Akershus, Norway Department of Computing
Coimbatore Institute of Technology
Coimbatore, India
ISSN 2198-0012 ISSN 2198-0020 (electronic)

ISBN 978-981-13-8362-5 ISBN 978-981-13-8363-2 (eBook)
https://doi.org/10.1007/978-981-13-8363-2
© Springer Nature Singapore Pte Ltd. 2019
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Foreword
Agriculture systems around the world, in general, and in India, in particular, are
subjected to severe climatic interannual variability and extreme events such as
floods, droughts, floods, and hailstorms, all intensified by climate change. Unless
climate change impacts are addressed, our agricultural systems will become more
vulnerable and put national food and global food security at risk.
Rice is a water-intensive crop and predominantly grown in irrigated conditions.
In India, 96 million tons of rice is produced on 44 million ha. This constitutes about
one-fourth of the total food grains produced from all the major cereal crops grown.
India still needs to produce more rice to feed an increasing population. Because
of the water scarcity, it is important to produce more rice with less water. With the
impacts of climate change and variability, this will be a formidable challenge for
achieving food security for the country.
A set of new interventions in rice production and management are required to
address climate change and its various impacts and to see how they could be
effectively developed and upscaled. Several studies have indicated that this is
possible but do not address how this can be done holistically at a national level.
Moreover, the economics of such adaptation strategies has not been studied in
detail. I find that in addition to quantifying the impact of climate change on rice for
all the regions, this book also evaluates the economics of management technologies
in different rice growing regions. This will be incredibly important for policy
analysis and formulation.
This book, a first of its kind, focuses on impacts of climate change on rice
production. It displays a region-specific impact of climate change on productivity
and production of rice and suggests relevant adaptation strategies to address these
impacts. I am sure the key messages from this book will be highly relevant to
planning future adaptation strategies not only in India but also in other rice growing
regions.
v
vi Foreword
I applaud the authors who have taken so much effort to put together this very
valuable information in a readable form. I congratulate the research team for this
rich contribution. This book will be of immense help to practitioners and policy
makers by suggesting necessary climate-smart adaptation measures not only in
India but also in other rice growing regions of the world.
David Molden
Director General
International Centre for Integrated Mountain
Development GPO Box 3226 Kathmandu Nepal
http://www.icimod.org
http://www.facebook.com/icimod
http://www.twitter.com/icimod
http://www.youtube.com/icimod
Preface
Warming of the climate system in recent decades is unequivocal. A warmer climate

may lead to an intensification of the hydrological cycle, resulting in higher rates of
evaporation, and increase/decrease in precipitation, resulting in with varying
intensity and extreme events of rainfall. Monsoon rainfall is considered very
important in the Indian subcontinent, and the utility of precipitation primarily
depends upon its spatial and temporal distribution. It will increase flood risk during
the rainy season and strongly reduce water supplies during the dry season. The
consequences of changing weather pattern and hydrological cycle might cause
different levels of water scarcity and consequently will have an impact on crop
yields by triggering associated factors like pests and diseases. The expected changes
in climate will ultimately alter regional agricultural systems, with consequences for
food production.
Understanding climate change is a challenge particularly due to difficulties in
precisely quantifying and effectively identifying the management options, and this
opens a wide range of opportunities for integrating adaptations taking into account
the economic and social development aspects. Understanding those challenges and
tackling them creatively can make climate change adaptation an important way to
help build a more resilient farming system. Therefore, the strategies and tools used
for coping with within and between crop season weather variability must be
strengthened to adapt to climate change. At the farm level, small-scale farmers are
interested in improved soil and land management technologies that can also benefit
them with carbon credits. They need technological support to address the chal-
lenges in water and soil management.
In recent years, researchers and practitioners are developing quantification
methods and identifying appropriate adaptation technologies and strategies to the
changing climate. However, these methods were applied in different time periods
covering different regions and various crops. Even though technologies/improved
practices are available to the farmers, the adoption level is very low which ranges
from 2 to 30%. Hence, what kind of policy and institutional support could help
address the adaptation is also needed in order to address these issues and challenges
relating to climate change and agriculture.
vii
viii Preface
To address these issues and the emerging challenges, a collaborative research

study was coordinated by the Norwegian Institute for Agricultural and
Environmental Research (Bioforsk) with funding from the Norwegian Embassy,
New Delhi. As part of this project, a study was carried out to quantify the impact of
climate change on rice productivity and identification of suitable adaptation
strategies for 13 major rice growing states of India applying uniform methodology
and data set. The study results could provide answers to most of the questions
relating to the quantification of climate change impacts on rice and the financial
feasibility of the identified adaptation strategies. As an outcome of this in-depth
research study, this book has been published.
Chapter 1 of this book gives an overview of climate change, its challenges to
ensure water and food security and the need for alternative adaptation strategies to
reduce the variability. Chapter 2 discusses in detail the rice area, production and
productivity trends in India. Decadal growth rates were explained for the central,
southern, northern, western and eastern regions of India. Chapter 3 provides the
climate change projections using the available data and narrows down the projec-
tions to the rice growing regions of India. Chapter 4 explains in detail the concepts,
tools and approaches that are used to quantify the climate change impacts. Chapters
5–9 provide results of the detailed analysis of the perception of the farmers on
climate change and the impact of climate change on rice productivity and pro-
duction in northern, eastern, western, central and southern regions, respectively.
Chapter 10 presents the climate-smart adaptation practices and their financial via-
bility at the farm level, and the last chapter summarizes the key findings and
provides recommendations for stabilizing the rice productivity and production in
future. We believe that the key messages from this book will have more impact in
mainstreaming and upscaling the appropriate adaptation strategies at both com-
munity and state levels.
The authors acknowledge with thanks the excellent technical support extended
by Dr C. Karthikeyan, Professor of Agrl. Extension, Dr R. Venkatraman, Agrl.
Economist and Former Director of Planning and Monitoring, Dr K. R. Karunakaran,
Professor of Agrl. Economics, and Dr T. Arivelarasan, Post doc Fellow of Tamil Nadu
Agricultural University while completing the final draft of this book.
New Delhi, India K. Palanisami

Hyderabad, India Krishna Reddy Kakumanu
Ås, Akershus, Norway Udaya Sekhar Nagothu
Coimbatore, India C. R. Ranganathan
Contents
1 Climate Change and Agriculture in India . . . . . . . . . . . . . . . . . ... 1

K. Palanisami, Krishna Reddy Kakumanu, Udaya Sekhar Nagothu
and C. R. Ranganathan
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Climate Change Impacts on Water and Agriculture . . . . . . . . . . 2
1.3 Relevance to Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Outline of the Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2 Rice Production in India: Analysis of Trend, Constraints
and Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....... 7
D. Suresh Kumar
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Trend in Area, Production and Productivity . . . . . . . . . . . . . . . 8
2.2.1 Trend in Area Under Rice . . . . . . . . . . . . . . . . . . . . . 8
2.2.2 Trend in Rice Production . . . . . . . . . . . . . . . . . . . . . . 11
2.2.3 Trend in Rice Productivity . . . . . . . . . . . . . . . . . . . . . 13
2.3 Constraints in Rice Production . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.4 Climate Change and Rice Production . . . . . . . . . . . . . . . . . . . . 22
2.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3 Climate Change Projections for Major Rice Growing States
of India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 35
S. Senthilnathan
3.1 Selection of Global Climate Models for Climate Change
Projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2 Approach and Methodology for Extracting Climate Data . . . . . . 38
3.3 Spatial Domains of Major Rice Growing States in India . . . . . . 38
3.4 Historical Observed Climate Data . . . . . . . . . . . . . . . . . . . . . . 40
ix
x Contents
3.4.1 Seasonality of Rainfall . . . . . . . . . . . . . . . . . . . . . . . . 40

3.4.2 Seasonality of Temperature . . . . . . . . . . . . . . . . . . . . . 41
3.5 Evaluation of Baseline Climate with IMD Observations . . . . . . 41
3.5.1 Rainfall Seasonal Cycle . . . . . . . . . . . . . . . . . . . . . . . 42
3.5.2 Temperature Seasonal Cycle . . . . . . . . . . . . . . . . . . . . 43
3.6 Future Climate Change Projections . . . . . . . . . . . . . . . . . . . . . 47
3.6.1 Seasonal Cycle of Rainfall Projections
of HadGEM2 Model . . . . . . . . . . . . . . . . . . . . . . . . .. 49
3.6.2 Seasonal Cycle of Maximum Temperature Projections
of HadGEM2 Model . . . . . . . . . . . . . . . . . . . . . . . . .. 49
3.6.3 Seasonal Cycle of Minimum Temperature Projections
of HadGEM2 Model . . . . . . . . . . . . . . . . . . . . . . . . .. 51
3.6.4 Rainfall, Maximum and Minimum Temperature
Projections of GFDL Model . . . . . . . . . . . . . . . . . . .. 55
3.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 55
4 Methodologies for Quantifying Climate Change Impacts
on Rice Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 57
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 57
4.2 Study Area and Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 58
4.3 Statistical Tools for Analysis of Primary Data . . . . . . . . . . . .. 61
4.4 A Review of Studies on Economic Impact of Climate Change
on Crop Yields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.5 Econometric Tools for Studying Climate Change Impacts . . . . . 69
4.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Appendix: Sample Gretl Scripts, Do File and Sample Outputs . . . . . . . 77
5 Climate Change Impacts on Rice Yield in Northern Region
of India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 87
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 87
5.2 Socio-economic Profiles of the Farmers . . . . . . . . . . . . . . . . .. 87
5.3 Farm Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 87
5.4 Climate Change Events, Shocks and Their Effects as Perceived
by Farmers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5.4.1 Perception of Farmers on Climate Change Events . . . . 91
5.4.2 Effects of Shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.4.3 Mitigation Strategies Used by Farmers . . . . . . . . . . . . 94
5.4.4 Opinions of Farmers on Government Initiatives . . . . . . 95
Contents xi
5.5 Analysis of the Effect of Climate Change on Rice Yield . . . . . . 96

5.5.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
5.5.2 Test of Normality of Yield Data . . . . . . . . . . . . . . . . . 102
5.5.3 Panel Root Test: Fisher’s Tests . . . . . . . . . . . . . . . . . . 103
5.5.4 Estimation of Just–Pope Production Function . . . . . . . . 104
5.5.5 Climate Change Scenarios and Rice Yield Changes . . . 108
5.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6 Climate Change Impacts on Rice Yield in Eastern Region
of India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
6.2 Socio-economic Profiles of the Farmers . . . . . . . . . . . . . . . . . . 115
6.3 Farm Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
6.4 Climate Change Events, Shocks and Their Effects
as Perceived by Farmers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
6.5.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
6.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
7 Climate Change Impacts on Rice Yield in Western Region
of India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
xii Contents

7.5.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
7.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
8 Climate Change Impacts on Rice Yield in Central
Region of India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
8.5.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
8.5.3 Panel Root Test: Fisher’s Type Tests . . . . . . . . . . . . . 176
8.5.4 Estimation of Just-Pope Production Function . . . . . . . . 176
8.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
9 Climate Change Impacts on Rice Yield in Southern Region
of India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
9.4 Climate Change Events, Shocks and Their Effects as Perceived
by Farmers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
9.4.1 Perception of Farmers on Climate Change . . . . . . . . . . 189
9.5.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
Contents xiii

9.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
10 Economics of Adoption of Rice Production and Management
Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
10.1 Rice Technology Adoption and Profitability . . . . . . . . . . . . . . . 209
10.1.1 Northern Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
10.1.2 Eastern Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
10.1.3 Western Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
10.1.4 Central Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
10.1.5 Southern Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
10.2 Comparison of Cost and Yield Attributes of Rice Due
to Technology Adoption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
10.3 Constraints in Technology Adoption . . . . . . . . . . . . . . . . . . . . 227
10.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
11 Climate Change and Rice Production in India:
A Way Forward . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
11.2 Overview of Climate Change Impact on Rice
Production in India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
11.3 A Way Forward . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
11.3.1 Future Rice Production Policy . . . . . . . . . . . . . . . . . . . 234
11.3.2 Upscaling and Mainstreaming the Adaptation
Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
11.3.3 Government’s Commitment to Address Climate
Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
11.4 Specific Measure Relevant to Rice . . . . . . . . . . . . . . . . . . . . . . 237
11.4.1 Augmenting and Managing Water Resources . . . . . . . . 237
11.4.2 Managing Soil Health . . . . . . . . . . . . . . . . . . . . . . . . . 238
11.4.3 Developing Multiple Abiotic Stress-Tolerant Rice
with Climate Resilience . . . . . . . . . . . . . . . . . . . . . . . 238
11.4.4 Climate Resilient Rice Production Technologies . . . . . . 239
11.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
About the Authors
K. Palanisami is an agricultural economist and international water resources

expert. He was a Visiting Professor at the University of Minnesota, USA, Principal
Researcher at the International Water Management Institute (IWMI), South Asia
Office, Hyderabad and Director of the Water Technology centre, Tamilnadu
Agricultural University. He is a Fellow of the National Academy of Agricultural
Sciences (NAAS), India. He has led several national and international research
projects on climate change, water and agriculture and published numerous papers
and books. He is currently serving as an agricultural and water expert for the
Interstate Water Resource Department of the Govt of Telangana, India.
Krishna Reddy Kakumanu is an Associate Professor at the Centre for Natural

Resource Management at the National Institute of Rural Development and
Panchayati Raj, Hyderabad, India. He completed his Ph.D. in Agricultural eco-
nomics from Justus Liebig University, Germany and worked as a Regional
Researcher at the International Water Management Institute (IWMI), Hyderabad &
New Delhi, India. He has broad expertise in the economics of conjunctive use of
water, energy pricing and climate change impact assessment and adaptation in
agricultural and water sectors in India. He has 10 years of research experience and
has published several research papers in international journals and books.
Udaya Sekhar Nagothu is a Research Professor and Director (Centre for

International Development) at the Norwegian Institute of Bioeconomy Research,
Norway. He holds an M.Sc. in Natural Resources Management and Ph.D. in
Development Studies from the University of Biosciences, Norway. He has more
than 25 years of research and development experience in natural resource man-
agement and environment related areas. He has coordinated several large inter-
disciplinary projects on climate change, water resources management and food
security in various countries. Sekhar has published articles in several international
journals, and edited and contributed to six books on climate change, sustainable
agriculture and food security.
xv
xvi About the Authors
C. R. Ranganathan is an Adjunct Professor at the Department of Computing,

Coimbatore Institute of Technology, Tamil Nadu. Prior to this, he was working at
Tamil Nadu Agricultural University, Coimbatore for 37 years. He has extensive
experience in teaching mathematics, statistics and data science and has led several
research projects at Tamil Nadu Agricultural University. He has published a
number of research articles in national and international journals, authored the book
entitled ‘A First Course in Mathematical Models of Population Growth’ and
co-authored the book ‘Climate Change and Agriculture in India’.
Chapter 1
Climate Change and Agriculture in India

1.1 Introduction
Climate change is a continuous phenomenon, and its variability is a complex phe-

nomenon. In order to assess its impact and to develop appropriate adaptation mea-
sures, an interdisciplinary approach is essential. A multi-level assessment and high-
resolution data analysis are required to map future climate scenarios and their impacts
with reduced uncertainty. This includes both natural and social science-related data,
including climate and weather, soils, water availability, cropping patterns, productiv-
ity and socio-economic variables that are necessary for impact assessment. Recent
literature shows that socio-economic factors are highly relevant as they determine
the magnitude of impact on the agriculture sector in the event of extreme weather
or climatic event. During the twentieth century, evidences through various studies
showed an increase in global temperature and changes in rainfall patterns and rates
(IPCC 2001; IPCC 2018; Jung et al. 2002). The IPCC (2007) has projected a rise in
temperature of 0.5–1.2 °C by 2020, 0.88–3.16 °C by 2050 and 1.56–5.44 °C by 2080
for South Asia. Similarly, changes in precipitation could be observed as the average
precipitation is expected to increase globally (IPCC 2001). However, the magnitude
of changes in precipitation among various regions varies with models, i.e. from 0 to
50%, where the direction of change is very strongly indicated and between −30 and
+30% where it is not.
Further, the IPCC, in its recent report (2018) indicated that the average global
temperature would rise by 1.5–2.0 °C by 2100. As a consequence, the climate-related
risks to human security, health, livelihood, food security, water supply and economic
growth are projected to increase with global warming.
In India, based on a multi-model climate change, it is likely the country will be
more warmer in the range of 1.7–2 °C by 2030 s and 3.3–4.8 °C by 2080s. In the
same trend, compared with baseline figure (1961–1990), the precipitation would also
tend to increase from 4 to 5% by 2030s and subsequently from 6 to 14% by 2080s
(Chaturvedi et al. 2012). In their study, Bal et al. (2016) show projections of maximum
temperature and rainfall from six models running over a period of 1970–2100. The
© Springer Nature Singapore Pte Ltd. 2019 1
K. Palanisami et al., Climate Change and Future Rice
Production in India, India Studies in Business and Economics,
https://doi.org/10.1007/978-981-13-8363-2_1
2 1 Climate Change and Agriculture in India
Table 1.1 Projected temperature and precipitation changes in India: 2070–2099

Region Jan–March April–June July–Sept Oct–Dec
Change in temperature (°C)
North-east 4.95 4.11 2.88 4.05
North-west 4.53 4.25 2.96 4.16
South-east 4.16 3.21 2.53 3.29
South-west 3.74 3.07 2.52 3.04
Change in precipitation (%)
North-east −9.30 20.30 21.0 7.5
North-west 7.20 7.10 27.2 57.0
South-east −32.9 29.70 10.9 0.7
South-west 22.30 32.32 8.80 8.5
Source Cline, 2007
study concludes that a rise in temperature would be between 2.5 and 4.4 °C and
rainfall would increase in the range between 15 and 24% by the end of the century.
Similarly, the IPCC (2014) had projected that in India during the twenty-first century,
the rainfall will increase by 10–12% with more frequent and heavy rainfall days while
the mean annual temperature will rise by 3–6 °C (Table 1.1).
The water is one of the primary resources that is affected by the climate change.
India is predominantly arid and semi-arid regions. Hence, any shortfall in water sup-
ply, due to climate change, will enhance competition for water use among economic,
social and environmental applications. Besides, population growth with improved
living conditions would also increase the demand for food, thereby increasing the
demand for water many folds, particularly, in the river basins (UNESCO-WWAP
2009). Among the various utility levels, a major portion of available water is used
for agriculture. Such demands pose a big challenge to water management. The sit-
uation will become more critical, at the time of unfavourable weather or climatic
conditions. The climate change in other regions is also expected to impact negatively
and by 2050, more than half of all food crops in sub-Saharan Africa and at least 22%
of the world area under important food crop, i.e. rice, will be affected (Campbell
et al. 2011).
1.2 Climate Change Impacts on Water and Agriculture
Agriculture system in India is subject to severe climatic inter-annual variability,

viz. droughts, floods, storms, etc. The system will become more vulnerable under
different scenarios of climate change. Due to climate change, among various crops,
rice, a major cereal crop would be highly affected. Around 95% of the cultivated
area under rice is grown in irrigated conditions and the crop, depending on the soil
1.2 Climate Change Impacts on Water and Agriculture 3
Fig. 1.1 State-wise production of rice (2009–2016) Source Indiastat 2013, and agricultural statistics
2014, 2016. Note 10 lakhs = 1 million
texture, structure and profile conditions, demands about 1200–2500 mm of water

(Reddy and Reddi 1995). In India, rice, with a production contribution of 96 million
tons from 44 million ha, constitutes 24% of the total food grains produced from all the
major cereal crops grown (Fig. 1.1). To meet the growing food demand, India needs to
produce at least 130 million tons of rice by 2030 (Gujja and Thiyagarajan 2009). The
projected climate scenario for mid (2021–2050)—and end centuries (2071–2100) in
Indian river basins also indicates a change in the future water availability and climatic
factors. Under these situations, farmers need to adapt new technologies to meet the
demand for rice (Palanisami et al. 2011).
The compounded climate factors would lead to a decline in productivity of the
crop. Consequently, an increase in the price of many of the important agricultural
crops could take place. Studies show that a rise in temperature during the last few
years has caused an overall loss in the crop production in the country. Some studies
also indicated with an increase in temperature by 2080–2100, and the probability of
loss in crop production in India would be 10–40% (Aggarwal et al. 2008). Using crop
simulation models (CERES-Rice and ORYZA1 N), Aggarwal and Mall (2002) and
Aggarwal et al. (1997) showed that an increase from 1 to 2 °C temperature without
any increase in CO2 would result in 3–17% decline in rice grain yield in different
regions.
Further, under climate change conditions, Mall et al. (2006), based on crop sim-
ulation modelling in India, evidenced a decline in the yields of important cereal
crops like rice and wheat. Barnwal and Kotani (2010) also found a negative effect on
rice yield and yield variability due to climate change in India. Pathak et al. (2003)
reported that during the growing season, the grain yield of rice would decline by
10% for every 1 °C increase in the temperature above 32 °C. Aggarwal et al. (2009)
also reported a decline in grain yield in rice to the tune of 5.4, 7.4 and 25.1% with
increases in temperatures by 1, 2 and 3 °C, respectively.
Palanisami et al. (2009) also projected the effect of climate variables on three
major crops, namely paddy, sugarcane and groundnut in Tamil Nadu. The results
showed a reduction in both area and yield of major crops by about 5.2–9.5% due
to the impact of climate change. Consequently, overall production would decrease
in the range from 9.00–22.00% for these crops in 2020. The National Action Plan
on Climate Change (GOI, 2004) identified that climate change will contribute to
significant reductions in crop yields and decline in the area cropped in arid and
semi-arid zones of India.
This implies a direct impact on the food security in India where rice is the staple
diet. Despite the constraints of water scarcity and to meet the demand for rice due to
population growth and rapid economic development, it is important to sustain rice
production by increasing rice productivity. Producing more rice with less water is
therefore a formidable challenge for achieving food security for the country.
So far, the popular method in rice cultivation was transplantation using ponded
water that ensured steady yields (Chen et al. 2009). With growing climate uncer-
tainty, insufficient irrigation water both in area commanded by canal and groundwa-
ter increased labour requirement and increased emissions, and such practices may
not be possible in the future. The rise in temperature could negatively impact the rice
yields as they are grown to their threshold (Kelkar and Bhadwal 2007). Further, the
requirement of labour is also comparatively higher in transplanted rice cultivation
and acts as a constraint. It is estimated that in rice cultivation, about 25–37 labourers
(mandays) per hectare are required for transplanting, about 50 mandays per hectare
for weeding and about 25 mandays per hectare for harvesting (Technical program
2011; Prakash et al. 2013). The availability of labour is also a major limiting factor
in most of the Indian states, as rural people migrate to nearby cities and towns for
higher wages or being hired in the government social welfare programmes (e.g. the
National Rural Employment Guarantee programme). Hence, the labour cost has sig-
nificantly increased during recent years (Farooq et al. 2011) and as a consequence,
farm mechanization is catching up fast in many of the rice growing regions in India.
Another major concern among scientists working on climate change is the methane
emissions from rice fields that contribute to global warming (Pathak et al. 2013). The
emissions from flooded fields are found to be higher than those from dry land (Komiya
et al. 2010). The methane is emitted when organic matter decays in anaerobic condi-
tions. Hou et al. (2000) showed that the factors affecting methane and nitrous oxide
emissions are soil temperature and soil redox potential, net irradiance and organic
matter content. The fourth assessment report of IPCC (2007) also indicated that 50%
of the methane is emitted from agriculture and 10–20% of these emissions come
from rice production alone (Reiner and Aulakh 2000). Overall, the South and East
Asian countries are responsible for 82% of methane emissions from rice produc-
tion. From India alone, annually 4.5 million tonnes of methane is being emitted
from rice cultivation (PepsiCo International 2009). There is a need to shift towards
1.2 Climate Change Impacts on Water and Agriculture 5
more climate-resilient rice cropping systems that can help in methane reduction
(Lakshmanan et al. 2009).
Hence, considering the above challenges and to sustain rice production vis-à-vis
ensuring food security, new adaptation strategies have to be developed and up-scaled.
Several studies had indicated that delay in onset of monsoon, insufficient canal
irrigation water at the tail ends of the command area, additional labour requirement
and costs are driving the farmers to adapt to different management practices like
Direct Seeded Rice (DSR), System of Rice Intensification (SRI), modified SRI,
machine transplanting, change in crop varieties, date of sowing/planting, alternate
wetting and drying (AWD) method of irrigation, supplemental irrigation, integrated
nutrient, pest management, etc. (Chapagain et al. 2011: Joshi et al. 2013; Mahajan
et al. 2013; Li et al. 2014; Kakumanu et al. 2018). But these practices are not
adopted fully and in large scale due to lack of timely and insufficient scientific
information, non-availability of inputs and other institutional constraints. Moreover,
the economics of such adaptation strategies has not been studied in detail. One of
the main focuses of this book will be to evaluate the economics of such technologies
in different agro-ecological situations.
1.3 Relevance to Policy
The results from this study will be highly relevant for future adaptation strategy
planning not only in the 13 states where the study was conducted, but also in other
rice growing regions. As rice is one of the main staple crops grown in the country,
governments have to seriously consider planning for the future, provide necessary
infrastructure and build capacity of the smallholders growing rice. The policy should
be supported by evidence-based research results. This book provides necessary rec-
ommendations that will be useful for policymakers at the state and national levels.
The different chapters in the book summarize results specific to the rice growing
regions for the benefit of policymakers and practitioners. Some of the future strate-
gies indicated by the Central Rice Research Institute (CRRI), India, in its Vision
2050 (CRRI 2013) are addressed in this book thus paving the way for a future line
of research and policy interface focusing on climate change and rice. These strate-
gies mainly include: (a) research emphasis on improving water and nutrient use
efficiencies with special focus on conservation agriculture, climate resilience, rice
and rice-based cropping and farming systems, (b) management of knowledge related
to rice, with due attention on extension services and fostering linkage and collabo-
ration among organizations namely public, private, national and international, and
(c) capacity building of scientists, farmers and other stakeholders to be able globally
competitive and to ensure food and nutritional security of the country.
1.4 Outline of the Book
Studies relating to climate change and impacts on rice are scattered and have not
focused much on the economics of the impacts and adoption. Researchers in the
past have used different approaches (ranging from micro-level field experimentation
to macro-level simulation models) for impact assessment. The book integrates and
analyses a large data set from different regions and states growing rice in India, and
presents the scenarios in a comprehensive manner. The study is one of the first of
its kind in the region focusing on the economics of the impacts and adoption. The
book addresses the impact of climate change on rice using a uniform methodol-
ogy/analytical approach and data set related to 13 major rice growing states of India
which are grouped into five rice growing regions. The book displays a region-specific
impact of climate change on productivity and production of rice and suggests relevant
adaptation strategies to address these impacts.
The book contains 11 chapters. The Introduction chapter gives an overview of
climate change, its challenges to ensure water and food security, and the need for
alternative adaptation strategies to reduce the variability. The second chapter dis-
cusses in detail the rice area, production and productivity trends in India. Decadal
growth rates were explained for the northern, central, western, eastern and southern
regions of India. The third chapter provides the climate change projections using the
available data and narrows down the projections to the rice growing regions of India.
Chapter four explains in detail the concepts, tools and approaches that are used to
quantify the climate change impacts. The chapters five to nine provide results of the
detailed analysis of the perception of the farmers on climate change and the impact of
climate change on productivity and production of rice in northern, eastern, western,
central and southern regions, respectively. Chapter ten presents the climate smart
adaptation practices and their financial viability at the farm level and last chapter
summarizes the key findings and provides recommendations for stabilizing the rice
productivity and production in the future.
The strength of the book, therefore, lies in its in-depth focus on climate change,
variabilities and their impacts on productivity and production of rice and the appro-
priate adaptation strategies to address them. Analysis of the economics of various
adaptation strategies relevant to different rice growing regions has added value to
the book. Thus, by addressing the climate change and variability impacts, the book
downsizes the climate change scenarios at state and region levels and uses relevant
econometric modelling methods to integrate and analyse natural and social science-
related data in the process. The book will therefore help practitioners and policy-
makers by suggesting necessary climate smart adaptation measures for stabilizing
rice yields and production in short term and long term.
Chapter 2
Rice Production in India: Analysis
of Trend, Constraints and Technologies
D. Suresh Kumar
2.1 Introduction
Rice is the staple food for more than 50% of the population across the globe. As
a labour-intensive crop, the rice cultivation provides livelihood and employment to
millions of people. The Green Revolution enabled many countries across the globe
to increase the production, and in India besides increasing production, the country
entered into a new era of input use. The usage of high-yielding varieties, fertilizers
and plant protection chemicals has increased manifold particularly in irrigated agri-
culture. The long-term analysis on area under rice, however, indicated that area is
almost stagnated during the past couple of decades. It is estimated that around 130
million tons of rice is needed to feed the population by 2030. With ever-increasing
demand for land and water due to urbanization and industrialization, there is only
little scope for expanding the area under rice. In addition, the other major production
constraints that pose challenges across states are labour and water. Hence, there is a
need to increase productivity (Gujja and Thiyagarajan 2009). Evidences show that
the irrigated rice consumes 34–43% of the world’s irrigation water (Bouman et al.
2007). It is estimated that the total value of rice production in the world is more than
US $150 billion per year (Shetty et al. 2013).
The Green Revolution technologies including adoption of HYVs, fertilizers, pes-
ticides and so on have contributed significantly to achieve increased rice yield in
many countries. It is evident that India with a total rice area of 44 million hectares
The chapter was commissioned by Dr. Palanisami to add different angles to this book. The author
would like to thank Dr. Palanisami for providing substantial inputs in preparing this chapter, without
which, the chapter would not have taken the shape that it has now.
D. Suresh Kumar (B)

Department of Agricultural Economics, Tamil Nadu Agricultural University, Coimbatore, India
e-mail: [email protected]

8 2 Rice Production in India: Analysis of Trend …
(26.3% of the world rice area) ranks first in area and ranks second in rice production
(169 million tonnes) during 2017.
India is not in a good position in terms of rice productivity. Recent estimates show
that India’s rice productivity (3590 kg/ha) is lower than many rice-producing coun-
tries like China (6686 kg/ha), Bangladesh (4219 kg/ha) and Myanmar (4081 kg/ha).
Though India has achieved significant increase in productivity of rice from 1.1 tons/ha
in 1965 to 2.33 t/ha in 2011, still the productivity is much lower than countries
like Egypt, China, Japan and Korea whose rice productivity ranges from 3.70 to
9.00 tons/ha. About 90% of the rice produced in India is used for domestic con-
sumption. Being a carbohydrate-rich cereal, it fulfils 30% of the calories in dietary
requirements (Mclean et al. 2002). Burgeoning population is the major driver for
ever-increasing demand for rice. It is estimated that the demand for rice will be
113.3 million tons by the year 2021–22 (Kumar et al. 2009). In order to meet the
growing demand for rice both for domestic consumption and export demand, it is
essential that the rice productivity and production should be increased to a higher
level.
Keeping the above issues in view, this chapter attempts to examine the growth
rate (compound growth rate) and trend in rice production and productivity across
Indian states. Such analysis will help in targeting appropriate policies to boost rice
production and productivity in poorly performing states.
2.2 Trend in Area, Production and Productivity
2.2.1 Trend in Area Under Rice
With wide variations in agro-economic conditions and socio-economic diversity, the

issue on regional differences in production and productivity of rice is of crucial
interest. Therefore, a proper understanding of dimensions of regional differences in
area, production and productivity of rice is essential for effective regional planning.
The regional differences in area of rice are presented in Table 2.1 (Figs. 2.1, 2.2, 2.3,
2.4 and 2.5).
India has registered only a little growth in area under rice. The rice area has
increased from 38.64 million hectares in 1970s to 43.08 million hectares during
2010s. The compound growth rate (CGR) was only 0.369% per annum. There are
regional differences in area under rice across regions. Among the southern states,
except Tamil Nadu, Andhra Pradesh and Karnataka have registered positive growth
in terms of area though the growth rate is lower. In Tamil Nadu, the area under rice
has registered a negative growth of 0.975%. This is mainly due to water scarcity,
labour scarcity, poor functioning of irrigation tanks and reduction in tank irrigated
area and so on.
Among the states in central India, except Madhya Pradesh, Maharashtra state has
witnessed a slight increase in area and positive growth rate. There is a small negative
2.2 Trend in Area, Production and Productivity 9
Table 2.1 Decadal trend in rice area across states in India (million ha)
Region/states 1970s 1980s 1990s 2000s 2010s Over all period
1970–71 to
2012–13
Northern region
Punjab 0.670 1.582 2.232 2.630 2.833 1.852
(12.456) (5.392) (2.491) (0.889) (4.403)
Haryana 0.339 0.545 0.830 1.062 1.235 0.731
(6.636) (2.396) (6.147) (2.008) (3.835)
Uttar Pradesh 4.700 5.332 5.602 5.691 5.822 5.365
(1.339) (0.027) (0.822) (−0.441) (0.577)
Central region
Madhya Pradesh 4.611 4.945 4.925 1.637 1.714 3.867
(0.877) (0.358) (7.386) (−1.329) (−2.938)
Maharashtra 1.402 1.505 1.524 1.522 1.536 1.491
(1.686) (−0.092) (−0.72) (−0.020) (0.239)
Eastern region
Bihar 5.265 5.189 4.992 3.436 3.141 4.610
(0.508) (0.252) (0.137) (−0.859) (−1.342)
Odisha 4.476 4.259 4.634 4.437 4.091 4.426
(−0.910) (0.405) (−1.33) (−0.016) (−0.056)
West Bengal 5.138 5.299 5.859 5.774 5.278 5.500
(0.050) (1.120) (0.587) (−0.025) (0.325)
Chhattisgarha 3.720 3.751 3.733
(−1.093) (0.163)
Western region
Gujarat 0.451 0.505 0.605 0.668 0.776 0.572
(0.266) (0.406) (2.024) (2.937) (1.413)
Southern region
Tamil Nadu 2.646 2.153 2.181 1.848 1.832 2.180
(0.234) (−2.018) (0.210) (0.210) (−0.975)
Karnataka 1.165 1.149 1.412 1.381 1.4 1.285
(0.699) (0.376) (3.452) (1.367) (0.691)
Andhra Pradesh 3.499 3.727 3.827 3.672 4.080 3.709
(1.910) (0.539) (0.394) (1.181) (0.253)
All India 38.639 40.654 43.201 43.404 43.081 41.586
(0.876) (0.413) (0.654) (−0.017) (0.369)
Note Figures in parentheses indicate estimated compound growth rates
1970s: 1970–71 to 1979–80; 1980s: 1980–81 to 1989–90; 1990s: 1990–91 to 1999–2000
2000s: 2000–01 to 2009–10; 2010s: 2010–11 to 2012–13
a Data not available for initial years as state was not formed
Compound growth rates for 2010s not worked out due to limited data
50
40
lakh ha
30
20
10
0
Tamil Nadu Karnataka Andhra Pradesh
Fig. 2.1 Trend in area of rice in Southern India
60
50
40
lakh ha
30
20
10
0
1970-71
1972-73
1974-75
1976-77
1978-79
1980-81
1982-83
1984-85
1986-87
1988-89
1990-91
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13
Madhya Pradesh Maharashtra
Fig. 2.2 Trend in area of rice in central India
70
60
50
lakh ha
40
30
20
10
0
Bihar Orissa West Bengal
Fig. 2.3 Trend in area of rice in eastern India
growth in Madhya Pradesh over the period of four decades. It is evident that the
area under rice in Madhya Pradesh has declined from 4.61 million hectares during
1970s to 1.714 million hectares during 2010s. It is mainly because of bifurcation of
Chhattisgarh during 2000 from Madhya Pradesh. As far as eastern India is concerned,
slight increase in area under rice is visualized in West Bengal, whereas the other states
like Bihar and Odisha have witnessed a decline in rice area.
9
8
7
lakh ha
6
5
4
3
2
1
0
Gujarat
Fig. 2.4 Trend in area of rice in western India
70
60
lakh ha
50
40
30
20
10
0
Punjab UƩar Pradesh Haryana
Fig. 2.5 Trend in area of rice in northern India
It is apparent that all states in northern India have witnessed a high growth rate in
area during the period from 1970–71 to 2012–13. The states like Punjab and Haryana
have registered a positive compound growth rate in rice area. Significant increase in
area under rice is seen in Punjab and Haryana. For instance, rice area in Punjab has
increased from 0.67 million hectares during 1970s to 2.833 million hectares during
2010s.
2.2.2 Trend in Rice Production
Rice production in India has witnessed a significant progress over the years. Total rice
production in the country has increased from 44.76 million tonnes during 1970s to
100.13 million tonnes during 2010s. More than doubling of rice production has been
achieved mainly because of Green Revolution and wider adoption of technologies
like modern rice varieties, HYV seeds, fertilizers and plant protection chemicals. At
all India level, the growth rate in rice production was 2.31% per annum (Table 2.2).
Table 2.2 Decadal trend in rice production across states in India (million tonnes)
1970s 1980s 1990s 2000s 2010s Over all period
1970–71 to
2012–13
Northern region
Punjab 1.675 4.925 7.430 10.000 9.539 6.253
(18.500) (6.740) (2.502) (2.670) (5.759)
Haryana 0.697 1.383 2.168 3.083 3.421 1.943
(11.280) (2.249) (4.429) (4.021) (4.931)
Uttar Pradesh 4.044 7.211 10.860 11.464 13.311 8.737
(1.644) (5.674) (3.057) (0.086) (3.408)
Central region
Madhya 3.302 4.292 5.287 1.485 2.777 3.534
Pradesh (−3.064) (2.032) (4.956) (1.573) (−1.739)
Maharashtra 1.741 2.176 2.322 2.416 3.078 2.227
(6.932) (−45.486) (0.499) (1.466) (1.318)
Eastern region
Bihar 4.678 5.273 6.318 4.569 6.011 5.265
(0.296) (4.135) (4.938) (−2.278) (0.222)
Odisha 3.890 4.555 6.287 6.327 7.866 5.446
(−1.435) (3.978) (−4.890) (4.680) (1.748)
West Bengal 6.358 8.149 11.797 14.497 13.475 10.446
(0.779) (6.828) (−0.015) (0.785) (2.605)
Chhattisgarha 4.742 7.116 5.759
(−12.968) (7.663)
Western region
Gujarat 0.469 0.624 0.903 1.133 1.618 0.840
(4.140) (−0.241) (2.926) (10.179) (3.414)
Southern region
Tamil Nadu 5.182 5.106 6.910 5.353 4.924 5.587
(0.477) (3.863) (1.899) (−0.838) (0.227)
Karnataka 2.048 2.238 3.507 3.594 3.546 2.895
(1.439) (0.236) (5.954) (2.696) (1.923)
Andhra 5.577 8.012 9.726 10.414 11.546 8.649
Pradesh (4.202) (2.514) (1.640) (5.675) (1.843)
All India 44.759 59.775 80.338 89.166 100.127 70.715
(1.901) (3.616) (2.031) (1.591) (2.312)
1970s: 1970–71 to 1979–80; 1980s: 1980–81 to 1989–90; 1990s: 1990–91 to 1999–2000
2000s: 2000–01 to 2009–10; 2010s: 2010–11 to 2012–13
160
140
lakh tonnes
120
100
80
60
40
20
0
Tamil Nadu Karnataka Andhra Pradesh
Fig. 2.6 Trend in production in southern India
It is evidenced that all the regions of rice growing areas in the country have
distinct behavioural pattern and production conditions, which had implications on
overall production. It could be observed that there has been a tremendous increase
in rice production across regions of India. In Southern India, a slight decline in
production (0.429 million tonnes) was observed in Tamil Nadu during 2010s as
compared with the earlier period. There was a decline of 1.557 million tonnes during
2000s when compared to 1990s. The other states, viz. Karnataka and Andhra Pradesh,
have registered a positive growth in rice production with a significant increase in
production of rice.
In central India, Maharashtra has registered a significant and positive growth
in rice production. However, Madhya Pradesh showed a decline in growth rate of
1.74% mainly due to bifurcation of Chhattisgarh from Madhya Pradesh. Similarly,
the rice production has witnessed a significant increase in the states of eastern and
northern India. Among the states of eastern India, Odisha witnessed a steady growth
in production from 3.890 million tonnes during 1970s to 7.866 million tonnes during
2010s. But the most interesting is the unprecedented growth of 13.47 million tonnes
recorded by West Bengal during 2010s. The compound annual growth rate of different
states rapidly increased during this period (Figs. 2.6, 2.7, 2.8, 2.9 and 2.10).
2.2.3 Trend in Rice Productivity
Doubling of rice productivity has been achieved in the country over a period of four
decades of development owing to Green Revolution technologies. The productivity
level has increased from 1156 kgs/ha during 1970s to 2357 kgs/ha during 2010s
(Table 2.3). Though there is a consistent increase in productivity not only in the
country but also in all the states, there are significant variations across regions and
states. The average rice productivity of states like Punjab (3199 kgs/ha), Tamil Nadu
(2627 kgs/ha), Haryana (2556 kgs/ha), Andhra Pradesh (2376 kgs/ha), Karnataka
70
60
lakh tonnes
50
40
30
20
10
0
Madhya Pradesh Maharashtra
Fig. 2.7 Trend in production of rice in central India
180
160
140
lakh tonnes
120
100
80
60
40
20
0
Bihar Orissa West Bengal
Fig. 2.8 Trend in production of rice in eastern India
20
18
16
lakh tonnes
14
12
10
8
6
4
2
0
1970-71
1972-73
1974-75
1976-77
1978-79
1980-81
1982-83
1984-85
1986-87
1988-89
1990-91
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13
Gujarat
Fig. 2.9 Trend in production of rice in western India

160
140
lakh tonnes
120
100
80
60
40
20
0
1970-71
1972-73
1974-75
1976-77
1978-79
1980-81
1982-83
1984-85
1986-87
1988-89
1990-91
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13
Punjab UƩar Pradesh Haryana
Fig. 2.10 Trend in production of rice in northern India
(2236 kgs/ha) and West Bengal (1929 kgs/ha) had shown productivity levels higher
than the country’s average level of 1648 kgs/ha. Among the states, the Punjab stands
first in terms of productivity with 3199 kgs/ha followed by Tamil Nadu (2627 kgs/ha)
and Haryana (2556 kgs/ha).
India registered annual CGR of 1.95% in yield of rice for the period from 1970–71
to 2012–13. Though Indian agriculture has recorded a positive growth rate in rice,
there are considerable variations in pattern of growth across regions. The state-wise
analysis shows that except Karnataka, all other states have registered positive annual
CGR over a period of four decades. Thus, one can speculate that though the area
under rice is declining over years, the positive growth in yield in most of the states
compensated the reduction in total rice production. The dominating yield effect
could be attributed to high level of input use, modern varieties, technologies and
agro-climatic factors. For instance, the increase in compounded annual growth rate
of rice in Punjab is mainly because of wider adoption of technologies such as semi-
dwarf high-yielding varieties and crop production technologies. The low productivity
performance at all India level indicates that there is a need to increase the average
yield to meet the ever-increasing demand.
The rice productivity in Tamil Nadu has registered a positive growth of 1.34%
per annum. The positive growth in yield is a major contributing factor for at least
maintaining the rice production in the state, around 5.5 million tonnes, over decades
in spite of a negative growth in area under rice in the state (Figs. 2.11, 2.12, 2.13,
2.14, 2.15).
2.3 Constraints in Rice Production
The rice production is subject to many constraints, and these limitations vary across
regions. The constraints are technical and socio-economic. The technical constraints
Table 2.3 Decadal trend in productivity of rice across states in India (kg/ha)
Region/states 1970s 1980s 1990s 2000s 2010s 1970–71 to
2012–13
Northern region
Punjab 2385.600 3092.500 3325.900 3797.600 3852.667 3199.395
(5.369) (1.278) (0.016) (1.763) (1.450)
Haryana 2007.600 2534.900 2640.900 2900.700 3031.667 2556.651
(4.354) (−0.150) (−1.641) (1.979) (1.145)
Uttar Pradesh 858.400 1349.900 1934.800 2010.600 2312.333 1592.419
(0.299) (5.646) (2.215) (0.529) (2.826)
Central region
Madhya 717.400 867.200 1101.400 913.000 1230.000 922.790
Pradesh −(3.896) (1.671) (−2.271) (2.907) (1.015)
Maharashtra 1229.000 1443.400 1591.600 1586.800 1852.667 1489.907
(5.153) (−0.579) (2.115) (1.494) (1.046)
Eastern region
Bihar 886.900 1010.800 1262.500 1319.700 1837.000 1170.000
(−0.214) (3.875) (4.788) (−1.435) (1.569)
Odisha 867.400 1064.500 1343.800 1422.600 1653.333 1207.977
(−0.523) (3.560) (3.725) (4.703) (1.666)
West Bengal 1238.000 1527.900 2209.600 2509.700 2704.333 1929.419
(0.829) (5.642) (4.203) (0.803) (2.435)
Chhattisgarha 1374 1669.667 1443.571
(−11.991) (4.801)
Western region
Gujarat 1017.800 1212.400 1438.478 1665.100 2052.333 1383.599
(3.873) (−0.647) (0.587) (6.997) (1.926)
Southern region
Tamil Nadu 1950.000 2396.500 3164.000 2864.600 3082.667 2627.884
(0.246) (6.003) (0.636) (−0.303) (1.338)
Karnataka 1834.300 1945.100 2437.600 2580.200 2733.000 2236.535
(1.526) (−0.136) (2.447) (−45.162) (1.199)
Andhra 1587.700 2140.500 2539.700 3020.500 3102.333 2376.535
Pradesh (2.251) (1.962) (1.224) (1.321) (1.988)
All India 1156.300 1467.100 1858.200 2052.300 2357.667 1648.00
(1.013) (1.371) (1.371) (1.606) (1.948)
1970s: 1970–71 to 1979–80; 1980s: 1980–81 to 1989–90; 1990s: 1990–91 to 1999–2000
2000s: 2000–01 to 2009–10; 2010s: 2010–11 to 2012–13
kg/ha kg/ha kg/ha
0
500
1000
1500
2000
2500
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0
500
1000
1500
2000
2500
3000
3500
4000
1970-71 1970-71 1970-71
1972-73 1972-73 1972-73
1974-75 1974-75 1974-75
1976-77 1976-77 1976-77
1978-79 1978-79 1978-79
1980-81 1980-81 1980-81
2.3 Constraints in Rice Production
1982-83 1982-83 1982-83

Tamil Nadu
Bihar
1984-85 1984-85 1984-85
1986-87 1986-87 1986-87
Fig. 2.12 Trend in yield of rice in central India
Fig. 2.13 Trend in yield of rice in eastern India

1988-89 Fig. 2.11 Trend in yield of rice in southern India
Madhya Pradesh
1988-89 1988-89
1990-91 1990-91 1990-91
Orissa
Karnataka
1992-93 1992-93 1992-93

1994-95 1994-95 1994-95
1996-97 1996-97 1996-97
1998-99 1998-99 1998-99
Maharashtra
2000-01 2000-01 2000-01
West Bengal
2002-03 2002-03 2002-03
Andhra Pradesh
2004-05 2004-05 2004-05

2006-07 2006-07 2006-07
2008-09 2008-09 2008-09
2010-11 2010-11 2010-11
2012-13 2012-13 2012-13
17
2500
2000
kg/ha
1500
1000
500
0
1970-71
1972-73
1974-75
1976-77
1978-79
1980-81
1982-83
1984-85
1986-87
1988-89
1990-91
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13
Gujarat
Fig. 2.14 Trend in yield of rice in western India
4500
4000
3500
3000
2500
kg/ha
Punjab
2000
UƩar Pradesh
1500
Haryana
1000
500
0
1990-91
1970-71
1972-73
1974-75
1976-77
1978-79
1980-81
1982-83
1984-85
1986-87
1988-89
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13
Fig. 2.15 Trend in yield of rice in northern India
include pest and diseases, soil, agronomic, physiological factors, etc. The socio-
economic constraints include small size of holding, inadequate supply of inputs,
non-availability of inputs, etc.
Studies on rice production across the states however revealed that the production
is constrained by many factors like socio-economic, biotic and abiotic factors. The
details of different constraints present in different regions are presented in Table 2.4.
It is evident that most of the constraints which limit rice production are location-
specific and vary across regions.
Andhra Pradesh: The major production constraints in the state are broadly clas-
sified into biotic and abiotic constraints (Cheralu, http://www.rkmp.co.in).
Biotic stresses: The insect pests attacking the rice crop are brown plant hopper
(BPH), gall midge, stem borer, cutworm, leaf folder, Gundhi bug and rice hispa. The
pests like panicle mite and thrips are newly emerging pests of significance in rice.
2.3 Constraints in Rice Production 19
Table 2.4 Major rice production constraints across states in India

Constraints States/agro-ecologies
Smallholder-dominant agriculture with poor Mostly eastern states of India
resource base, suboptimal level of input use
Erratic rainfall with poor soils Madhya Pradesh, Odisha and some parts of
Uttar Pradesh
Floods, water logging due to poor drainage West Bengal, north Bihar and eastern Uttar
Pradesh
Poor adoption of modern and high-yielding Mostly eastern states
rice varieties
Suboptimal level of fertilizer use, imbalanced North-eastern and eastern states
use of nutrients
Late sowing and lengthy transplanting period Mostly rainfed lowlands
due to monsoon failure and water scarcity
Low and poor adoption of crop production Mostly in uplands and lowlands
technologies
Saline and alkali soils West Bengal, Odisha, Andhra Pradesh, Tamil
Nadu, Kerala, Karnataka, Maharashtra,
Gujarat, Western Uttar Pradesh, Punjab,
Haryana, etc.
Source http://www.rkmp.co.in
BPH is a dominant pest of rice specifically in coastal Andhra Pradesh. Recently,

it has become widespread throughout the state. In addition to BPH, white backed
plant hopper (WBPH) incidence also observed and they occur in mixed population
and causing severe damage (hopper burn) to the rice crop. It is estimated that the
yield loss due to the WBPH reported to be 10 to 100%. Recently, panicle mite has
become an important constraint. Losses caused by the weeds are reduction in yield,
loss of nutrients from the soil, poor quality of produce, increased incidence of pests
and diseases, competition for resources and increased cost of cultivation. Estimated
yield losses due to weeds in low lands are 10–30%, while it is 40–60% in case of
rainfed rice.
The abiotic factors that limit rice yields are severe drought, water logging, temper-
ature, zinc deficiency, salinity and so on. In addition, the socio-economic constraints
like illiteracy, less inclination to know the latest technologies, labour scarcity, poor
resource base and investment form other important constraints.
Karnataka: Poor adoption of modern varieties, poor water management, use of
imbalanced nutrition, sudden outbreak of pest and diseases in irrigated areas and
moisture stress and poor nutrient uptake in rainfed areas are the major constraints
that limit rice production in the state (Rajanna, http://www.rkmp.co.in).
Tamil Nadu: Rice is predominantly grown under irrigated conditions. In some
pockets, it is also grown under rainfed conditions. The rainfed rice (upland rice)
often subjected to severe drought or high rainfall. In addition, soils in upland rice
cultivating areas may not be suited well to rice cultivation and hence led to poor
fertilizer–yield response (Thiyagarajan and Kalaiyarasi, http://www.rkmp.co.in).
It is found that the major rice production constraints in the state are: (i) non-
availability of HYV seeds which led to cultivation of traditional rice varieties, (upland
and rainfed lowland), (ii) poor fertility status of soils due to wind and water erosion,
(iii) high incidence of pest and diseases, weeds, etc., (iv) poor germination percentage,
poor population and uneven growth, (v) monsoon failure and consequent delayed
sowing and lengthy transplanting period.
Shanmugam et al. (2007) found that the top ten rice yield constraints which limit
rice production in the state are leaf folder, yellow stem borer, drought, blast, ear head
bug, zinc deficiency, RTV, sheath rot, BPH and bacterial leaf blight. The yield loss
due to all these constraints are estimated to be 11.40, 9.35, 7.84, 7.14, 7.12, 5.26,
5.20, 4.63, 4.31 and 4.01%, respectively, which account for a total production loss
of 1.7 million tonnes.
Madhya Pradesh: The rice production in Madhya Pradesh is limited by various
biotic and abiotic constraints. The important diseases include leaf blast, brown spot,
bacterial blight, sheath blight, false smut and seed discoloration, and the pests are
leaf folder, case worm, Gundhi bug, WBPH, BPH, Stem borer etc. The climatic fac-
tors such as poor and erratic rainfall and high temperature are the major constraints.
The critical socio-economic constraints include suboptimal and imbalanced use of
fertilizers, interrupted power supply to agriculture and inadequate supply of fertiliz-
ers. Insufficient research to address the major rice production constraints and poor
linkages with the development departments hinder rice production (Rao, http://www.
rkmp.co.in).
Maharashtra: In Maharashtra, rice is grown mainly in four major regions, namely
Konkan, Marathwada, Western Maharashtra and Vidarbha. Among the four regions,
the Konkan region has witnessed maximum productivity contributing around 43%
of rice production in the state. Though Vidarbha region has larger extent of area
under rice in the state, its productivity is very low when compared to other regions
(Thaware et al. http://www.rkmp.co.in).
Rice production in the state is constrained by many diseases and pests. Rice blast,
bacterial leaf blight, sheath blight and false smut are the major diseases causing yield
loss in rice. Similarly, the major pests are yellow stem borer, BPH, leaf folder, army
worm and gall midge. The other important constraints are salinity, poor adoption of
HYVs, low soil fertility, inefficient water management, suboptimal level of fertilizer
use, dry spells at critical stages and poor resource base of farmers. The eastern part
of Vidarbha suffers heavily due to erratic rainfall particularly during kharif season.
Uttar Pradesh: The rice growing seasons in the state are classified as wet season
(kharif), winter season (boro) and summer season (zaid). Kharif is the main rice
growing season in the state, and more than 98% rice is cultivated during this season
covering short-, medium- and long-duration varieties.
The average productivity of rice in the state is low (1592 kgs/ha) which is even
lower than the national average. This might be due to the fact that major rice area
is under rainfed. The frequent occurrence of various biotic and abiotic constraints
is also responsible for the low productivity. Rice blast, bacterial leaf blight, brown
2.3 Constraints in Rice Production 21
spot, sheath blight and neck blast are the major rice diseases that cause significant
yield loss. Similarly, pink stem borer and yellow stem borer are the major pests
which cause around 40% of yield loss in rice. The other insects including leaf folder,
white grub, Gundhi bug and termite also limit the rice yield in the state. In rainfed
conditions, weeds and moisture stress are the major constraints limiting rice yields.
The nutrient deficiency such as zinc, iron and boron and iron toxicity limit the higher
rice yields.
The other factors which constrain higher rice yield in the state are deteriorating
soil health conditions due to imbalanced use of fertilizers, inadequate application of
organic manures, secondary and micro-nutrients, low seed replacement rate, lack of
farm mechanization, inefficient utilization and improper water management, small
size of holding, and inefficient extension mechanism for transfer of technologies
(Dwivedi, http://www.rkmp.co.in).
Odisha: In Odisha, the biotic constraints such as brown plant hopper, white backed
plant hopper, stem borer, gall midge, sheath rot and bacterial leaf blight are found
to be most yield reducing constraints. It is also found that lack of HYVs, resistance
to stem borer and bacterial leaf blight and tolerance to water logging conditions are
the major constraint to higher rice yields in the state (Das http://www.rkmp.co.in).
West Bengal: Rice is an important crop in West Bengal, cultivated in different
production environments such as different soil types, water regimes, water logging
conditions and so on. In general, there are four major types of production environ-
ments in which rice is grown. They include irrigated, rainfed (lowland and upland),
coastal saline and flood-prone areas. On an average, the yield gap II, estimated as
1.32 tons/ha, is due to various technical and socio-economic constraints. The major
constraints causing the yield gap are: (i) submergence and droughts at early stages
of plant growth (ii) important pests like stem borer, brown plant hopper, leaf folder
and green leafhopper, (iii) plant diseases such as bacterial leaf blight, brown spot
and sheath rot and (iv) soil salinity and weeds as the other constraints. In upland
rice-cultivating areas, droughts, stem borer, green leaf hopper, blast, brown spot,
acid soils and weeds limit higher rice yields (Adhikari et al. http://www.rkmp.co.in).
Chhattisgarh: The state is not exempted from the constraints that limit rice pro-
duction. The common constraints prevalent in the state are poor and erratic rainfall,
low pH, losses of nutrients particularly nitrogen and potassium due to leaching,
poor fertility status, diseases (blast and brown spot), pests like leaf hopper, Gundhi
bug, termite, rootknot nematodes and so on. The other constraints which limit rice
production in the state are non-availability of high-quality seeds, poor adoption of
technologies, etc. (Pandey et al. http://www.rkmp.co.in).
Punjab: Though Punjab state is leading in rice productivity and production among
states of India, there is scope for improvement in productivity. The rice production
in the state is also affected by different constraints like pests (yellow stem borer, leaf
folder, white backed plant hopper, brown plant hopper) and diseases (sheath blight,
sheath rot, false smut, foot rot and brown spot) (Rang et al. http://www.rkmp.co.in).
Gujarat: The major diseases which limit to higher rice yields are bacterial leaf
blight (BLB), rice blast, grain discoloration and sheath rot. Rice stem borer, brown
plant hopper, WBPH, leaf roller and wireworm are the major insect pests in the state.
Weed infestation particularly in upland drilled rice and rice nurseries also are found
to limit rice production in the state (Pathak et al. http://www.rkmp.co.in).
2.4 Climate Change and Rice Production
Climate change and its impacts have received much attention in recent decades. It
is expected that the climate change may result in significant negative impacts on
growth and development, poverty alleviation and achievement of the Millennium
Development Goals (MDGs). Consistent efforts taken globally to achieve the devel-
opmental objectives may be thwarted by climate change. It is argued by many that
the climate change may increase risk, particularly for those who depend on weather
patterns, agriculture, water and other natural resources for their livelihoods (El-Ashry
2009). It is found that increase in temperature in India could reduce farm net rev-
enues by 9–25% (http://www.eldis.org/static/DOC11253.htm). Climate variability
and change add to the vulnerability in general and those who are already poor in
particular. With the climate change on its way, even though mitigation measures are
taken up by both Central and State governments, there is a dire need for adapting to
the changes like increase in average temperature, aridification of the dry regions and
more extreme floods and droughts (Hoff 2004).
Both the Central and State governments have been implementing various adap-
tive measures. They include policy measures (watershed development programmes,
bringing water from the neighbouring states), technological and structural measures
(demand-side management of water technologies like drip and sprinkler irrigation,
system of rice intensification), change of use activity or location (changes in land
use like shifting towards rainfed cultivation, farm diversification such as inclusion of
livestock), monitoring and forecasting (weather forecasts, automatic weather fore-
cast station) and risk sharing and spreading (crop insurance, weather-based crop
insurance, micro-insurance). In order to prepare for the climate variability, each pro-
duction environment will require a specific set of adaptation measures addressing
various scales of intervention and vulnerable groups.
The climate challenge facing Indian agriculture needs to be addressed in a proper
way. It is evident that the annual mean rainfall and also that during monsoon peri-
ods decline over the years and the variability in rainfall is also increasing. Another
threatening issue is an increase in temperature as well (Table 2.5).
The changes in temperature, radiation, rainfall and CO2 levels can affect the yield
of rice through their direct effect as well as indirect effects such as higher incidence
of pests and diseases, and availability of water for irrigation.
A time series analysis of trend in rice yield in the Indo–Gangetic plains shows
a declining trend, and this is partly related to the gradual change in weather condi-
tions in the last two decades (Agarwal 2008). The simulation analysis indicated that
irrigated rice is likely to lose the yields up to 23% in Upper Ganga Basin. Gupta
et al. (2012) observed that climate change is likely to reduce the yields of rice pro-
ductivity in 16 major agriculture-intensive states of India. It is also found that a 4°
2.4 Climate Change and Rice Production 23
Table 2.5 Details on agro-climatic and weather parameters

1951/52 to 1967/68 1968/69 1981/82 1991/92 1997/98 2002/03 2007/08
to to to to to to
1980/81 1990/91 1996/97 2001/02 2006/07 2011/12
Annual rainfall (cm)
Mean 122.5 118.7 120.1 121.0 118.5 113.7 111.7
Std. 12.5 10.2 11.5 7.2 8.3 9.4 10.0
deviation
Coefficient 10.20 8.59 9.58 5.95 7.00 8.27 8.95
of varia-
tion (%)
Monsoon rainfall (cm)
Mean 91.9 88.8 88.8 90.0 87.8 83.9 86.6
Std. 10.1 9.6 11.0 6.5 5.5 7.9 9.7
deviation
Coefficient 10.99 10.81 12.39 7.22 6.26 9.42 11.20
of varia-
tion (%)
Annual temperature anomaly from normal (°C)
Mean 0.04 −0.03 0.09 0.19 0.34 0.56 0.65
Std. 0.28 0.24 0.03 0.10 0.22 0.11 0.26
deviation
Coefficient 700.00 −800.00 33.33 52.63 64.71 19.64 40.00
of varia-
tion (%)
Source Climate bulletins and other publications of the India Meteorological Department
Government of India, Twelfth Five Year (2012–2017), Economic Sectors, Volume-II, Government
of India, P. 3
rise in temperature will lead to 41% reduction in rice yields (Geethalakshmi et al.
2011). Nandhini et al. (2006) mentioned that rice cultivable land has declined due
to scarcity of inputs and scanty rainfall in Tamil Nadu (India). Similarly, Hundaland
Kaur (2007) found that 1° increase in minimum temperature leads to a reduction in
yield of rice by 3% in Punjab (India). Saseendran et al. (2000) found that change
in temperature up to 5 °C can lead to a continuous decline in the yield of rice and
every one degree increase in temperature will lead to a 6% decline in yield of rice in
Kerala. Haris et al. (2010) found that rice production may decline by 31% in 2080
due to climate change in Bihar.
Researchers found that the grain yield of rice will decline by 10% for every 1 °C
increase in temperature above 32 °C (Pathak et al. 2003). The climate change impact
on the productivity of rice in Punjab (India) has shown that with all other climatic
variables remaining constant, temperature increase of 1, 2 and 3 °C would reduce
the grain yield of rice by 5.4, 7.4 and 25.1%, respectively (Aggarwal et al. 2009).
2.5 Summary
This chapter examined the trend in area, production and yield of rice in major rice
growing states of India. From the analysis, it is found that there are variations in area,
yield and production of rice across regions and states. The overall growth of rice area
during the four decades was 0.4%, the growth in rice productivity was 1.95%, and the
growth in rice production was 2.36%. Also, there exists significant yield gap in rice
production across states and the rice production is constrained by many technical,
socio-economic and institutional constraints. In the case of climate change impact,
raise in temperature in the crop season has reduced the rice yield.
Significant policy supports will help minimize the yield gap and increase both pro-
ductivity and production of rice in the long run. Hence, the policy focus may be tilted
towards developing and managing the irrigation infrastructure, access to institutional
credit, supply of inputs, adoption of technologies particularly use of high-yielding
variety seeds, integrated nutrition and pest management, water-saving methods, cli-
mate smart rice management practices, development of market infrastructure and
adequate price supports.
Appendix
Appendix
States Yield gap and its Constraints Reasons Impact of technologies Ruling varieties/ Strategies to increase
reasons recommended the production of rice
varieties crop
Southern
Tamil Nadu a. Maximum Saline, acidic soils Difficulty in removal of salts Encouraging the use of soil IR-64, CO-47, Adoption of SRI
Attainable Yield is and alkali soils in the coastal region ameliorants for improving saline, ADT-36, ADT-37, cultivation
the rice yields of Knowledge gaps Toxicity due to high pH and alkaline and acidic soils ADT-43, ADT-45, Creating awareness
experimental/on-farm Use of traditional due to the presence of sodium Potential of conservation ADT-47, ADT-48, and motivating hybrid
plots with no varieties either as carbonate or as agriculture including zero or ASD-16, ASD-17, rice cultivation
physical, biological Poor and imbalanced bicarbonate in the alkali or minimum tillage, direct-seeded ASD-20 through
and economic use of fertilizers saline–alkali soils. Highly rice, precision agriculture, demonstration
constraints and with Poor adoption of crop dispersed soil under alkaline site-specific nutrient management Improving soil
the best-known production or saline–alkali situation and fertility through crop
management practices technologies drainage is a problem rotation
at a given time and Lack of awareness about Encouraging balanced
ecology high-yielding varieties and use of fertilizers and
b. Farm-Level Yield hybrid rice technology plant nutrients
is the average yield of Continuous use of traditional Adoption of improved
farmers (southern) varieties due to the water management
obtained in a given non-availability of seeds and practices
target area at given farmers lack of awareness Creating awareness
time and ecology about high-yielding varieties through
(upland, rainfed lowland and demonstrations,
deep-water areas) farmer’s field school
Poor adoption of improved Attractive Market
crop production technology Price and Insurance
due to economic coverage
backwardness of the farmers
(continued)
25
26
(continued)
varieties crop
Karnataka Erratic rainfall Integrated nutrient management MAS 946-1 Transfer of recent
Poor in adoption of (organic manures + biofertilizers MAS 26, Rasi, technologies on rice
modern varieties + chemical fertilizers) Farmer grown locals production through
Global climatic Enhance water productivity Doddabairanelli proper extension
changes through wider adoption of SRI methods like FLDs,
Poor water Aerobic rice cultivation in Video conferencing,
management, use of water-limited areas like tail-end Rice Knowledge
imbalanced nutrition, areas of the canal, tank-fed areas Management Portals
sudden outbreak of and borewell-irrigated areas if, the and extension
pest and diseases in cultivation of rice is inevitable services supported by
irrigated areas and Aerobic rice cultivation Mobile phone based
moisture stress and A new mid-early, high-yielding MMSs and SMS
poor nutrient uptake variety MAS 946-1 is recently Enhance water
in rainfed areas released by the University for productivity through
Aerobic Cultivation wider adoption of
SRI, Aerobic and
AWD method of
cultivation in
water-limited areas
like tail-end areas of
the canal, tank-fed
areas and bore
Well irrigated areas if,
the cultivation of rice
is inevitable
(continued)
2 Rice Production in India: Analysis of Trend …
Appendix
(continued)
varieties crop
Andhra Though farmers adopt Genetic improvement BPT-5204, Kavya Collection and
Pradesh HYVs, often they of rice yield and MTU maintenance of
ignore adoption of strengthening of germplasm bank
other crop production quality seed Identification of
technologies and production suitable areas for
solving other related Promoting hybrid rice hybrid rice seed
problems cultivation production in the state
Cyclones and floods Stabilizing yield Ensuring proper
are the frequently through molecular transfer of technology
occurring constraints breeding Promotion of
which limit rice Improving resource vermicompost, green
production in the state use of water manuring, application
Water-saving and of gypsum, zinc
yield-enhancing sulphate etc.
strategies Educating the farmers
through capacity
building on improved
crop management
practices likeINM,
IPM, Post harvest
technology, Water
management
(continued)
27
28
(continued)
varieties crop
Central
Maharashtra Reduction in soil Ratnagiri 24, Promotion of
organic matter Sahyadri—2, Integrated Nutrient
Widening NPK Ratio Sahyadri—3 Management
Micro-nutrient Adoption of
deficiencies Conservation
Declining Agriculture Practices
yield-fertilizer Optimal use fertilizers
response in terms of Evolution of
grains/kg fertilizer high-yielding,
Heavy metal pollution semi-dwarf, fertilizer
Low fertilizer use responsive rice
efficiency varieties with
different duration and
grain types suitable
for different
ecosystem
Developing
nutrition-rich rice
varieties
Developing pest and
disease resistant
varieties
(continued)
(continued)
Appendix

varieties crop
Madhya Poor adoption of Mahamaya, Ratnagiri Creating awareness
Pradesh hybrid rice 3 through campaigns
Poor adoption of and demonstration on
management practices SRI, aerobic rice etc.
led to only marginal Promoting wider
increase in hybrid rice adoption of hybrid
and significant rice under SRI
variations across method and proper
regions water management
Higher seed cost practices
Poor
Non-availability of
quality seeds in time
and at reasonable
price
Eastern
Bihar Erratic and uneven Suboptimal level of PA 6444, PHB 71, Promotion of optimal
distribution of rainfall input use Pusa RH 10, KRH 2, use of different inputs
Poor extension Rajlaxmi, Sahyadri, through development
services DRRH programmes and
Severe weeds Rajashree, Sita, interventions
Rajendra Mahsuri, KVKs may be
Prabhat encouraged to ensure
proper transfer of
technologies and
adoption of various
crop production
technologies
(continued)
29
(continued)
30
varieties crop
Odisha Erratic and uneven Poor marketing and Rajalaxmi, KRH-2, Soil amelioration by
distribution of rainfall credit facilities, poor PHB-71, PA 6201, PA lime (7 lakh ha).
Non-availability of adoption of modern 6444, PA 6129 and Paper mill sludge for
quality seeds and rice production IR 64, IR 36, Pooja, acidic soil treatment
fertilizers in time and technologies CR 1018, Savitri, in Odisha
at reasonable prices Ratna, Khitish, Minikits of saline
Annada, CR 1017, tolerant varieties
Lunishree, CR 1014 More use of quality
seeds
Promoting farm
mechanization
wherever possible
Developing propoer
post-harvest
technologies and
infrastructure
West Bengal Yield gap II is not Major rice production Development of location-specific Khitish, Satabdi, Developing
able to manage by constraints are floods rice production technologies Sabita, Shashi, Bipasa post-harvest
farmers and drought in Evolving varieties for drought- technologies
addition to other tolerant/flood-tolerant/salt-tolerant Promotion of
biotic and abiotic high-yielding rice varieties for public-private
constraints flood-prone upland, salinity entrepreneurship
affected coastal land and Promoting Seed
flood-prone lowland ecosystems, Village’ programme
respectively Adoption of low-cost,
Adoption of rain water resource conserving
harvesting structures eco-friendly
Adopting IPNS technique technologies
Encouraging organic farming and
biofertilizers application Page|
(continued)
Appendix
(continued)
varieties crop
Chhattisgarh Poor and imbalanced Water stress and Light texture of soil, shallow IR 64, IR 36, Adoption of improved
use of fertilizers flooding, poor soil depth, un-bounded Mahamaya, Kranti, crop production
drainage Sloppy and undulating Purnima, Swarna, technologies and
Suboptimal use of topography MTU 1010, efficient water
fertilizers and plant Accumulation of water in Danteswari, management
nutrients lowland areas creates hurdles Bamleshwari technologies
Socio-economic in adoption of
Technological management practices
Poor and uneven distribution
of rainfall
Resource poor farmers
Insecurity of land tenure
Inequality in land ownership
Labour scarcity,
non-availability of inputs in
time
Poor adoption of improved
crop production technologies
Non-availability of
high-quality seeds in time
Lack of suitable rice varieties
Lack of suitable machinery
(continued)
31
32
(continued)
varieties crop
Western
Gujarat Blast, Grain SRI method may be encouraged to GR-4, GR-5, GR-7, Assured canal water
Discoloration and ensure higher rice yields GR-12, NAUR-1, supply at the time of
Sheath Rot are the IR-28 raising nurseries and
major diseases in the during dry spell may
state. Whereas; Stem be ensured
Borer, Brown Plant Treatment of
Hopper, WBPH, Leaf seedlings with
roller and Wire worm biofertilizers prior to
are the major insect transplanting
pest in the Gujarat SRI and aerobic
Salinity and drought method of cultivation
are the major abiotic may be encouraged
constraints in rice Water harvesting in
cultivation rainfed rice
Evolvingsuitable
hybrid rice
Creating awareness
through circulation of
print outs on different
technologies,
conducting
demonstrations etc.
(continued)
(continued)
Appendix
varieties crop
Northern
Punjab Yield gap is due to Major constraints for Basmati 217, Basmati Adoption of improved
various technical and these ecosystems are 370, Basmati 385, land, crop water
socio-economic (i) Submergence and Basmati 386, Punjab management practices
constraints droughts at seedling Basmati 1, Super
stage (ii) stem borer, Basmati
brown plant hopper,
leaf folder and green
leafhopper as the
major insects; (iii)
bacterial leaf blight,
brown spot and sheath
rot as the major
diseases and (iv) soil
salinity and weeds as
the other constraints
Labour shortage
Haryana Fluctuating water CSR 30 and CSR 27 Adoption of improved
tables, salinity, Pusa Basmati 1, Pusa crop production
declining soil health Sugandh 4, Pusa technologies
and low crop Sugandh 5 and HBC
productivity 19
Taroari Basmati
Uttar Problem soils Suboptimal of input level of Adoption of INM Narendra Dhan Encouraging optimal
Pradesh Poor adoption of crop input use, unscientific crop Effective technology transfer 3112-1, Prakhar, use of fertilizers and
production rotation and poor use of NDR 2064, Narendra inputs
technologies organic manures Usar Sankar Dhan 3 Extension mechanism
Erratic rainfall, water poor adoption of modern may be strengthened
logging and floods technologies
Source Status paper on Rice for different states, Rice Knowledge Management Portal, Directorate of Rice Research, Rajendranagar, Hyderabad 500030. Field surveys
Visit Rice Knowledge Management Portal http://www.rkmp.co.in
33
Chapter 3
Climate Change Projections for Major
Rice Growing States of India
S. Senthilnathan
This chapter presents evaluation of the model baselines with respect to the observa-
tions. Further, it also describes the methodology and approach used for extracting
future climate change projections for the major rice growing states to predict the rice
production in India.
3.1 Selection of Global Climate Models for Climate Change

Projections
A detailed climate projection has been derived from the global coupled model results
now available from Coupled Model Intercomparison Project 5th Phase (CMIP5). The
5th IPCC Assessment Report (AR5) was based on analysis of CMIP5. Results of
multi-member ensembles of present-day and future climate change scenario simu-
lations from over 25 models run at leading modelling centres throughout the world
are now available. Modelling groups across the world have contributed data from
more than one model versions. Table 3.1 gives a list of the recent CMIP5 models with
daily and monthly data sets available (http://cmip-pcmdi.llnl.gov/cmip5/data_portal.
html). In CMIP5, a detailed diagnostics of the historical (current climate) simulations
performed by these models indicate that eight of them are capable of simulating the
regional distribution of mean monsoon precipitation and its spectrum of variabil-
ity over South Asia during south-west monsoon season (June through September)
The chapter was commissioned by Dr. Palanisami to add different angles to this book. The author
would like to thank Dr. Palanisami for providing substantial inputs in preparing this chapter, without
which, the chapter would not have taken the shape that it has now.
S. Senthilnathan (B)
Department of Agricultural Economics, Tamil Nadu Agricultural University,
Coimbatore, India
e-mail: [email protected]
36 3 Climate Change Projections for Major Rice Growing States of India
Table 3.1 List of GCMS participated in CMIP5 modelling groups

S. No. Modelling centre Institute ID Model name
1 Commonwealth Scientific CSIRO-BOM ACCESS1.0
and Industrial Research
Organization (CSIRO) and
Bureau of Meteorology
(BOM), Australia
2 Beijing Climate Center, BCC BCC-CSM1.1
China Meteorological
Administration
3 College of Global Change GCESS BNU-ESM
and Earth System Science,
Beijing Normal University
4 Canadian Centre for CCCMA CanESM2
Climate Modelling and CanCM4
Analysis CanAM4
5 National Center for NCAR CCSM4
Atmospheric Research
6 Community Earth System NSF-DOE-NCAR CESM1(CAM5)
Model Contributors
7 Centre National de CNRM-CERFACS CNRM-CM5
Recherches
Meteorologiques/Centre
European de Recherche et
Formation Avancees en
CalculScientifique
8 Commonwealth Scientific CSIRO-QCCCE CSIRO-Mk3.6.0
and Industrial Research
Organization in
Collaboration with
Queensland Climate
Change Centre of
Excellence
9 EC-EARTH Consortium EC-EARTH EC-EARTH
10 LASG, Institute of LASG-CESS FGOALS-g2
Atmospheric Physics,
Chinese Academy of
Sciences and CESS,
Tsinghua University
11 LASG, Institute of LASG-IAP FGOALS-gl
Atmospheric Physics, FGOALS-s2
Chinese Academy of
Sciences
12 The First Institute of FIO FIO-ESM
Oceanography, SOA,
China
(continued)
3.1 Selection of Global Climate Models for Climate Change Projections 37
Table 3.1 (continued)

S. No. Modelling centre Institute ID Model name
13 NOAA Geophysical Fluid NOAA GFDL GFDL-CM3a
Dynamics Laboratory GFDL-ESM2G
GFDL-ESM2 M
14 NASA Goddard Institute NASA GISS GISS-E2-H
for Space Studies GISS-E2-R
15 National Institute of NIMR/KMA HadGEM2-AO
Meteorological
Research/Korea
Meteorological
Administration
16 Met Office Hadley Centre MOHC HadCM3
HadGEM2-CCa
HadGEM2-ES
17 Institute for Numerical INM INM-CM4
Mathematics
18 Institute Pierre-Simon IPSL IPSL-CM5A-LR
Laplace IPSL-CM5A-MR
19 Japan Agency for MIROC MIROC-ESM
Marine-Earth Science and MIROC-ESM-CHEM
Technology, Atmosphere
and Ocean Research
Institute (The University
of Tokyo), and National
Institute for
Environmental Studies
20 Atmosphere and Ocean MIROC MIROC4 h
Research Institute (The MIROC5
University of Tokyo),
National Institute for
Environmental Studies,
and Japan Agency for
Marine-Earth Science and
Technology
21 Max Planck Institute for MPI-M MPI-ESM-LR
Meteorology MPI-ESM-P
22 Meteorological Research MRI MRI-CGCM3
Institute
23 Non-hydrostatic NICAM NICAM.09
Icosahedral Atmospheric
Model Group
24 Norwegian Climate NCC NorESM1-M
Centre
a Two models capable of simulating the regional features over South Asia were used for the present
study
(Sperber et al. 2013). Of them, the latest versions of the UK Meteorological Office
model HadGEM2_CC and GFDL_CM3 developed at Geophysical Fluid Dynamics
Laboratory, NOAA, Princeton, USA, can be used to get the future projections for
the major rice growing states in India. The simulations from these CMIP5 global
models are performed at a higher horizontal and vertical resolution compared to the
CMIP3 groups, an added value for understanding and assessing regional changes in
a warmer climate.
Monthly data for GFDL_CM3 and HadGEM2_CC for each of these GCMs from
CMIP5 website was downloaded. The major climate variables required for the study
included rainfall, maximum and minimum temperatures. We have selected the two
GCMs (GFDL_CM3 and HadGEM2_CC) which have good skill in simulating real-
istic characteristics of Asian monsoons (Annamalai et al. 2013; Cherchi et al. 2014),
and those models have better skill for simulating the future climate change scenarios.
The present approach considered the two selected global climate models (GCMs),
and rainfall and temperature climatologies constructed from the past 30 years of the
historical model simulations have been compared with observations. The robust-
ness has been assessed based on the global climate models ability in representing
the current climate regional features of our study regions. Hence, these models are
considered for their capability in capturing the future projections in the study regions.
3.2 Approach and Methodology for Extracting Climate

Data
The current and future climate change scenarios of rainfall, minimum and maximum
temperature changes were generated for the 13 major rice growing states of India
from the two selected GCMs. Figure 3.1 represents the schematic illustration for the
downscaling approach adopted to prepare current and climate change scenarios for
major rice growing states in India.
3.3 Spatial Domains of Major Rice Growing States in India
The demarcated spatial domain for the present study for the 13 states is given in
Table 3.2. The larger domain will enable to ensure climatological consistency for
the particular state which is used for impact studies. The monthly climate data, viz.
rainfall, maximum and minimum temperatures for the respective states, was extracted
for the current and future periods from the two different global climate models using
the domain given below.
3.3 Spatial Domains of Major Rice Growing States in India 39
Downscaling from CMIP5 GCMs
Selection of the best two GCMs

(Correlation and RMSE analysis with observations/reanalysis products)
GFDL CM2.1 HadGEM2-CC

Historical Climate-1971-2000 Historical Climate-1971-2000
Mid Century Scenario-2021-2050 Mid Century Scenario-2021-2050
End Century Scenario-2071-2100 End Century Scenario-2071-2100
Downscaled historical climate (Baseline data)
IMD observations
Validation of the global climate model

(GCM baseline data compared with IMD observations to
reflect the regional characteristic features)
Downscaled future projection

(Mid and End century data)
GCM data utilized to predict the future rice

production in India
Fig. 3.1 Schematic diagram for extracting current and future climate change scenarios’ data gen-
eration
Table 3.2 Spatial domain S. No. Name of the Longitude (°) Latitude (°)
used for the selected states to states
extract climate data
1 Andhra Pradesh 77–84.4 13–20
2 Bihar 83.3–88.2 24.5–27.5
3 Chhattisgarh 80.2–84.2 17.8–24
4 Gujarat 68.3–74.6 20–24.7
5 Haryana 74.5–78 27.7–31
6 Karnataka 74.2–78.6 11.5–18.5
7 Madhya Pradesh 74.1–82.7 21.1–26.6
8 Maharashtra 72.7–80.7 15.5–22
9 Odisha 81.8–87.5 18.0–22.5
10 Punjab 73.9–76.8 29.7–32.4
11 Tamil Nadu 76.5–80.3 8–13.6
12 Uttar Pradesh 77.2–84.6 24–30.5
13 West Bengal 85.8–89.1 21.5–26.5
3.4 Historical Observed Climate Data
Climate data relevant to the study area was collected from meteorological stations
located in the respective states. These observational data was collected from the
network stations maintained by India Meteorological Department (IMD), agricultural
universities and State Department of Agriculture. We have taken observational data
having long time series of nearly 30 years that is suitable for climatological trend
analysis.
3.4.1 Seasonality of Rainfall
Climate normal of a variable is represented as an average value, typically for a period

of 30 years. The seasonality can be expressed as the temporal evolution of rainfall
which is capable of illustrating the timing of the maximum and minimum during
a year. This is an effective indicator of climate of a particular state. The seasonal
cycle of rainfall from observations for major rice growing states in India is shown
in Fig. 3.2. It could be observed from Fig. 3.2 that rainfall peak is clearly observed
during the south-west monsoon season from June to September in all the states except
Tamil Nadu in which north-east monsoon is observed from October to December as
a major monsoon season. This indicates that the clear climatological feature has been
expressed by all the states from the observations. However, Karnataka receives the
maximum quantity of rainfall, and Uttar Pradesh receives the lowest rainfall among
the selected states.
3.4 Historical Observed Climate Data 41
1000
800
Rainfall (mm)
600
400
200
0
Jan Feb Mar Apr May Jun July Aug Sep Oct Nov Dec
AP Bih Cha Guj Har

Kar Mah MP Ori Pun
TN UP WB
Fig. 3.2 Seasonal cycle of rainfall in major rice growing states of India
3.4.2 Seasonality of Temperature
The seasonal cycle of maximum and minimum temperatures from observations for
major rice growing states in India is shown in Fig. 3.3. The highest values of max-
imum and minimum temperatures are encountered during the summer months of
April–May, while the lowest values are observed during winter months December—
January. This indicates that the maximum and minimum temperatures clearly brought
out by the seasonal evolution in all the states.
3.5 Evaluation of Baseline Climate with IMD Observations
For all climate models, one of the most important challenging aspects is the simulation
of rainfall climatology during monsoon seasons, viz. summer (June–September) and
winter (October–December) monsoon periods. A realistic simulation of the basic
state of monsoon rainfall climatology is a key feature, for assessing the future changes
due to anthropogenic forcing over the Indian monsoon region. The rainfall extracted
from the GCMs has been validated against observed rainfall climatology for all the
major rice growing states of India. Compared to CMIP3, higher horizontal resolutions
employed in CMIP5 models capture the topographically induced rainfall over the
monsoon region. Therefore, it is useful to compare the seasonal cycles of rainfall and
temperatures over the particular state. Comparing observations with model baseline
data will provide a good indicator of the model performance to simulate climate
change projections.
The selected CMIP5 models were evaluated for their ability to simulate precip-
itation and temperature patterns in the respective rice growing states in India. The
50
40
Tmax ( ° C)
30
20
10
0
AP Bih Cha Guj Har Kar Mah
MP Ori Pun TN UP WB
30
25
Tmin ( °C)
20
15
10
5
0
MP Ori Pun TN UP WB
Fig. 3.3 Seasonal cycle of temperature in major rice growing states of India
baseline period selected for the study regions was from 1971 to 2000. The future
time horizons up to 2100 have been analysed for this study for two future scenar-
ios, viz. RCP4.5 and RCP8.5, representing medium- and high-emission futures. The
future scenario selected for the period representing the mid- and end-century is,
respectively, from 2021 to 2050 and 2071 to 2100 (Table 3.3).
3.5.1 Rainfall Seasonal Cycle
Seasonal cycle of rainfall for major rice growing states in India is given in Fig. 3.4. In
all the states, the highest monthly normal rainfall was received during the south-west
monsoon (June–September) season except Tamil Nadu. The peculiarity of Tamil
Nadu rainfall seasonal cycle which is dominating influence of north-east monsoon
is clearly seen during the months from October to December (Fig. 3.4k). The states
3.5 Evaluation of Baseline Climate with IMD Observations 43
Table 3.3 Observational, baseline and future scenarios climate model data sets used in the study
regions
S. No. Product Parameters Scenarios Time period Frequency
1 Observed Rainfall (mm) Station 1971–2000 Monthly
climate data Maximum observatory
from IMD temperature (°C) located in
stations Minimum different
temperature (°C) places
2 GFDL-CM3 Rainfall (mm) Baseline 1971–2000 Monthly
Maximum Projection 2021–2050
temperature (°C) RCP4.5 and and
Minimum RCP8.5 2071–2100
temperature (°C)
3 HadGEM2-CC Rainfall (mm) Baseline 1971–2000 Monthly
Maximum Projection 2021–2050
temperature (°C) RCP4.5 and and
Minimum RCP8.5 2071–2100
temperature (°C)
like Bihar, Chhattisgarh, Haryana, Odisha, Punjab, Tamil Nadu and West Bengal
indicated that both the model rainfall has very closer agreement with the observations.
The climate model underestimates the rainfall as compared to observations in the
states like Gujarat, Karnataka, Madhya Pradesh and Maharashtra. This indicates that
the degree of uncertainty prevails in rainfall observations. However, both the climate
models predict the seasonal peak and magnitude of rainfall very well in all the states
compared with observations.
3.5.2 Temperature Seasonal Cycle
Maximum temperature seasonal cycle for major rice growing states in India is shown
in Fig. 3.5. It is observed that both the climate models capture the seasonal evolution of
maximum temperature with slight cold bias in almost all the states except Karnataka
which shows warm bias. It is observed that the monthly temperature evolution in all
the states of baseline temperature has good similarity with the IMD observations.
However, the magnitude of the maximum temperature is smaller than the rainfall. The
states like Chhattisgarh, Madhya Pradesh, Odisha, Andhra Pradesh and Maharashtra
have very close agreement between the model and observations compared to other
states.
Seasonal cycle of minimum temperature for major rice growing states in India is
given in Fig. 3.6. It is observed that both the climate models capture the timing of the
highest and lowest temperatures in annual seasonal cycle thus confirming the good
indicator of the model performance. In all the states, model minimum temperature
agrees well with the observations.
(a) Bihar (b) Chattisgarh

500 500
Rainfall (mm)
400
Rainfall (mm)
400
300 300
200 200
100 100
0 0
Jan
Feb
Mar
Apr
May
Jun
July
Aug
Sep
Oct
Nov
Dec
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
July
Aug
Sep
Oct
Months Months
Obs GFDL HadGEM2 Obs GFDL HadGEM2
(c) Gujrat (d) Haryana

700
250
600
Rainfall (mm)
Rainfall (mm)
500 200
400 150
300
100
200
100 50
0 0
Mar
May
July
Aug
Jan
Feb
Apr
Jun
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
July
Aug
Jun
Sep
Oct
Nov
Dec
Months Months
(e) Karnataka (f) Madya Pradesh

1000 600
800 500
Rainfall (mm)
Rainfall (mm)
400
600
300
400
200
200 100
0 0
Mar
Aug
Nov
Jan
Feb
Apr
May
Jun
July
Sep
Oct
Dec
Nov
Jan
Apr
May
Feb
Mar
Jun
July
Aug
Sep
Oct
Dec
Months Months
Fig. 3.4 Seasonal cycle of rainfall for major rice growing states in India: Evaluation of GCM
(GFDL_CM3 and HadGEM2_CC) baseline rainfall with observations
3.5 Evaluation of Baseline Climate with IMD Observations 45
(g) Maharastra (h) Orissa

400 500
Rainfall (mm)
400
Rainfall (mm)
300
300
200 200
100 100
0
0
Jan
Feb
Mar
Apr
May
Jun
July
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Dec
Jun
July
Aug
Sep
Oct
Nov
(i) Punjab (j) Uttar Pradesh
250 350
300
Rainfall (mm)
200
250
Rainfall
150 200
100 150
100
50 50
0 0 Jan
Feb
Mar
Apr
May
Jun
July
Aug
Sep
Oct
Nov
Dec
July
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
(k) Tamil Nadu (l) Andhra Pradesh

300 350
250 300
200 250
Rainfall
Rainfall
200
150
150
100 100
50 50
0 0
May
Aug
Jan
Feb
Mar
Apr
Sep
Jun
July
Oct
Nov
Dec
Aug
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
July
Sep
Oct
(m) West Bengal

500
Rainfall (mm)
400
300
200
100
Months
0
Obs GFDL HadGEM2
Fig. 3.4 (continued)

45 Bihar Chattisgarh
45
40 40
35 35
30 30
Tmax (°C)
Tmax (°C)
25 25
20 20
15 15
10 10
5 5
0 0
Feb
Sep
Nov
Jan
Apr
Jun
July
Aug
Oct
May
Mar
Dec
Feb
Sep
Nov
Jan
Apr
Jun
July
Aug
Oct
Mar
May
Dec
Gujrat Haryana
45 45
40 40
35 35
30
Tmax (°C)
Tmax (°C) 30
25 25
20 20
15 15
10 10
5 5
0 0
Feb
Sep
Nov
Jan
Apr
Jun
July
Aug
Oct
Mar
May
Dec
Feb
Sep
Nov
Jan
Apr
Jun
July
Aug
Oct
Mar
May
Dec
Months Months
Karnataka Madya Pradesh
40 50
35
30 40
Tmax (°C)
Tmax (°C)
25 30
20
15 20
10
10 Obs GFDL HadGEM2
5
0 0
Sep
Nov
Feb
Jan
Jun
Apr
July
Aug
Oct
Mar
May
Dec
Sep
Nov
Feb
Jan
Apr
Jun
July
Aug
Oct
Mar
May
Dec
Months
Orissa Andhra Pradesh
40 40
30 30
Tmax (°C)
Tmax (°C)
20 20
10 Obs GFDL HadGEM2 10

0 0
Nov
Dec
Jan
May
Aug
Feb
Mar
Apr
Jun
July
Sep
Oct
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
July
Aug
Sep
Fig. 3.5 Seasonal cycle of maximum temperature for major rice growing states in India: Evaluation
of GCM (GFDL_CM3 and HadGEM2_CC) baseline with observations
3.6 Future Climate Change Projections 47
Maharastra Punjab
50 50
40 40
Tmax (°C)
Tmax (°C)
30 30
20 20
10 10
0 0
Jan
Feb
Mar
Apr
May
Jun
July
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
July
Aug
Sep
Oct
Nov
Dec
Tamil Nadu Uttar Pradesh
40 50
30 40
Tmax (°C)
Tmax (°C)
30
20
20
10 10
0 0
Nov
Oct
Dec
Jan
Feb
Mar
Apr
May
Jun
July
Aug
Sep
Nov
Jan
July
Feb
Mar
Apr
May
Jun
Aug
Sep
Oct
Dec
West Bengal
40
30
Tmax (°C)
20
10
0
Obs GFDL HadGEM2
3.6 Future Climate Change Projections
The detailed future climate change projections were generated for the major 13 states
from the two GCMs in the CMIP5. A new set of emission scenarios called the Rep-
resentative Concentration Pathways (RCPs) was introduced in CMIP5. A particular
RCP represents radiative forcing and CO2 concentration reached by the year 2100,
without being linked to any specific socio-economic development storylines as in
the case of the earlier Special Report on Emission Scenarios (SRES). The CMIP5
in their Assessment Report 5 (AR5) models used the four RCP scenarios such as
RCP2.6, RCP4.5, RCP6.0 and RCP8.5 in their future climate projection model runs.
For this study, the baseline and future time horizons up to 2100s for two future sce-
narios, viz. RCP4.5 and RCP8.5, representing medium- and high-emission scenarios
were analysed. RCP4.5 represents the increases in radiative forcing (~4.5 W/m2 ) and
CO2 equivalent (~650 p.p.m) up to 2100 and stabilizes after 2100 without overshoot
Observations
30
25
20
Tmin (°C)
15
10
5
0
Months
MP Ori Pun TN UP WB
30 GFDL - Baseline
25
20
Tmin (°C)
15
10
5
0
-5
Months
MP Ori Pun TN UP WB
HadGEM2 - Baseline
30
25
Tmin (°C)
20
15
10
5
0
Months
MP Ori Pun TN UP WB
Fig. 3.6 Seasonal cycle of minimum temperature for major rice growing states in India: Evaluation
of GCM (GFDL_CM3 and HadGEM2_CC) baseline with observations
relative to preindustrial conditions. Similarly, RCP8.5 represents the increases in

radiative forcing (>8.5 W/m2 ) and CO2 equivalent (>1370 p.p.m) in 2100 and rising
even after 2100 (Moss et al. 2010). The period 1971–2000 was defined as the baseline
period. Then two time periods to represent the future, viz. mid-century (2021–2050)
and end-century (2071–2100), were selected. Hence, the climate change projection
was grouped under these two time periods and two emission scenarios for the present
study.
3.6.1 Seasonal Cycle of Rainfall Projections of HadGEM2

Model
The rainfall climatology of baseline and future projections of HadGEM2 model

were given in Fig. 3.7. The Hadley centre model predicts that the precipitation will
increase throughout the year in all the states. Ratio of increase is prominent during
the monsoon months compared to other months. Further, monthly rainfall climatol-
ogy predicts the seasonal peak and trough shows a more confidence of the model
projecting the future rainfall in all the states. Seasonal peak is noticed during the
south-west monsoon season during June to September in almost all the states except
Tamil Nadu with high north-east monsoon during October to December which is the
major indicator of model projecting the future climate. For instance, Bihar rainfall
increases about 7 and 22% under RCP4.5 and RCP8.5 scenarios, respectively, dur-
ing the end of the century. In Tamil Nadu, both the scenarios during the mid-century
show decrease in rainfall due to low radiative forcing. However, high anthropogenic
force increases the rainfall about 3 and 27% under RCP4.5 and RCP8.5 scenarios,
respectively, during end-century.
3.6.2 Seasonal Cycle of Maximum Temperature Projections

of HadGEM2 Model
The maximum temperature climatology of baseline and future projections of

HadGEM2 model were given in Fig. 3.8. The Hadley centre model projections indi-
cate that the maximum temperature climatology shows an increasing tendency in
all the states. It varies from 0.5 °C (Punjab) under RCP4.5 scenario during the mid-
century and reaches the maximum increase of 6.1 °C (Gujarat) during the end-century
under RCP8.5 scenario. For instance, Odisha maximum temperature increases about
1.07 and 1.17 °C under RCP4.5 scenario during the mid- and end-century, respec-
tively. Further, it increases about 3.05 and 4.09 °C under RCP8.5 scenario during the
mid- and end-century, respectively.
50
Rainfall (mm) Rainfall (mm) Rainfall (mm) Rainfall (mm)
0
20
40
60
80
100
0
20
40
60
80
100
120
0
20
40
60
80
100
0
100
200
300
400
500
600
700
Jan Jan Jan

Jan
Feb Feb Feb Feb
Mar Mar Mar Mar
Apr Apr Apr Apr
for two RCP scenarios

May May May May
June June June June
Bihar
Gujrat
July July July July
Karnataka
Tamil Nadu
Aug Aug Aug Aug
Sep Sep Sep
Sep
Oct Oct Oct
Oct
Nov Nov Nov
Nov Dec
Dec Dec
Dec
Rainfall (mm)
Rainfall (mm) Rainfall (mm)
Rainfall (mm)
0
50
100
150
200
250
300
350
0
50
100
150
200
250
300
0
50
100
150
200
0
20
40
60
80
100
120
140
160
Jan Jan
Jan
Feb Jan Feb
Feb
Mar Mar Feb Mar
Apr Apr Mar Apr
May Apr May
May
June May June
June
July June July
July
July
Haryana
Aug Aug Aug

Chaƫsgarh
Maharastra
Aug
Andhra Pradesh
Sep Sep Sep

Sep
Oct Oct Oct Oct
Nov Nov Nov Nov
Dec Dec Dec Dec
Fig. 3.7 HadGEM2 baseline and climate change projections of rainfall during 2050s and 2100s
3 Climate Change Projections for Major Rice Growing States of India
250 Madya Pradesh Orissa

350
200 300
Rainfall (mm)
Rainfall (mm)
250
150 200
150
100
100
50 50
0
0
May
Nov
Dec
Jan
Feb
Mar
Apr
June
July
Aug
Sep
Oct
Jan Mar May July Sep Nov
300 Punjab 500 UƩar Pradesh

450
250 400
Rainfall (mm)
Rainfall (mm)
200 350
300
150 250
200
100 150
50 100
50
0 0
Jan
Feb
Mar
Apr
May
June
July
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
June
July
Aug
Sep
Oct
Nov
Dec
West Bengal
600
500
Rainfall (mm)
400 HadGEM2_Baseline
HadGEM2_RCP4.5_2021-2050
300
HadGEM2_RCP4.5_2071-2100
200
HadGEM2_RCP8.5_2021-2050
100 HadGEM2_RCP8.5_2071-2100
0
Months
3.6.3 Seasonal Cycle of Minimum Temperature Projections

of HadGEM2 Model
Baseline and climate change projections for minimum temperature of HadGEM2

model during 2021–2050 and 2071–2100s for two RCP scenarios are given in
Fig. 3.9. The minimum temperature projections show increasing trend over the dif-
ferent states of India in almost all the months. The minimum temperature increase
52
Tmax ( °C) Tmax ( °C) Tmax ( °C) Tmax ( °C)
10
20
30
40
50
10
20
30
40
50
10
20
30
40
50
0
0
0
10
15
20
25
30
35
40
0
5
Jan Jan Jan Jan
Feb Feb Feb Feb
Mar Mar Mar Mar
Apr Apr Apr Apr
May May May May
June June June June
Bihar
Gujrat
July July July July
Karnataka
Tamil Nadu
Aug Aug Aug Aug
Sep Sep Sep Sep
Oct Oct Oct Oct
2050s and 2100s for two RCP scenarios

Nov Nov Nov Nov
Dec Dec Dec Dec
Tmax ( °C) Tmax ( °C) Tmax ( °C) Tmax ( °C)
10
20
30
40
50
10
20
30
40
50
10
20
30
40
50
10
20
30
40
50
0
0
0
0
Jan Jan Jan Jan

Feb Feb Feb Feb
Mar Mar Mar Mar
Apr Apr Apr Apr
May May May May
June June June June
July July July July
Haryana
Chaƫsgarh
Maharastra
Aug Aug Aug Aug
Andhra Pradesh
Sep Sep Sep Sep

Oct Oct Oct Oct
Nov Nov Nov Nov
Dec Dec Dec Dec
Fig. 3.8 HadGEM2 baseline and climate change projections for maximum temperature during
3 Climate Change Projections for Major Rice Growing States of India
Madya Pradesh Orissa

50 45
45 40
40 35
35 30
Tmax (°C)
Tmax ( °C)
30
25
25
20
20
15 15
10 10
5 5
0 0
Jan
Feb
Mar
Apr
May
June
July
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
June
July
Aug
Sep
Oct
Nov
Dec
45
Punjab 40
UƩar Pradesh
40 35
35 30
30 Tmax ( °C)
Tmax ( °C)
25
25
20
20
15 15
10 10
5 5
0 0
Jan
Feb
Mar
Apr
May
June
July
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
June
July
Aug
Sep
Oct
Nov
Dec
45 West Bengal
40
35 HadGEM2_Baseline
30
Tnax (°C)
25 HadGEM2_RCP4.5_2021-2050
20
15 HadGEM2_RCP4.5_2071-2100
10 HadGEM2_RCP8.5_2021-2050
5
0 HadGEM2_RCP8.5_2071-2100
Jan
Mar
May
Feb
Apr
June
July
Aug
Sep
Oct
Nov
Dec
Months
varies between 1.07 °C (Odisha) and 1.63 °C (Gujarat) under RCP4.5 scenarios dur-
ing the mid-century. Further, it also varies between 4.55 °C (Andhra Pradesh) and
5.27 °C (Chhattisgarh) under RCP4.5 scenarios during the mid-century.
HadGEM2_Baseline AP
30
TN
25 Bih
Chat
20 Guj
Tmin ( °C)
Har
15 Kar
Mah
10
MP
5 Ori
Pun
0 UP
Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec WB
HadGEM2_RCP4.5_2021-2050 35 HadGEM2_RCP4.5_2071-2100
35
30 30
25 25
Tmin ( °C)
Tmin ( °C)
20 20
15 15
10 10
5 5
0
0
May
Jan
Feb
Mar
Apr
June
July
Aug
Sep
Nov
Oct
Dec
Jan
Feb
Mar
Apr
May
June
July
Aug
Sep
Oct
Nov
Dec
HadGEM2_RCP8.5_2021-2050 35 HadGEM2_RCP8.5_2071-2100
35
30 30
25 25
Tmin ( °C)
Tmin ( °C)
20 20
15 15
10 10
5 5
0 0
Mar
May
Dec
Mar
May
Dec
June
Apr
June
Aug
Apr
Aug
Jan
July
Oct
Jan
July
Oct
Feb
Sep
Nov
Feb
Sep
Nov
Fig. 3.9 HadGEM2 baseline and climate change projections for minimum temperature during
2050s and 2100s for two RCP scenarios
3.6.4 Rainfall, Maximum and Minimum Temperature

Projections of GFDL Model
Baseline and climate change projections for annual rainfall, maximum and minimum
temperatures of GFDL model during 2021–2050 and 2071–2100s for two RCP sce-
narios are given in Fig. 3.10. The GFDL model predicts that the annual precipitation
will increase in all the states. The maximum and minimum temperature projections
show increasing trend over all the states of India. The minimum temperature increase
will be up to 2.26 °C in Haryana and 4.26 °C in Uttar Pradesh during the mid- and
end-century, respectively, under RCP4.5 scenarios. The GFDL model projections
indicate that the maximum temperature will increase up to 1.93 and 6.23 °C in Uttar
Pradesh during the mid- and end-century, respectively, under RCP8.5 scenarios.
3.7 Summary
Future climate change projections for the major rice growing states are presented
which help to predict the future rice production in India. Major focus is given for
the selection of global climate modelling approach and methodology for extracting
climate data, spatial domains of major rice growing states in India, historical observed
climate data, evaluation of baseline climate with IMD observations and future climate
change projections.
The two GCMs have good skill in simulating realistic characteristics of Asian
monsoons, and those models have better skill for simulating the future climate change
scenarios. The monthly climate data, viz. rainfall, maximum and minimum temper-
atures for the respective states, was extracted for the current and future periods from
the two different global climate models using the demarcated spatial domain. Climate
data relevant to the study area was collected from meteorological stations located in
the respective states. The rainfall peak is clearly observed during the south-west mon-
soon season from June to September in all the states except Tamil Nadu in which
north-east monsoon from October to December is observed as a major monsoon
season. However, Karnataka receives the maximum quantity of rainfall, and Uttar
Pradesh receives the lowest rainfall among the selected states. The highest values of
maximum and minimum temperatures are encountered during the summer months of
April–May, while the lowest values are observed during winter months December—
January. The rainfall extracted from the GCMs has been validated against observed
rainfall climatology for all the major rice growing states of India.
The selected CMIP5 models were evaluated for their ability to simulate precipi-
tation and temperature patterns in the respective rice growing states in India. It was
observed that the monthly temperature evolution in all the states of baseline tem-
perature has good similarity with the IMD observations. It was also observed that
both the climate models capture the timing of the highest and lowest temperatures
in annual seasonal cycle showing the good indicator of the model performance. The
1800
1600
1400
Rainfall (mm)
1200
1000
800
600
400
200
0
40
35
30
Tmax ( °C)
25
20
15
10
5
0
30
25
20
Tmin (°C)
15
10
5
0
HadGEM2_Baseline HadGEM2_RCP4.5_2021-2050
HadGEM2_RCP4.5_2071-2100 HadGEM2_RCP8.5_2021-2050
HadGEM2_RCP8.5_2071-2100
Fig. 3.10 GFDL baseline and climate change projections for annual rainfall, minimum and maxi-
mum temperatures during 2050s and 2100s for two RCP scenarios
detailed future climate change projections were generated for the major 13 states
from the two GCMs in the CMIP5.
The baseline and future time horizons were analysed up to 2100s for two future
scenarios, viz. RCP4.5 and RCP8.5, representing medium- and high-emission sce-
narios. Seasonal peak is noticed during the south-west monsoon season during June
to September in almost all the states except Tamil Nadu which dominates north-east
monsoon during October to December, thus showing the major indicator of model
projecting the future climate.
Chapter 4
Methodologies for Quantifying Climate
Change Impacts on Rice Production

4.1 Introduction
In order to study how climate change affects agriculture, scientists normally use
farm survey data collected by them, viz. primary data. They also use data from
various other sources such as government publications, called secondary data. The
data together is used in econometric models to get an understanding of the impact of
climate change on the economy. Farm survey data usually results in cross-sectional
data on farm characteristics. Farmers may suffer crop and income losses due to
climate change impacts like droughts, floods and diseases to crops. Quantification
of these losses is important but inadequately done due to lack of adequate data and
relevant measurement methodologies. Hence, results of analysis of farm surveys data
help us to quantify economic loss/gain, and this information can be used in short-
term and long-term planning. Also using farm survey questionnaires, researchers can
elicit information from the farmers about their opinion on climate change, impact of
climate change shocks on agriculture and also the adaptation and mitigation strategies
followed by them to control the negative effect of climate change on agriculture.
Long-term effects of climate change, from a macro-perspective, can be well stud-
ied from time series or panel data on yield of crops and climate-related variables.
Econometric modelling of these data sets and using climate change projections will
help to predict the possible increase/decrease in crop yield in future years. The infor-
mation so derived can be used by policymakers for optimal cropping pattern.
Various statistical and econometric tools starting from simple averages to
advanced econometric models are available at the disposal of the climate change
modeller to understand how yield of crop is affected. These procedures help to for-
mulate appropriate adaptation strategies which can be followed by government. In
this chapter, various analytical tools used in assessing the influence of climate change
on rice yield are presented.

58 4 Methodologies for Quantifying Climate Change …
4.2 Study Area and Data
For the present study, 13 Indian states where rice is predominantly grown were
selected and grouped into five regions: southern, central, eastern, northern and west-
ern. From each state, one district which performed best in terms of productivity of
rice for the past 30 or more years was identified. The districts selected from each
state and the sample sizes from each district are given in Table 4.1 and Fig. 4.1. The
study used both cross-sectional and panel data collected from the chosen districts.
Using the cross-sectional data collected from farmers, the awareness of farmers on
climate change and its effects were examined, while the panel data was used to
predict long-term effects of climate variables on rice yield.
Cross-sectional data is collected from farmers through questionnaire. The survey
was undertaken in 2014, and following information were collected from the famers.
(i) General information on the study area which includes the names of dis-
trict/block/village, date of interview and names of the enumerators.
Table 4.1 Study area and the sample size

Sl. No. State District selected Sample size (no. of farmers)
Region: Southern
1 Tamil Nadu Tanjore 112
2 Andhra Pradesh West Godavari 101
3 Karnataka Shimoga 100
Region: Central
4 Madhya Pradesh Balaghat 70
5 Maharashtra Bhandara 100
Region: Eastern
6 Bihar Jehanabad 100
7 Odisha Bhadrak 500
8 Chhattisgarh Rajpur 70
9 West Bengal Bankura 151
Region: Western
10 Gujarat Navsari 103
Region: Northern
11 Punjab Sangrur 100
12 Haryana Karnal 100
13 Uttar Pradesh Chandauli 100
Total 1707
Note The sample sizes vary from 70 to 151 in all the states except for Odisha. This variation is due
to availability of complete farm-level information. For Odisha, the sample size was 500 and it was
drawn from a major research study conducted in that state
4.2 Study Area and Data 59
Fig. 4.1 Study regions
(ii) Socio-economic profile of the farm household. These include age, gender, edu-
cational level, occupation, main source income and membership any farmers
associations.
(iii) Farm information such as land owned, type and source of irrigation, investments
on irrigation, area grown under rice in each season, net income.
(iv) Farmers were questioned on their perception of climate change. They were
asked to rank 16 climate change events (Table 4.2).
(v) Further, farmers were questioned on their opinion about the effects of climate
change shocks (Table 4.3). They were asked to rank the following effects.
(vi) Farmers were asked to state the rice technologies or practices they adopt and the
constraints if any in adopting. The crop management technologies or practices
and non-crop-oriented strategies included were:
1. Crop management oriented:
i.Direct sown rice (DSR)

ii.System of rice intensification (SRI)
iii. Modified SRI (MSRI)
iv. Machine transplanting (MT)
v. Improved management practices (change in variety, planting dates, efficient
fertilizer and water use) (IMP)
vi. Irrigation at minimum depth (IMD)
vii. Alternate wetting and drying (AWD)
Table 4.2 List of events Sl. No. Variable Notation

included for studying used
farmers’ perception of
climate change 1 Drought Drought
2 Hailstorm Hailstorm
3 Flood Flood
4 Animal disease AD
5 Increase in pest damage IPD
6 Illness of family members IFM
7 Untimely rains UR
8 Irregular weather IW
9 Temperature fluctuation high TFH
10 Temperature fluctuation low TFL
11 Change in soil salinity and SSDI
decrease/increase in soil moisture
12 Major changes in crop pattern CCP
13 Major changes in livestock asset CLA
14 Major changes in farm investment CFI
(capital formation)
15 Crop failure CF
16 Depletion in groundwater DGW
Table 4.3 List of effects of Sl. No. Variable Notation used

climate change
1 Decline in crop yield DCY
2 Loss of assets LA
3 Loss of income LI
4 Food insecurity/shortage FIS
5 Death of livestock DL
6 Decline in consumption DC
7 Decline in health DH
viii. Supplemental irrigation through groundwater (SITG)

ix. Integrated pest management (IPM)
x. Machine harvesting (MH)
xi. Application of more organic manure (AMOM)
xii. Drip irrigation (DI)
xiii. Subsurface drainage (SSD)
2. Non-crop-oriented strategies
i. Borrowed money from relatives/others

ii. Received assistance from government/NGOs
4.2 Study Area and Data 61
iii. Less food consumption

iv. Shifted to non-farm employment
v. Outmigration to cities
vi. Government initiations to tackle climate change impacts.
At a macro-level, panel data was collected from historical records and government
publications. The panel data consisted of
a. area, production and productivity of rice crop in the selected district over years
b. monthly data on temperature and rainfall in the selected district over years.
4.3 Statistical Tools for Analysis of Primary Data
a. Garrett Ranking technique
In the present study, farmers were asked to rank a number of alternatives (or factors)
regarding climate change events (as listed in Table 4.2) and we are interested in
knowing which factor is most preferred by the farmers. For this purpose, the Garrett
Ranking technique (Garrett and Woodworth 1971) was applied. For example, in
a farm survey in Tamil Nadu (Palanisami et al. 2014), farmers were asked about
changes in climate over the last ten years and they were asked to rank the factors
according to their perception:
1. Incidence of drought
2. Amount of rainfall/moisture
3. Rainfall distribution/delays
4. Extreme weather events
5. Temperature changes
6. Growing season.
In the Garrett Ranking method, the farmers’ rankings are first converted as

Percent Position = 100 Ri j − 0.5 /N j
where
Ri j Rank given for the ith factor by the jth farmer, and
N j Number of factors ranked by the jth respondent.
By referring to Garrett’s table, the per cent positions are converted into scores. Then
for each factor, the scores of each individual are added from which the average scores
can be obtained.
The frequency of the factor rankings can be listed in a table. An example of such
a table is given below (Table 4.4).
The average scores were obtained as shown in Table 4.5.
Table 4.4 Factors influencing farmers towards climate change

Factor Number of farmers who gave rank Total
1 2 3 4 5 6
Incidence of drought 15 14 25 30 41 55 180
Amount of rainfall/moisture 23 52 53 29 19 4 180
Rainfall distribution/delays 45 62 41 15 14 3 180
Extreme weather events 83 31 28 18 12 8 180
Temperature changes 6 13 24 58 53 26 180
Growing season 8 8 9 30 41 84 180
Garrett score 77 63 54 46 36 23
Table 4.5 Garrett ranking: selection of factors

Factor Rank Total garrett Average score Rank
1 2 3 4 5 6 score
Incidence of 1155 882 1350 1380 1476 1265 7508 41.7 5

drought
Amount of 1771 3276 2862 1334 684 92 10,019 55.7 3
rain-
fall/moisture
Rainfall 3465 3906 2214 690 504 69 10,848 60.3 2
distribu-
tion/delays
Extreme 6391 1953 1512 828 432 184 11,300 62.8 1
weather
events
Temperature 462 819 1296 2668 1908 598 7751 43.1 4
changes
Growing 616 504 486 1380 1476 1932 6394 35.5 6
season
Table 4.5 shows that farmers perceive extreme events as the most important climate
change factor followed by rainfall distribution/delays.
The same procedure was employed in the present study with events as given in
Table 4.2.
(a) Friedman Rank Test
Nonparametric tests can be applied to test the consistency of the rankings. For exam-
ple, to test whether the farmers were consistent in ranking the six factors stated above,
Friedman’s test was employed. It is a nonparametric test, and it resembles two-way
analysis of variance used in parametric tests. It is used to test for differences between
groups when the dependent variable being measured is ordinal, for example, ranks.
Friedman Rank Test is used to determine whether c groups have been selected from
populations having equal medians.
4.3 Statistical Tools for Analysis of Primary Data 63
H0 : m 1 = m 2 = m 3 = · · · = m c
against the alternative, H1 : Not all mj are equal (j = 1,2,…c).

The Friedman test statistic is given by
12 c
F= R 2 − 3r (c + 1)
r c(c + 1) j=1 j
where R j = Sum of the ranks for the jth group.

This statistic follows a χ 2 distribution with number of degrees of freedom equal
to c − 1.
Thus, for any chosen level of significance, α, the null hypothesis is rejected if the
calculated value of the F statistic is greater than the upper-tail critical value for the
chi-square distribution with c − 1 degrees of freedom.
The results of the test, when applied to the above data set, are as given in Table 4.6.
In the above example, the test statistics has 5 degrees of freedom and the calculated
value is 310.952. The table value of χ 2 at 5% significance level is 12.59, and so the
null hypothesis is rejected which implies that the farmers did not consistently rank
the factors; thus, the farmers’ perceptions of climate change effects differ across the
sample.
(b) Kendall’s coefficient of concordance
Kendall’s coefficient of concordance (Kendall and Smith 1939) is another statistical
measure to test the consistency of rankings. This test statistic is defined by
12 S
W =
m2 M3 − M
where m is the number of objects or factors (equal to 6 in the above example)

and M is the number of judges (farmers = 180 in the above example) and S =
Table 4.6 Results of Factor Rj

friedman rank test R 2j
Incidence of drought 773 597,529

Amount of rainfall/moisture 521 271,441
Rainfall distribution/delays 440 193,600
Extreme weather events 409 167,281
Temperature changes 757 573,049
Growing season 880 774,400
Total 3780 2,577,300
F statistic 310.95
χ 2 -table value (at 5 degrees of freedom) 12.59
i=M 2
i=1 Ri − R̄ ; Ri = sum of the ranks of factor i and R̄ = m(M+1)
2
. The quantity
W is always positive but less than 1. If W = 1, there is perfect agreement. But, if W
= 0 there is no overall agreement among the components in ranking the regions. To
test the significance of W, use the statistic
χ 2 = m(M − 1)W
and it has a chi-square distribution with M − 1 degrees of freedom.

For the data set discussed above, the results of Kendall’s test are also in agreement
with Friedman’s test results. The Kendall’s coefficient of concordance, W, was 0.346
which was found to be significant at 1% level implying no concordance or agreement
in rankings.
4.4 A Review of Studies on Economic Impact of Climate

Change on Crop Yields
There are many studies to estimate the impact of environmental changes on crop
productivity. These studies used agro-economic models or multiple regression tech-
niques. A brief review of studies done mostly after 2000 is presented below.
In agronomic–economic methods, crop models are used. They are developed
from field experiments (FAO 2000; Kumar and Parikh 1998a). In these experiments,
experiments are conducted in field or laboratory conditions with many valid future
climates and carbon dioxide levels, but all farming practices in experimental condi-
tions are held fixed. The aim is to elicit differences in outcomes which are due to
temperature, precipitation or carbon dioxide. The cross-sectional approach is called
as Ricardian method. In this method, the performances are studied across climate
zones (Mendelsohn et al. 1994, 1996; Kumar and Parikh 1998b). This approach was
used in studies across the USA (Mendelsohn et al. 1994, 1996)
Table 4.7 provides a selected list of papers on economic impact of CC on agricul-
ture published during 2000–14. The major econometric model used in these studies
is Just–Pope production function (Just–Pope 1978). The reason being it helps to esti-
mate both mean yield and variability in yield simultaneously, and both these variables
are affected by CC and so estimation of the effects of CC on these variables become
easy.
Seo and Mendelsohn (2008) used structural Ricardian models and demonstrated
that increases in temperature induce African farmers to adopt mixed farming and
avoid specialized farms or livestock farms. Wang et al. (2009), in a study on China’s
agriculture, applied Ricardian modelling, and the crop net profit was the dependent
variable and temperature and rainfall were the independent variables. Their study
showed that temperature rise will affect rainfed farms, but it will be profitable to
farmers in irrigated area.
Table 4.7 Selected list of papers on economic impacts of climate change published during 2000–2014
S. No. Authors Year Country Crop/variable Technique used CC variables
1 Mendelsohn et al. 2000 Africa GDP Simulation IPCC forecast
2 Chen et al. 2004 US Major US agricultural JP Average temperature and Annual
crops Rainfall
3 Isik and Devadoss 2006 US Wheat, Barley, Potato JP Average temperature and Annual
and Sugar beet Rainfall
4 Schlenker and 2008 US Corn, Soybeans and Regression Temperature data using Parameter
Roberts Cotton -elevation Regressions on
Independent Slopes Model (PRISM)
5 Seo and Mendelsohn 2008 Africa Farming Income Structural Summer, Winter temperature and
Ricardian Rainfall
Model
6 McCarl et al. 2008 US Corn, cotton, sorghum, JP Average temperature, SD
soybean temperature and Annual Rainfall
and winter wheat
7 Wang et al. 2009 China Crop net revenue Ricardian Temperature and Rainfall
model
8 Deressa and Hassen 2009 Ethiopia Crop net revenue Ricardian Seasonal Temperature and Rainfall
model
9 Kim and Pang 2009 Korea Rice JP Average temperature and Annual
Rainfall
10 Ranganathan 2009 India Nine major crops grown JP Average temperature and Annual
4.4 A Review of Studies on Economic Impact of Climate Change on Crop Yields
in Tamil Nadu Rainfall

11 Barnwal and Kotani 2010 India Rice JP and Qreg Temperature, rainfall, SD
Temperature and SD Rainfall
12 Cabas et al. 2010 Canada Corn, soybean and JP Grow days, temperature and rainfall
winter wheat
65
(continued)
66
S. No. Authors Year Country Crop/variable Technique CC variables

used
13 Joshi et al. 2011 Nepal Six major crops in Nepal Time series analysis
14 Rainer et al. 2011 China Aggregate grain output JP Annual average temperature, annual
from rice, wheat, corn, precipitation
sorghum, millet, tubers and average sunshine
and beans
15 Palanisami et al. 2011 India Rice JP Average temperature and Annual
Rainfall
16 Boubacar 2010 West Africa Maize, Millet and JP Standardized Precipitation Index
Sorghum (SPI), degree
days, Average temperature and
Annual Rainfall
17 Sarker et al. (a) 2012 Bangladesh Rice JP Min and Max temperature, Rainfall
18 Sarker et al. (a) 2012 Bangladesh Rice JP Min and Max temperature, Rainfall
19 Acquah et al. 2012 Ghana Maize JP Average temperature and Annual
Rainfall
20 Zainal et al. 2013 Malaysia Crop net revenue Ricardian Seasonal Temperature and Rainfall
model
21 Hasanthika et al. 2013 Sri Lanka Rice JP Min and Max temperature, Rainfall
22 Pattanayak and 2013 India Rice Regression, Min and Max temperature, Solar
Kavikumar simula- Radiation, Rainfall
tion
22 Mahadeb et al. 2014 Nepal Rice, Maize and Wheat JP Temperature and annual Rainfall
4 Methodologies for Quantifying Climate Change …
4.4 A Review of Studies on Economic Impact of Climate Change on Crop Yields 67
In a study to understand the effect of CC on crop production in Ethiopia, Deressa

and Hassen (2009) applied Ricardian modelling approach. Their study revealed that a
small increase in temperature in summer and winter would significantly reduce crop
net revenue per hectare. But increase in rainfall precipitation during spring would
significantly increase net crop revenue per hectare. Similarly, Zainal et al. (2013)
applied and analysed the CC impact on crop net revenue. The study indicated that
temperature and rainfall had significant negative impact on production of paddy in
the study area.
The agro-ecological zone-based approach was also used (AEZ) (FAO 1992, 1996).
Variability in crop yields is important because agricultural production is very sen-
sitive to changes in rainfall and temperature. Not much attention was paid to study
this important aspect (Bindi et al. 1996; Mearns et al. 1997). Econometric models
are usually applied to study the effect of climate change on impact of climate yield
variability. Mendelsohn et al. (2000) used simulation of IPCC forecast to estimate
the GDP loss due to CC in African agriculture. They found that African agriculture
is very vulnerable to CC and that potential damage may be large both in absolute
terms and as a fraction of agricultural GDP. Schlenker and Roberts (2008) stud-
ied the nonlinear temperature effects on crop yields. Chen et al. (2004) and Isik
and Devadoss (2006) have applied regression technique by employing a Just–Pope
(1978) production functional form. They estimated both the mean and variability in
crop productivity. Chen et al. (2004) demonstrated that changes in climate variables
modify crop yield values and variances.
Isik and Devadoss (2006) used Just–Pope production function to study the impacts
of projected climate change on the yield of wheat, barley, potato and sugar beet in
Idaho, USA. The independent variables employed were total precipitation, temper-
ature and trend. They showed that climate change will have limited effects on the
mean yields, but will significantly decrease the variance and covariance for most of
the crops used in the study. Ranganathan (2009) used Just–Pope function methodol-
ogy to estimate effects of climate change on nine major crops grown in Tamil Nadu.
Further, by using multi-goal linear programming technique, he formulated optimal
cropping plan for Tamil Nadu for 2021 and 2026. Palanisami et al. (2011) studied
the CC impacts of agriculture in Godavari basin, Andhra Pradesh, India. The study
revealed that rice production will decrease during mid- and end-century periods by
16 and 36%, respectively. But if water and labour-saving technologies are employed,
the reduction in rice yield will be removed during mid-century and it will be only 19%
during end-century period. Also if these technologies are followed, on the whole,
water saving will be about 20%
Barnwal and Kotani’s (2010) study revealed that increase in temperature as well
as inter-annual variance of temperature and rainfall adversely affected the average
yield of rice in Andhra Pradesh, India. Also, increase in average temperature, rainfall
and their respective variances are likely to increase inter-annual variability in rice
yield. Using quantile regression equations, they showed that changes of rice yield to
climate differ significantly across the quantiles of the yield distribution. Sarker et al.
(2012) studied the climate change effects on three major rice varieties: Aus, Aman
and Boroin Bangladesh. They used Just–Pope function to estimate the mean and
variability in yield. Their study revealed that the effects of climate change variables
vary among the varieties. Kim and Pang (2009) analysed the climate change effects
on rice yield in Korea by employing Just–Pope production function. They found that
mean yield is positively related to temperature and negatively related to rainfall while
both climate variables are risk increasing for variability in yield. Hasanthika et al.
(2013), using Just–Pope production function, studied the climate variability, risk and
paddy production. The study showed that climate factors, rainfall and maximum
temperature, and production factors such as labour, machinery cost and cultivation
extents are positively related to the probability distribution of yield. Therefore, their
variability induces increase in risk.
In a study on crop yields and yield variability on maize, millet and sorghum in
Sahel, Boubacar (2010) found that increase in degree days will be harmful to crop
yields, but soil property index and precipitation intensity index statistically significant
positive impacts in mean yields. Cabas et al. (2010) examined the effect of climatic
and non-climatic factors on the mean and variance of corn, soybeans and winter
wheat in southern Ontario, Canada, over a period of 26 years, 1981–2006, using
Just–Pope modelling framework. They included input change, area change, monthly
temperature and monthly rainfall from April to October as explanatory variables.
In Just–Pope modelling, an important assumption is stationarity of yield distribu-
tion. This hypothesis was examined by McCarl et al. (2008). They used temperature,
precipitation, variance of intra-annual temperature, index of rainfall intensity and
Palmer Drought Severity Index (PDSI) as the main variables in the analysis. They
found that stationarity of yield distribution is not tenable when the mean and variance
of key climate variables change over time. Also they showed that the mean of the
crop yields is affected by the average temperature and precipitation. They concluded
that stationarity is a doubtful assumption and stressed the importance of testing this
assumption before going for probabilistic models.
One study on positive impact of climate change has been made by Holst et al.
(2010). They used JP methodology and found that grain production in South China
might be a beneficiary of global warming in the short run and estimated that 1 °C
increase in annual average temperature in South China would increase the grain
output by 3.79 million tons. In contrast to this, the study revealed that the impact of
global warming in North China is negative, small and insignificant.
De-Graft Acquah, Henry and Kyei, Clement (2012) using JP framework studied
the effect of CC on maize yield. The study showed that mean maize yield has positive
relation related to crop area but has negative association with rainfall and temperature.
The study also indicated that when crop area and temperature increase, maize yield
variability also increases. But when precipitation decreases, variability in maize yield
decreases.
4.5 Econometric Tools for Studying Climate Change Impacts 69
4.5 Econometric Tools for Studying Climate Change

Impacts
This section describes various statistical and econometric tools employed in the
present study.
a. Testing for normality of data
Testing for normality of yield data is the first step in any econometric modelling
as various tests of the parameters are based on this assumption especially if we
want to apply Maximum Likelihood Estimate (MLE) methods. There are many tests
available for this purpose. The important tests are (i) Shapiro–Wilk test and Lilliefors
test. These tests have been applied for yield data prior to modelling. Both of these
tests can be done using software packages such as Gretl.
b. Just–Pope Production Function
The centre of interest of the present study is on the impact of climate change on
rice crop production in the five regions of India using an econometric approach.
Specifically, we applied Just–Pope production function and quantile regression. The
analysis is done separately for each region. The mathematical relation between pro-
ductivity (y) of rice and a set of explanatory variables (X) denoted by the vector is
given by the Just and Pope stochastic production function (Just and Pope 1978):
y = f (X ; β) + ωh(X ; δ)0.5 (4.1)
where ω is the random term with mean zero and variance 1,β and δ are the constants
to be estimated from past data. The error term,ωh(X ; δ)0.5 in Eq. (4.1), shows that
Just–Pope model has heteroscedasticity error terms. The variable y is a panel data
consisting of the yield per hectare observed over a period of time across the states
belonging to the region. The set of explanatory variables include trend, average tem-
perature, annual precipitation and dummy variables representing the states belonging
to the region. The expected rice productivity is E(y) = f (X ; β), and crop variability
is V (y) = h(X ; δ). Therefore, f (X ; β) and h(X ; δ) are called mean and variance
functions, respectively. The derivatives of h(X ; δ) w.r.t. precipitation and tempera-
ture can be used to check if a climate variable increases or decreases crop variability.
Therefore, if, h x = ∂∂hx > 0, it shows that the corresponding independent variable x
will increase the risk. On the other hand, if h x < 0, it means risk decreasing. This
is the main advantage of using the Just–Pope production function. We can estimate
both the mean yield and also variability in yield simultaneously.
The procedure to estimate the production function can be considered as estimation
with heteroscedastic errors as in the following equation (Saha et al. 1997; Kumbhakar
1997)
y = f (X ; β) + u (4.2)
where u = ωh(X ; β)0.5 with E(u) = 0 and Var(u) = h(X ; δ). There are two
approaches suggested in many studies to estimate the mean and variance functions
of the Just–Pope production function. They are Feasible Generalized Least Squares
(FGLS) and Maximum Likelihood Estimation (MLE).
The two methods have their own advantages and disadvantages. The FGLS
approach is employed in many studies. For large samples, FGLS approach is used
to estimate fixed effects models (Judge et al. 1988). Further, our panel data consists
of yield of rice across different districts observed over a period of about 30 years.
So the estimation may lead to the problems of heteroscedasticity and autocorrelation
(Gujarati 2004). As stated by Sarker et al. (2012), these two problems are better
addressed in FGLS methodology. On the other, the MLE approach produces consis-
tent and more efficient estimates for the parameters than FGLS approach and more
suitable for small samples (Saha et al. 1997). The two methods are described below.
The essential steps for estimating the Just–Pope production function using FGLS
are as follows:
1. Estimate the mean function by running OLS regression of y (yield of rice per ha)
on f (X ; β).
2. Compute the residuals u it .
3. Regress natural logarithm of u it2 , i.e. log u it2 on log(h(X ; δ)), and get the resid-
uals in the second regression.
4. Obtain the antilogarithms of the residuals obtained in Step 3. These are consistent
estimates of the variances.
5. Again estimate the original mean function f (X ; β) by weighted least squares
with inverse of variances (obtained in step 4) as weights. That is, we apply WLS
to estimate the equation
yh −1/2 (X ; δ) = f (X ; β)h −1/2 (X ; δ) + uh −1/2 (X ; δ). (4.3)
In the MLE approach, the log-likelihood function is maximized. If the error term ω is
assumed to be N (0, 1), Eq. (4.1) shows that the variable y has a normal distribution
with mean f (X ; β) and variance h(X ; δ) and so the log-likelihood function can be
written as
yh −1/2 (X ; δ) = f (X ; β)h −1/2 (X ; δ) + uh −1/2 (X ; δ) (4.4)
The parameters β and δ are estimated by maximizing the above function.

Our experience with data from farm survey shows that, in general, the optimum
value of log-likelihood function under MLE is greater than that estimated by FGLS
implying that MLE estimates are superior. But for some data, converge problems
occur when estimation is done using MLE while no such difficulty arises under
FGLS. Hence in the present study, both methods are employed. We first apply MLE
and estimate the parameters and if iterations do not converge we use FGLS approach.
The two methods of estimation can be implemented using Gretl Econometrics

package or STATA. The sample Gretl Script file and Do file of STATA are shown in
Appendix.
The J-P production function was estimated by identifying the explanatory vari-
ables and then explicitly stating the forms of the mean and variance functions:
f (X ; β) and h(X ; δ). In the literature, for rice crop, different climate variables and
different functional forms have been used (see Table 4.8 for details). Studies by Chen
et al. (2004), Isik and Devadoss (2006), Ranganathan (2009), Palanisami et al. (2011),
Kim and Pang (2009) have used a quadratic functional form for the mean function
using average temperature during growing season and annual rainfall. For the vari-
ance function, either exponential or Cobb–Douglas form was used. Aye and Ater
(2012) used linear, quadratic and square root forms for the mean function and linear
form for the variance function. McCarl et al. (2008) introduced the standard deviation
of temperature in the mean and variance functions to study the climate change effects
on corn, cotton, sorghum soybeans and winter wheat. Similarly, Barnwal and Kotani
(2010) included standard deviation of rainfall and temperature in the two functions
because variation in temperature and rainfall adversely affects the mean and variance
in rice yield. Recent studies show that rather than average temperature, maximum and
minimum temperatures have stronger effects on rice yield. For example, Pattanayak
Table 4.8 Functional forms of Just–Pope function

Sl. No. Variables Functional formsa for
Mean Variance
1 Trend, AvgT and RainT (i) Linear in trend, AvgT (i) Exponential w.r.t.
and RainT trend, AvgT and RainT
(ii) Linear in trend and (ii) Cobb–Douglas w.r.t.
quadratic in AvgT and trend, AvgT and RainT
RainT
(iii) Linear in trend and
Cobb–Douglas in AvgT
and RainT
(iv) Linear in trend and
square root in AvgT and
RainT
2 MaxT, MinT, RainT, trend (i) Linear in trend and Cobb–Douglas w.r.t.
quadratic in MaxT, MinT MaxT, MinT, RainT and
and RainT trend
(ii) Linear in trend and
Cobb–Douglas in AvgT
and RainT
3 AvgT, RainT, SD(AvgT), Linear w.r.t. all variables Linear w.r.t. all variables
SD(RainT) and trend
AvgT = Average temperature; RainT = Annual/seasonal rainfall; SD(T) = standard deviation of
rainfall; SD(RainT ) = Standard deviation of rainfall; MaxT = Maximum temperature; MinT =
Minimum temperature and t = trend
a The functional forms are given omitting the dummy variable terms
and Kavi Kumar (2013) have shown that both minimum and maximum tempera-
tures adversely affect rice yield during different growth phases. They have estimated
an annual loss of 172 million tons. Similarly, Sarker et al. (2012) have studied the
effect of changes in maximum and minimum temperatures on rice yield. They have
both Cobb–Douglas and quadratic forms for the mean function and Cobb–Douglas
form for the variance function. Similarly, Hasanthika et al. (2013) included these two
variables to study the climate variability and risk in paddy production in Sri Lanka.
Thus, we find that while average temperature and rainfall are most commonly
included in the mean and variance functions, use of maximum, minimum and standard
deviation of temperature and rainfall are more plausible from the biological point of
view in explaining the mean and variability in rice yield.
The above discussion shows that the two functions can be stated as:
Model 1
Mean Function:
Linear

i=S−1
f (X ; β) = β0 + β1 t + β2 T + β3 P + β6+i Di (4.5a)
i=1
Quadratic

i=S−1
f (X ; β) = β0 + β1 t + β2 T + β3 P + β4 P 2 + β5 T 2 + β6 P T + β6+i Di (4.5b)
i=1
Square root

i=S−1
f (X ; β) = β0 + β1 t + β2 T + β3 P + β4 P 2 + β5 T 2 + β6 P T + β6+i Di (4.5c)
i=1
where t represents time trend, T is the average temperature during the rice growing
months in a year, P is the annual rainfall, Di , i = 1, 2 . . . S are the district dummy
variables which take values 1 and 0 and S is the number of states included in the region.
Annual rainfall not only reflects rainfall falling directly on the crop but also water
accumulated between seasons within the year (Isik and Devadoss 2006). These forms
of mean function are consistent with the postulates of Just and Pope (1978). Further,
these forms are flexible enough to estimate using MLE or FGLS. The inclusion of
time trend represents technological progress in agricultural front during the sample
periods.
Variance Function
The variance function h(X ; δ, η) was assumed to have exponential form
Exponential:
h(X ; δ) = exp(δ0 + δ1 t + δ2 T + δ3 P) (4.6a)
Cobb–Douglas:
β
h(X ; δ) = β0 t α xj j (4.6b)
j
The exponential was employed first by Harvey (1976) and after that by many authors
(Palanisami et al. 2011; Ranganathan 2009; Isik and Devadoss 2006; Isik and Khanna
2003; Asche and Tveteras 1999). The important benefit of this mathematical form
is that it guarantees positive output variance. Also, the riskiness of an input variable
can be quickly obtained by examining the sign of the coefficient of that variable in
the function. For example, with the above functional form, it can be easily checked
β
that h(X ; δ) = β0 t α j x j j . As the mathematical function h is always positive,
precipitation will be risk increasing if δ1 > 0 and it will be risk decreasing if δ1 < 0.
In view of these advantages, in the present study, the exponential form of the variance
function is employed.
Thus, Model 1 has three submodels:
(i) Linear mean function and exponential variance function
(ii) Quadratic mean function and exponential variance function and
(iii) Square-root mean function and exponential variance function.
Model 2
Mean Function
Linear:

S−1
f (X ; β) = β0 + β1 t + β2 MaxT + β3 MinT + β4 RainT + β9+i Di (4.7a)
i=1
Quadratic
f (X ; β) = β0 + β1 t + β2 MaxT + β3 MinT + β4 RainT + β5 MaxT 2 + β6 MinT 2

+ β7 MaxT × MinT + β8 MaxT × RainT

S−1
+ β9 MinT × RainT + β9+i Di (4.7b)
i=1
Variance Function
Exponential:
h(X ; δ) = exp(δ0 + δ1 t + δ2 MaxT + δ3 MinT + δ4 RainT ) (4.8)
So Model 2 has two submodels:

(i) Linear mean function and exponential variance function.
(ii) Quadratic mean and exponential variance function.
Model 3

s
f (X ; β) = β0 + β1 t + β2 T + β3 S DT + β4 P + β5 S D P + β5+i (T × Statei )
i=1

$1
S−1
+ β5+S+i (P × Statei )i + β5+2s+i Di (4.9)
i=1 i=1
Variance Function
Same as the mean function with and without interaction terms.
In our present study, we restrict ourselves with Model 2 wherein we use maximum
temperature, minimum temperature and rainfall as climatic variables. As already
stated, these variables have been used in recent studies to estimate the impact of
climate change on rice yield and the best model among all the submodels (linear or
quadratic) is selected based on Akaike Information Criterion (AIC) and other model
characteristics.
c. Akaike Information Criterion (AIC)
Introduced by Akaike, Hirotugu (1973) provides a means of selecting the best model.
This selection is based on the AIC score given by
AIC = 2K − 2 ln(L) (4.10)
where K is the number of parameters in the model (including constant and variance)
and L is the maximized value of the likelihood function of the model. When we have
small samples, that is, when n/K < 40, we apply a correction and the corrected AIC
is given by
2K (K + 1)
AICc = AIC + (4.11)
n−K −1
In the above model, n is the number of observations. The best model is one for which
the AIC or AICc is a minimum.
d. Fixed Effect or Random Effect Model
Panel data is commonly analysed using either fixed effect or random effect models
(Baltagi 2005). In the present study, fixed effects models are purposively used. The
main reason is that it helps to estimate state-specific effects; we can include dummy
variables for the states in the mean and variance functions. This will accommodate
the gap between mean and variance functions of states. Another reason, as pointed out
by Barnwal and Kotani (2010), Sarker et al. (2012), is that there may be association
between unobserved time invariant properties and the variables included in the study.
The presence of such correlations violates the random effects model assumptions.
For example, as emphasized by Barnwal and Kotani (2010), states with relatively
more suitable climate may have developed better irrigation infrastructure and better
soil fertility over time. In our present study, we have selected one district in each
state which has best performance in terms of yield during a period of time. Hence,
it is more likely that these districts have developed better soil fertility and irrigation
sources. Hence, it is reasonable to assume that there is good correlation between
the yield and these unobserved time invariant characteristics. So the assumptions of
random effects model are violated. So fixed effects models are more suitable, and
they will provide unbiased estimates. Studies by McCarl et al. (2008), Cabas et al.
(2010), Barnwal and Kotani (2010) and Sarker et al. (2012) also support the choice
of fixed effects model, and so the same model is employed in the present study also.
e. Panel Root Tests
Just and Pope production function assumes that the variables in the model are sta-
tionary (Chen et al. 2004). If this assumption is violated, it will give spurious results
(Chen and Chang 2005; Granger and Newbold 1974). Granger and Newbold (1974)
showed that deterministic and stochastic trends in the variables can induce spuri-
ous correlations between the variables. Time series data with more than 20 years
should be tested for stationarity because a spurious result could arise because of
non-stationarity on the time series variable in question (Chen et al. 2004; McCarl
et al. 2008; Sarkar et al. 2012). Even though inclusion of time trend will minimize
this problem, it may not completely solve it when spurious correlations are present.
So the variables included in the panel data must be checked for stationarity before
we estimate the model using FGLS or MLE method. That is, it is important to test
for the presence of unit root for each variable prior to estimation of the model. If the
variable is non-stationary, i.e. if they possess I(1) property1 , then they must be differ-
enced before we perform the analysis. Traditional unit root tests (e.g. Dickey–Fuller
(DF) test and augmented Dickey–Fuller (ADF) test) are suitable for testing temporal
series data on single variable only. They cannot be applied for panel data where we
have time series data on different cross sections. But a number of panel root tests
have been developed and some of them are developed recently.
The main objective of these tests is to test whether a given series (panel) is sta-
tionary for all the individual panel units (in our study, states belonging to the region).
In all these tests, the given panel data is modelled by the structure
yit = (1 − φi )μi + φi yit−1 + εit , i = 1, 2, . . . N ; t = 1, 2, . . . T (4.12)
Rearranging the terms, the above equation can be written as
yit = αi + βi yit−1 + εit , i = 1, 2, .N ; t = 1, 2, . . . T (4.13)
1A time series is defined to be stationary, if its statistical properties do not change over time. For
example, for a stationary time series mean and variance remain constants. A time series which is
not stationary can be converted into a stationary time series by differencing. If a time series {yt } is
not stationary, but, if the differenced series {yt − yt−1 } is stationary, then the original series {yt } is
said to possess I (1) property.
where
βi = φi − 1, i = 1, 2, . . . N (4.14)
We want to test the hypothesis that φi = 1 for all i. That is, we want to test for unit
root. This is equivalent to testing the hypothesis that H0 : βi = 0 for all i.
There are a number of panel unit root tests. Hurlin and Mignon (2007) provide an
overview of such tests. The popular tests are Levin, Lin and Chu (LLC) 2002; Im,
Pesaran and Shin (IPS) 2003 and Fisher’s test proposed by Maddala and Wu (1999)
and Choi (2001). Fisher’s unit root tests are widely applied. For example, Sarker et al.
(2012) have applied this test as it provides more precise results and achieves higher
power compared to other tests. Barnwal and Kotani (2010) have used the same test
for performing panel root tests. This test combines the p-values of the unit root test
statistics in each cross-sectional unit and this test, unlike LLC and IPS tests, does
not require the panel data to be balanced. In the present study, the same test has been
used for panel unit root testing. The null and alternative hypotheses are:
H0 : All panels contain unit roots
Ha : At least one panel is stationary.
Panel root tests are applied under two models: a) with individual effects without
trend (i.e. with intercept only) and b) with individual effects with trend. The second
option is suitable when we believe the presence of trend in the concerned variable.
For our panel data, we have data for about 25–30 years. Hence, it is important to
check for the presence of trend before choosing model. To perform the panel root
tests, the MATLAB coding by Hurlin C., and downloaded from the website http://
www.execandshare.org/CompanionSite, was used. This MATLAB coding computes
the Fisher panel unit root tests, proposed by Choi (2001) and Maddala and Wu
(1999). Both tests combine the significant levels obtained from individual augmented
Dickey–Fuller (ADF) tests.
f. Quantile Regression
One important question is whether climate change factors affect the quantiles of
the yield distribution uniformly or differently. Usually, it is hypothesized that lower
levels of yield are more likely to be sensitive for changes in climate variables (Barnwal
and Kotani (2010). In order to address the above stated hypothesis, this study uses
quantile regression. Quantile regressions are variations of ordinary least squares
(OLS). The relationship between average yield and a set of explanatory variables
is developed in OLS, whereas quantile regression finds the relationship between
a quantile yield (e.g. median yield, 25% quartile yield, 75% quartile yield, etc.)
and the given set of explanatory variables. It was first introduced by Koenker and
Bassett (1978). In our study, it can be used to explain how a given quantile k (0<k<1)
of the rice yield changes due to changes in one or more climatic variables. It has
been shown that when the probability distribution of the response variable (in our
study rice yield) is non-normal, quantile regression estimators are more efficient than
OLS estimators (Buchinsky 1998). So when the yield data of crops is non-normally
distrusted, quantile regression estimators give deeper insights into the relationship
between yield and climate variables. The quantile regression model can be specified
as
yi = X i βθ + u 6i with Quantθ (yi |X i ) = X i βθ (4.15)
where Quantθ (yi |X i ) represents θ th conditional quantile of rice yield y and X denotes
a set of explanatory variables, trend, temperature, rainfall and state dummies. The
quantile regression functions can be fitted by using Gretl software package.
4.6 Summary
This chapter summarizes the methodologies applied in the present study for quanti-
fying the climate change impacts on rice production. A total of 1707 farmers from
13 states, grouped into 5 regions, were selected for the study. Garrett Ranking tech-
nique, Friedman Rank Test and Kendall’s coefficient of concordance were used as the
statistical tools to analyse primary data collected on farmers’ perception of climate
change. Econometric tools such as Just–Pope production function, Akaike Informa-
tion Criterion, fixed effect model and panel root tests are employed in the current
study to assess the economic impact of climate change on rice production. A review
of past studies on economic impact of climate change on crop yields is also presented
in the chapter.
Appendix: Sample Gretl Scripts, Do File and Sample

Outputs
1. Sample Gretl Script to estimate Just–Pope production function Using FGLS
# Estimation of Just Pope Production Function Using FGLS

# Script by Dr.C.R.Ranganathan 18-08-14
#
# Opening the Data File
open G:\CRRDataFiles\NorthRAvgJPData.gdt
# Generating Square, Interaction of CC variables and Dummy Variables for States
genr AvgT2=AvgT^2
genr RainT2=RainT^2
genrAvgTRainT=AvgT*RainT
genr Punjab=(StateID=1)
genr Haryana=(StateID=2)
# First Stage ols and residuals
ols Yield const Time AvgTRainT AvgT2 RainT2 AvgTRainT \
Punjab Haryana
series uh= $uhat
# Second Statge: Estimation of Variance Function

genr Luh2=log(uh^2)
ols Luh2 const Time AvgTRainT Punjab Haryana
# Third Stage Weighted Least Square Method to estimate mean function
seriesLr=$yhat
genr W=1/exp(Lr)
wls W Yield const Time AvgTRainT AvgT2 RainT2 AvgTRainT \
Punjab Haryana
Sample Output
gretl version 1.9.91
Current session: 2014-09-30 12:49

#
? open G:.gdt
Read datafile G:.gdt

periodicity: 1, maxobs: 93
observations range: 1 to 93
Listing 6 variables:
0) const 1) StateID 2) Yield 3) Time 4) AvgT
5) RainT

? genr AvgT2=AvgT^2
Generated series AvgT2 (ID 6)
? genr RainT2=RainT^2
Generated series RainT2 (ID 7)
? genr AvgTRainT=AvgT*RainT
Generated series AvgTRainT (ID 8)
? genr Punjab=(StateID=1)
Generated series Punjab (ID 9)
? genr Haryana=(StateID=2)
Generated series Haryana (ID 10)
? ols Yield const Time AvgT RainT AvgT2 RainT2 AvgTRainT Punjab Haryana
Model 1: OLS, using observations 1-93
Dependent variable: Yield
coefficient std. error t-ratio p-value
--------------------------------------------------------------
const 0.785161 35.3682 0.02220 0.9823
Time 0.0380004 0.00416873 9.116 3.47e-014 ***
AvgT 0.291484 2.38967 0.1220 0.9032
RainT -0.00751912 0.00503529 -1.493 0.1391
AvgT2 -0.00841372 0.0404213 -0.2082 0.8356
RainT2 1.15559e-06 4.51529e-07 2.559 0.0123 **
AvgTRainT 0.000195514 0.000171757 1.138 0.2582
Punjab 1.58892 0.149044 10.66 2.80e-017 ***
Haryana 0.554109 0.193199 2.868 0.0052 ***
Mean dependent var 2.854064 S.D. dependent var 0.835066
Sum squared resid 9.807372 S.E. of regression 0.341693
R-squared 0.847130 Adjusted R-squared 0.832571
F(8, 84) 58.18564 P-value(F) 4.82e-31
Log-likelihood -27.36115 Akaike criterion 72.72230
Schwarz criterion 95.51570 Hannan-Quinn 81.92563
Excluding the constant, p-value was highest for variable 4 (AvgT)
Appendix: Sample Gretl Scripts, Do File and Sample Outputs 79
? series uh= $uhat

Generated series uh (ID 11)
? genr Luh2=log(uh^2)
Generated series Luh2 (ID 12)
? ols Luh2 const Time AvgT RainT Punjab Haryana
Model 2: OLS, using observations 1-93
Dependent variable: Luh2
--------------------------------------------------------
const -2.15791 13.3502 -0.1616 0.8720
Time 0.0163648 0.0268282 0.6100 0.5435
AvgT -0.0942311 0.436214 -0.2160 0.8295
RainT 0.00179283 0.00131444 1.364 0.1761
Punjab -0.330535 0.874377 -0.3780 0.7063
Haryana -0.328832 1.15343 -0.2851 0.7763
Mean dependent var -3.604329 S.D. dependent var 2.270875

F(5, 87) 1.485777 P-value(F) 0.202705
#ols Luh2 const Time AvgT RainT

? series Lr=$yhat
Generated series Lr (ID 13)
? genr W=1/exp(Lr)
Generated series W (ID 14)
? wls W Yield const Time AvgT RainT AvgT2 RainT2 AvgTRainT Punjab Haryana
Model 3: WLS, using observations 1-93
Variable used as weight: W

--------------------------------------------------------------
const -2.37226 26.3066 -0.09018 0.9284
Time 0.0393390 0.00375067 10.49 6.15e-017 ***
AvgT 0.458150 1.75891 0.2605 0.7951
RainT -0.00402308 0.00547036 -0.7354 0.4641
AvgT2 -0.0106407 0.0293972 -0.3620 0.7183
RainT2 8.59025e-07 4.91175e-07 1.749 0.0840 *
AvgTRainT 9.13326e-05 0.000186397 0.4900 0.6254
Punjab 1.64280 0.164616 9.980 6.38e-016 ***
Haryana 0.541451 0.193409 2.800 0.0063 ***
Statistics based on the weighted data:

F(8, 84) 75.14196 P-value(F) 5.05e-35
Statistics based on the original data:


2. Sample Gretl Script to estimate Just–Pope production function using MLE

#
open G:\CRRDataFiles\NorthRJPData.gdt
genr Punjab=(StateID=1)
genr Haryana=(StateID=2)
genr MaxT2=MaxT^2
genr MinT2=MinT^2
genr RainT2=RainT^2
genr MaxTMinT=MaxT*MinT
genr MaxTRainT=MaxT*RainT
genr MinTRainT=MinT*RainT
ols Yield const Time MaxT MinT RainT MaxT2 MinT2 RainT2 MaxTMinT \
MaxTRainT MinTRainT Punjab Haryana
series uh= $uhat
genr Luh2=log(uh^2)
ols Luh2 const Time MaxT MinT RainT Punjab Haryana
#genr elastTV=$coeff(AvgT)*mean(AvgT)
#genr elastRV=$coeff(RainT)*mean(RainT)
scalar a0=$coeff(const)
scalar a1=$coeff(Time)
scalar a2=$coeff(MaxT)
scalar a3=$coeff(MinT)
scalar a4=$coeff(RainT)
series Lr=$yhat
genr W=1/exp(Lr)
wls W Yield const Time MaxT MinT RainT MaxT2 MinT2 RainT2 MaxTMinT \
MaxTRainT MinTRainT Punjab Haryana
#scalar elastTE=$coeff(AvgT)*mean(AvgT)/mean(Yield)
#scalar elastRE=$coeff(RainTcm)*mean(RainTcm)/mean(Yield)
scalar b0=$coeff(const)
scalar b1=$coeff(Time)
scalar b2=$coeff(MaxT)
scalar b3=$coeff(MinT)
scalar b4=$coeff(RainT)
scalar b5=$coeff(MaxT2)
scalar b6=$coeff(MinT2)
scalar b7=$coeff(RainT2)
scalar b8=$coeff(MaxTMinT)
scalar b9=$coeff(MaxTRainT)
scalar b10=$coeff(MinTRainT)
scalar b11=$coeff(Punjab)
scalar b12=$coeff(Haryana)
set bfgs_richardson on
mle logl=-0.5*log(hv) + log(dnorm(Z))
series mf = b0 + b1*Time +b2*MaxT + b3*MinT + b4*RainT + b5*MaxT2 \
+b6*MinT2 + b7*RainT2 + b8*MaxTMinT + b9*MaxTRainT +b10*MinTRainT \
+ b11*Punjab + b12*Haryana
series hv =exp(a0 +a1*Time +a2*MaxT + a3*MinT +a4*RainT)
series Z =(Yield-mf)/sqrt(hv)
deriv a0 =-0.5*(1-Z^2)
deriv a1 =-0.5*(1-Z^2)*Time
deriv a2 =-0.5*(1-Z^2)*MaxT
deriv a3 =-0.5*(1-Z^2)*MinT
deriv a4 =-0.5*(1-Z^2)*RainT
deriv b0 =Z/sqrt(hv)
deriv b1 =(Z/sqrt(hv))*Time
deriv b2 =(Z/sqrt(hv))*MaxT
deriv b3 =(Z/sqrt(hv))*MinT
deriv b4 =(Z/sqrt(hv))*RainT
deriv b5 =(Z/sqrt(hv))*MaxT2
deriv b6 =(Z/sqrt(hv))*MinT2
deriv b7 =(Z/sqrt(hv))*RainT2
deriv b8 =(Z/sqrt(hv))*MaxTMinT
deriv b9 =(Z/sqrt(hv))*MaxTRainT
deriv b10 =(Z/sqrt(hv))*MinTRainT
deriv b11 =(Z/sqrt(hv))*Punjab
deriv b12 =(Z/sqrt(hv))*Haryana
end mle
Sample Output
gretl version 1.9.91

Current session: 2014-09-30 12:51

#
? open G:.gdt
Read datafile G:.gdt

periodicity: 31, maxobs: 93
observations range: 1:01 to 3:31
Listing 9 variables:
0) const 1) StateID 2) Year 3) Yield 4) Time
5) AvgT 6) MaxT 7) MinT 8) RainT
? genr Punjab=(StateID=1)
Generated series Punjab (ID 9)
? genr Haryana=(StateID=2)
Generated series Haryana (ID 10)
? genr MaxT2=MaxT^2
Generated series MaxT2 (ID 11)
? genr MinT2=MinT^2
Generated series MinT2 (ID 12)
? genr RainT2=RainT^2
Generated series RainT2 (ID 13)
? genr MaxTMinT=MaxT*MinT
Generated series MaxTMinT (ID 14)
? genr MaxTRainT=MaxT*RainT
Generated series MaxTRainT (ID 15)
? genr MinTRainT=MinT*RainT
Generated series MinTRainT (ID 16)
? ols Yield const Time MaxT MinT RainT MaxT2 MinT2 RainT2 MaxTMinT MaxTRainT \par
MinTRainT Punjab Haryana
Model 1: Pooled OLS, using 93 observations
Included 3 cross-sectional units
Time-series length = 31

--------------------------------------------------------------
const -2.69786 39.3139 -0.06862 0.9455
Time 0.0424002 0.00409913 10.34 2.08e-016 ***
MaxT 4.06101 2.46250 1.649 0.1030
MinT -5.54125 1.88343 -2.942 0.0043 ***
RainT 0.00710528 0.00592957 1.198 0.2343
MaxT2 -0.140689 0.0525708 -2.676 0.0090 ***
MinT2 -0.0885605 0.0570486 -1.552 0.1245
RainT2 -3.34060e-07 5.74530e-07 -0.5814 0.5626
MaxTMinT 0.260415 0.0899561 2.895 0.0049 ***
MaxTRainT -0.000970450 0.000258727 -3.751 0.0003 ***
MinTRainT 0.00109967 0.000291935 3.767 0.0003 ***
Punjab 1.53502 0.179225 8.565 6.30e-013 ***
Haryana 0.541471 0.194454 2.785 0.0067 ***

F(12, 80) 50.35151 P-value(F) 3.47e-32
rho 0.363012 Durbin-Watson 1.195375
Excluding the constant, p-value was highest for variable 13 (RainT2)
? series uh= $uhat

Generated series uh (ID 17)
? genr Luh2=log(uh^2)
Generated series Luh2 (ID 18)
? ols Luh2 const Time MaxT MinT RainT Punjab Haryana
Model 2: Pooled OLS, using 93 observations

Included 3 cross-sectional units
Time-series length = 31
Dependent variable: Luh2

----------------------------------------------------------
const -2.17281 12.5729 -0.1728 0.8632
Time 0.00490119 0.0255307 0.1920 0.8482
MaxT 0.0260893 0.407787 0.06398 0.9491
MinT -0.0810797 0.441439 -0.1837 0.8547
RainT 0.000257499 0.00129560 0.1987 0.8429
Punjab -1.92298 0.954147 -2.015 0.0470 **
Haryana -0.499248 1.09491 -0.4560 0.6496
Mean dependent var -3.792231 S.D. dependent var 2.164226

F(6, 86) 2.540886 P-value(F) 0.025916
rho 0.054128 Durbin-Watson 1.814540
Excluding the constant, p-value was highest for variable 6 (MaxT)
#genr elastTV=$coeff(AvgT)*mean(AvgT)
#genr elastRV=$coeff(RainT)*mean(RainT)
? scalar a0=$coeff(const)
Generated scalar a0 = -2.17281
? scalar a1=$coeff(Time)
Generated scalar a1 = 0.00490119
? scalar a2=$coeff(MaxT)
? scalar a3=$coeff(MinT)
Generated scalar a3 = -0.0810797
? scalar a4=$coeff(RainT)
? series Lr=$yhat
Generated series Lr (ID 19)
? genr W=1/exp(Lr)
Generated series W (ID 20)
? wls W Yield const Time MaxT MinT RainT MaxT2 MinT2 RainT2 MaxTMinT MaxTRainT \par
MinTRainT Punjab Haryana
Model 3: WLS, using 93 observations

Variable used as weight: W

-------------------------------------------------------------
const 27.0829 37.3277 0.7255 0.4702
Time 0.0488063 0.00321416 15.18 2.81e-025 ***
MaxT 2.26438 2.42206 0.9349 0.3527
MinT -5.46233 1.86166 -2.934 0.0044 ***
RainT 0.00542750 0.00583105 0.9308 0.3548
MaxT2 -0.116323 0.0543280 -2.141 0.0353 **
MinT2 -0.100708 0.0608617 -1.655 0.1019
RainT2 -4.67472e-07 4.20998e-07 -1.110 0.2702
MaxTMinT 0.270352 0.0995194 2.717 0.0081 ***
MaxTRainT -0.00102960 0.000248053 -4.151 8.22e-05 ***
MinTRainT 0.00126102 0.000297232 4.243 5.90e-05 ***
Punjab 1.43799 0.204759 7.023 6.41e-010 ***
Haryana 0.469765 0.230385 2.039 0.0447 **
Statistics based on the weighted data:

F(12, 80) 77.24636 P-value(F) 8.22e-39
Statistics based on the original data:

Excluding the constant, p-value was highest for variable 8 (RainT)

#scalar elastTE=$coeff(AvgT)*mean(AvgT)/mean(Yield)
#scalar elastRE=$coeff(RainTcm)*mean(RainTcm)/mean(Yield)
? scalar b0=$coeff(const)
Generated scalar b0 = 27.0829
? scalar b1=$coeff(Time)
? scalar b2=$coeff(MaxT)
? scalar b3=$coeff(MinT)
Generated scalar b3 = -5.46233
? scalar b4=$coeff(RainT)
? scalar b5=$coeff(MaxT2)
? scalar b6=$coeff(MinT2)
? scalar b7=$coeff(RainT2)
Generated scalar b7 = -4.67472e-007
? scalar b8=$coeff(MaxTMinT)

? scalar b9=$coeff(MaxTRainT)
? scalar b10=$coeff(MinTRainT)
? scalar b11=$coeff(Punjab)
? scalar b12=$coeff(Haryana)
? set bfgs_richardson on
? mle logl=-0.5*log(hv) + log(dnorm(Z))
? series mf = b0 + b1*Time +b2*MaxT + b3*MinT + b4*RainT + b5*MaxT2 +b6*MinT2 \par
+ b7*RainT2 + b8*MaxTMinT + b9*MaxTRainT +b10*MinTRainT + b11*Punjab + \par
b12*Haryana
? series hv =exp(a0 +a1*Time +a2*MaxT + a3*MinT +a4*RainT)
? series Z =(Yield-mf)/sqrt(hv)
? deriv a0 =-0.5*(1-Z^2)
? deriv a1 =-0.5*(1-Z^2)*Time
? deriv a2 =-0.5*(1-Z^2)*MaxT
? deriv a3 =-0.5*(1-Z^2)*MinT
? deriv a4 =-0.5*(1-Z^2)*RainT
? deriv b0 =Z/sqrt(hv)
? deriv b1 =(Z/sqrt(hv))*Time
? deriv b2 =(Z/sqrt(hv))*MaxT
? deriv b3 =(Z/sqrt(hv))*MinT
? deriv b4 =(Z/sqrt(hv))*RainT
? deriv b5 =(Z/sqrt(hv))*MaxT2
? deriv b6 =(Z/sqrt(hv))*MinT2
? deriv b7 =(Z/sqrt(hv))*RainT2
? deriv b8 =(Z/sqrt(hv))*MaxTMinT
? deriv b9 =(Z/sqrt(hv))*MaxTRainT
? deriv b10 =(Z/sqrt(hv))*MinTRainT
? deriv b11 =(Z/sqrt(hv))*Punjab
? deriv b12 =(Z/sqrt(hv))*Haryana
? end mle
Using analytical derivatives
Tolerance = 1.81899e-012
Function evaluations: 332
Evaluations of gradient: 74
Model 4: ML, using 93 observations
logl = -0.5*log(hv) + log(dnorm(Z))
Standard errors based on Outer Products matrix
estimate std. error z p-value
------------------------------------------------------------
a0 4.95313 8.19414 0.6045 0.5455
a1 0.00893487 0.0260521 0.3430 0.7316
a2 -0.525693 0.358395 -1.467 0.1424
a3 0.407885 0.443294 0.9201 0.3575
a4 0.000351394 0.000951061 0.3695 0.7118
b0 -7.49489 39.7426 -0.1886 0.8504
b1 0.0455863 0.00384772 11.85 2.21e-032 ***
b2 4.20032 2.68857 1.562 0.1182
b3 -5.40478 2.05361 -2.632 0.0085 ***
b4 0.00994506 0.00641027 1.551 0.1208
b5 -0.151541 0.0667235 -2.271 0.0231 **
b6 -0.110422 0.0687613 -1.606 0.1083
b7 -4.62258e-07 5.08444e-07 -0.9092 0.3633
b8 0.287166 0.115094 2.495 0.0126 **
b9 -0.00104226 0.000294570 -3.538 0.0004 ***
b10 0.00109221 0.000317638 3.439 0.0006 ***
b11 1.50951 0.148421 10.17 2.69e-024 ***
b12 0.508953 0.160703 3.167 0.0015 ***
3. Sample Do file and executing it for obtaining MLEs
. use
"G:\FDriveFiles\13StatesCCStudy\DataRegionwiseAnalysed\NorthR\NorthRJPData.dta"
. do "G:\FDriveFiles\RIHNFellowship\JustANDPope-Model-Analysis\myjpmle1.do"
. program define myjpmle1

1. version 1.0
2. args lnf Xb Xd
3. quietly replace `lnf' = -`Xd'/2+ln(normd(($ML_y1-
`Xb')/sqrt(exp(`Xd'))))
4. end
.
end of do-file
. ml model lf myjpmle1 ( yield = time maxt mint raint maxt2 mint2 raint2 maxtmint
maxtraint mintraint punjab harya
> na)( yield = time maxt mint raint)
. ml maximize
initial: log likelihood = -496.43998

alternative: log likelihood = -284.52857
rescale: log likelihood = -187.40136
rescale eq: log likelihood = -148.53285
Iteration 0: log likelihood = -148.53285 (not concave)
Iteration 5: log likelihood = -33.377034
Number of obs = 93
Wald chi2(12) = 817.32
Log likelihood = -9.7477702 Prob > chi2 = 0.0000
------------------------------------------------------------------------------
| Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
eq1 |
time | .045632 .0036608 12.47 0.000 .038457 .052807
maxt | 4.217792 2.395352 1.76 0.078 -.4770114 8.912596
mint | -5.373193 1.734861 -3.10 0.002 -8.773458 -1.972929
raint | .0098913 .0056688 1.74 0.081 -.0012193 .0210019
maxt2 | -.1519693 .0540734 -2.81 0.005 -.2579512 -.0459873
mint2 | -.1113685 .0576984 -1.93 0.054 -.2244554 .0017184
raint2 | -4.61e-07 5.06e-07 -0.91 0.362 -1.45e-06 5.31e-07
maxtmint | .287581 .0932881 3.08 0.002 .1047398 .4704223
maxtraint | -.0010415 .0002469 -4.22 0.000 -.0015255 -.0005576
mintraint | .0010933 .0002731 4.00 0.000 .0005581 .0016285
punjab | 1.511339 .1582188 9.55 0.000 1.201235 1.821442
haryana | .5099374 .1773931 2.87 0.004 .1622534 .8576214
_cons | -8.144124 35.03346 -0.23 0.816 -76.80845 60.5202
-------------+----------------------------------------------------------------
eq2 |
time | .0089379 .0209578 0.43 0.670 -.0321386 .0500144
maxt | -.5244976 .2692086 -1.95 0.051 -1.052137 .0031416
mint | .400668 .3068341 1.31 0.192 -.2007157 1.002052
raint | .0003664 .0008137 0.45 0.653 -.0012285 .0019612
_cons | 5.076084 5.472121 0.93 0.354 -5.649077 15.80124
------------------------------------------------------------------------------
. estat ic
-----------------------------------------------------------------------------
Model | Obs ll(null) ll(model) df AIC BIC
-------------+---------------------------------------------------------------
. | 93 . -9.74777 18 55.49554 101.0823
-----------------------------------------------------------------------------
Note: N=Obs used in calculating BIC; see [R] BIC note
Chapter 5
Climate Change Impacts on Rice Yield
in Northern Region of India

5.1 Introduction
The north region for the present study covers three states: Punjab, Haryana and
Uttar Pradesh. The districts selected from the three states were: Sangrur, Karnal and
Chandauli, respectively. As stated in the methodology chapter, the present study is
based on (i) primary data: farm survey among 100 farmers from each of the three
districts and (ii) secondary data: time series data (1980–2010) on productivity (yield)
of rice and climate variables from the three districts. The location of the study area
is given in Fig. 5.1.
5.2 Socio-economic Profiles of the Farmers
All the 300 farmers surveyed were males. Out of them, 200 farmers in Punjab and
Haryana have crop with supplemental irrigation typology, while all the 100 farmers
in Uttar Pradesh have irrigated crop-based farming. The average ages of the farmers
in the three states are 39, 40 and 53 years, respectively. So, farmers in Uttar Pradesh
are relatively older than those from Punjab and Haryana. About 23% of the farmers
have no formal education, 21% have studied up to primary level, and the rest 56%
have studied up to secondary level (Fig. 5.2); 86% of the farmers from Uttar Pradesh
have education up to secondary level.
5.3 Farm Information
Farmers in the study region own land and wells. They use different methods of
irrigation like drip, sprinkler and others, and they invest on digging borewells. They
cultivate rice crop only during the kharif season. Farmers in Punjab and Haryana
88 5 Climate Change Impacts on Rice Yield in Northern …
86
90
80
Percentage of farmers
70 Punjab
56.3
60 Haryana
45
50
38 UP
40 33
29 29 Total
26
30 22.7 21
20 13
10 1
0
No formal Primary Level Secondary
EducaƟon Level
Fig. 5.2 Educational status of the sample farmers

5.3 Farm Information 89
Table 5.1 Summary State Mean S.D. Min Max

statistics of irrigated land
owned (hectares) Punjab 4.98 6.35 0.40 60.70
Haryana 4.01 2.79 0.61 16.19
UP 2.10 1.74 0.40 8.09
Overall 3.68 4.29 0.40 60.70
S.D standard deviation
Table 5.2 Distribution of Number of farmers owning

number of borewells owned
across states No. of borewells Punjab Haryana UP Total
0 0 0 1 1
1 59 73 92 224
2 34 25 7 66
3 6 2 0 8
>3 1 0 0 1
derive income from rice and other crops only during kharif season as they are not
raising any crop during rabi and summer. Though farmers in Uttar Pradesh could get
income from rice crop only in kharif, they could get income from other crops also
during all the seasons.
(a) Physical assets:
(i) Land owned: All the farmers own only irrigated land. Table 5.1 gives the sum-
mary statistics.
The mean area under irrigation varies from 2.10 to 4.98 ha across the three states,
and the overall mean is 3.68 ha. Landholdings in Punjab have the largest standard
deviation of 6.35 ha, and the minimum and maximum holdings are, respectively,
0.40 and 60.7 ha.
(ii) Wells owned: No farmer possesses open well, and they own only borewells.
Table 5.2 gives the frequency distribution of number of borewells owned by the
farmers.
Out of 300 farmers, 224 of them, that is, 75% of the farmers, possess one borewell,
and in Uttar Pradesh, 92% of the farmers possess on borewell each. Also, overall,
22% of the farmers possess 2 borewells each.
(iii) Sources of irrigation: Farmers have two sources of irrigation, viz. canal and
well.
All farmers of Haryana have well and that is the main source of their irrigation,
whereas in Punjab, 61 of them use wells and the remaining 39 use both canal and
well. But in Uttar Pradesh, 86% of the farmers use both sources and only 13% use
well (Fig. 5.3).
100 0
80 39
Canal+Well
60 86
100 Well
40
61 Canal
20
13
0 0 0 1
Punjab Haryana UP
Fig. 5.3 Frequency distribution of sources of irrigation
(iv) Investment in irrigation: None of the farmers in Haryana made any investment
on irrigation structures and farmers in Punjab and Uttar Pradesh invested in
digging borewells. On the average, farmers in Punjab invested Rs. 34,800 and
farmers in Uttar Pradesh spent, on the average, Rs. 61,630 in digging borewells.
(v) Area under rice: All the farmers have cultivated rice only during the kharif
season. During this season, on the average every farmer cultivates rice in 4.43,
3.84 and 2.02 ha in the three states, respectively. Similarly, other crops are
grown, respectively, in 0.071, 0.41 and 0.53 ha in the three states during the same
season. No crop is cultivated by the sample farmers during rabi and summer
season in Punjab and Haryana, whereas in Uttar Pradesh, on the average, other
crops are grown in 2.1 and 1.34 ha, respectively, during these two seasons.

as Perceived by Farmers
With a view to have an in-depth understanding of farmers’ perception on the climate

change, its effects on agriculture and various adaptation strategies/techniques fol-
lowed by the farmers, detailed survey was conducted and the results are discussed
below.
5.4 Climate Change Events, Shocks and Their Effects … 91
5.4.1 Perception of Farmers on Climate Change Events
Farmers were asked to rank the 16 events as listed in the methodology chapter
according to their perceptions of climate change (CC). To find what the farmers
perceive as the most significant effects of climate change, Garrett ranking technique
was applied. Farmers in Punjab and Uttar Pradesh ranked all the 16 CC events, while
farmers in Haryana ranked only 8 CC events. Hence, the rankings of these two states,
viz. Punjab and Uttar Pradesh, were combined to represent northern region. Garrett’s
ranking technique, as described in the methodology chapter, was used. Tables 5.3
and 5.4 give the results of the Garrett Ranking. Table 5.4 shows that untimely rain
(UR) is the most important climate change event (Rank 1) perceived by rice farmers
in northern region. It earned an average Garrett Score of 66.385 as given in Table 5.4.
It was given Ranks 1, 2 and 3 by 31, 41 and 47 farmers, respectively. The second
and third important climate change events are, respectively, incidence of pests and
diseases (IPD) and depletion of groundwater (DGW).
The results for Haryana are presented in Tables 5.5 and 5.6. Depletion in ground-
water is the foremost concern for Haryana farmers as all of them have given Rank
1 to it. The same CC event has been given Rank 3 by farmers of Punjab and Uttar
Pradesh. Incidence of pest and disease and untimely rains were given Ranks 2 and
3 by Haryana farmers. The same two CC events were given Ranks 2 and 1 by the
farmers of Punjab and Uttar Pradesh. Thus, as a whole, these three events are the
most important CC events as perceived by the farmers belonging to the northern
region. Similar Garrett ranking analysis was done separately for Punjab and Uttar
Pradesh. The results are presented in Appendix (Tables 5.20, 5.21, 5.22, 5.23).
Friedman Rank Test was applied to test for differences between rankings of the
climate change events (a listed in Table 5.3) by the farmers. The methodology as
described in Chap. 4 was followed. The calculated value of Friedman’s F statistics
was 672.9, while the table value (corresponding to 15 degrees of freedom at 5%
level of significance) is 24.996. This shows that the farmers did not consistently rank
the events. Thus, the farmers’ perceptions of climate change effects differ across the
sample chosen. The same conclusion was arrived by applying Kendall’s coefficient
of concordance also. Similar analysis was done for individual states, viz. Punjab,
Haryana and Uttar Pradesh. The results show that there is no consistency in the
rankings of the events by the farmers.
5.4.2 Effects of Shocks
Farmers were questioned on their perception about the effects of climatic shocks.
They were asked to Rank 7 effects of climatic shocks, as described in methodology
chapter. Farmers ranked decline in crop yield (DCY) as the most important shock
(Tables 5.7 and 5.8). Out of 300 farmers, 199 of them have given Rank 1 to this
effect. The second and third ranks are given to loss of income (LI) and loss of assets
92
Table 5.3 Factors influencing farmers perception of climate change—Garrett ranking: northern region
Climate change event Rank Total no. of farmers
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 2 1 2 6 11 12 9 10 13 9 7 11 13 25 49 20 200
Hailstorm 48 17 11 6 3 4 5 6 3 7 8 12 13 21 22 14 200
Flood 9 36 14 9 1 4 8 5 9 4 7 5 17 16 23 33 200
AD 0 5 5 6 7 6 15 20 23 18 20 14 25 15 13 8 200
IPD 25 32 17 34 15 24 22 11 7 1 2 1 4 2 1 2 200
IFM 3 6 7 3 13 11 13 4 3 6 8 11 5 10 16 81 200
UR 31 41 47 10 14 13 13 6 1 5 3 6 3 4 2 1 200
IW 17 23 26 30 13 16 14 8 9 7 4 4 4 10 10 5 200
TFH 5 15 12 17 15 15 7 14 11 15 24 21 19 5 4 1 200
TFL 1 8 19 13 19 9 6 10 7 13 11 30 30 16 7 1 200
SSDI 0 1 4 7 8 12 28 33 41 21 18 10 7 6 3 1 200
CCP 0 0 1 2 6 8 12 14 11 18 26 30 16 19 24 13 200
CLA 0 0 0 1 4 9 14 19 32 27 21 16 17 30 7 3 200
CFI 0 1 2 24 20 27 14 18 16 24 11 16 14 4 7 2 200
CF 4 5 24 19 33 17 9 8 9 11 24 7 10 4 7 9 200
DGW 55 9 9 13 18 13 11 14 5 14 6 6 3 13 5 6 200
Total 200 200 200 200 200 200 200 200 200 200 200 200 200 200 200 200 3200
Garrett score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
5 Climate Change Impacts on Rice Yield in Northern …
Table 5.4 Rankings of factors representing climate change using Garrett ranking: northern region
Climate Rank Total Average Rank
change 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
event
(factors) Total score
Drought 172 76 140 390 671 696 495 520 624 405 294 429 455 750 1176 280 7573 37.865 15
Hailstorm 4128 1292 770 390 183 232 275 312 144 315 336 468 455 630 528 196 10654 53.27 5
Flood 774 2736 980 585 61 232 440 260 432 180 294 195 595 480 552 462 9258 46.29 11
AD 0 380 350 390 427 348 825 1040 1104 810 840 546 875 450 312 112 8809 44.045 12
IPD 2150 2432 1190 2210 915 1392 1210 572 336 45 84 39 140 60 24 28 12827 64.135 2
IFM 258 456 490 195 793 638 715 208 144 270 336 429 175 300 384 1134 6925 34.625 16
UR 2666 3116 3290 650 854 754 715 312 48 225 126 234 105 120 48 14 13277 66.385 1
IW 1462 1748 1820 1950 793 928 770 416 432 315 168 156 140 300 240 70 11708 58.54 4
5.4 Climate Change Events, Shocks and Their Effects …
TFH 430 1140 840 1105 915 870 385 728 528 675 1008 819 665 150 96 14 10368 51.84 7
TFL 86 608 1330 845 1159 522 330 520 336 585 462 1170 1050 480 168 14 9665 48.325 10
SSDI 0 76 280 455 488 696 1540 1716 1968 945 756 390 245 180 72 14 9821 49.105 9
CCP 0 0 70 130 366 464 660 728 528 810 1092 1170 560 570 576 182 7906 39.53 14
CLA 0 0 0 65 244 522 770 988 1536 1215 882 624 595 900 168 42 8551 42.755 13
CFI 0 76 140 1560 1220 1566 770 936 768 1080 462 624 490 120 168 28 10008 50.04 8
CF 344 380 1680 1235 2013 986 495 416 432 495 1008 273 350 120 168 126 10521 52.605 6
DGW 4730 684 630 845 1098 754 605 728 240 630 252 234 105 390 120 84 12129 60.645 3
93
Table 5.5 Factors influencing farmer’s perception of climate change—Garrett ranking: Haryana
Climate change Rank Total no.
event 1 2 3 4 5 6 7 8 of farmers
AD 0 0 0 0 0 67 33 0 100
IPD 0 90 0 0 0 8 2 0 100
IFM 0 8 0 0 0 0 65 27 100
UR 0 2 90 0 8 0 0 0 100
IW 0 0 2 78 0 0 0 20 100
TFH 0 0 8 0 92 0 0 0 100
TFL 0 0 0 22 0 25 0 53 100
DGW 100 0 0 0 0 0 0 0 100
Total 100 100 100 100 100 100 100 100 800
Garrett score 80 67 60 53 47 40 33 20
Table 5.6 Rankings of factors representing climate change using Garrett ranking: Haryana
Climate Rank Rank
change 1 2 3 4 5 6 7 8 Total Average
event score score
Total score
AD 0 0 0 0 0 2680 1089 0 3769 37.69 6
IPD 0 6030 0 0 0 320 66 0 6416 64.16 2
IFM 0 536 0 0 0 0 2145 540 3221 32.21 8
UR 0 134 5400 0 376 0 0 0 5910 59.1 3
IW 0 0 120 4134 0 0 0 400 4654 46.54 5
TFH 0 0 480 0 4324 0 0 0 4804 48.04 4
TFL 0 0 0 1166 0 1000 0 1060 3226 32.26 7
DGW 8000 0 0 0 0 0 0 0 8000 80 1
(LA), respectively. This is understandable because decline in crop yield will reflect
in income and asset generation.
5.4.3 Mitigation Strategies Used by Farmers
Farmers apply several mitigation strategies to tide over the negative impacts of climate
change. Table 5.9 provides a summary. It shows that 100% of the farmers in all the
three states have restricted rice-growing season to one. Farmers in Uttar Pradesh have
used only this strategy. All the farmers in Punjab and Haryana have changed crop
variety and planting dates. Also 52% of the farmers in Punjab have sold livestock.
All other strategies are followed by less than 6% of the farmers.
Table 5.7 Rankings of effects of climatic shocks by farmers

Rank
Factors 1 2 3 4 5 6 7 Total
DCY 199 26 14 12 8 20 21 300
LA 7 41 81 30 36 81 24 300
LI 61 136 75 8 5 10 5 300
FIS 5 40 21 74 77 43 40 300
DL 5 16 57 100 68 20 34 300
DC 1 3 25 49 47 45 130 300
DH 22 38 27 27 59 81 46 300
Total 300 300 300 300 300 300 300 2100
Garrett score 79 66 57 50 43 34 21
Table 5.8 Garrett ranking of effects of climatic shocks

Factors Rank
1 2 3 4 5 6 7 Total Average Rank
score score
Total score
DCY 15721 1716 798 600 344 680 441 20,300 67.67 1
LA 553 2706 4617 1500 1548 2754 504 14,182 47.27 3
LI 4819 8976 4275 400 215 340 105 19,130 63.77 2
FIS 395 2640 1197 3700 3311 1462 840 13,545 45.15 5
DL 395 1056 3249 5000 2924 680 714 14,018 46.73 4
DC 79 198 1425 2450 2021 1530 2730 10,433 34.78 7
DH 1738 2508 1539 1350 2537 2754 966 13,392 44.64 6
5.4.4 Opinions of Farmers on Government Initiatives
From time to time, both Central and State governments implement several mitigation
and long-term measures to offset the negative effects of climate change on agricul-
ture. With a view to know the awareness of these measures among the farmers, they
were questioned on seven government initiatives (as described in the methodology
chapter). They were asked to state their replies as either ‘yes’, or ‘no’ or ‘not aware’.
Figure 5.4 gives a pictorial representation of the awareness on these initiatives by the
farmers belonging to the entire region. About 93% of the farmers are well aware of
introduction of new crop varieties, 60% have good understanding of water regula-
tion by government through irrigation department, and 53% of the farmers have good
understanding on the subsidised inputs (like fertilizers, drip and machinery). Only
29% of the farmers participated in the field demos of the listed adaptation practices.
Further, weather-based cropping insurance schemes are not popular among the farm-
Table 5.9 Frequency distribution of mitigation strategies followed by farmers

Strategies Punjab Haryana UP Total no. of farmers
Did nothing 0 0 0 0
Changed crop variety 100 100 0 200
Changed planting date 100 100 0 200
Left land fallow 1 0 0 1
Sold part of land 0 0 0 0
Leased out part of land 5 0 0 5
Restricted number of seasons 100 100 100 300
for rice-growing to 1
Sold livestock 52 0 0 52
Shifted to other crops 1 0 0 1
Borrowed money from 4 0 0 4
relatives/others
Received assistance from 5 0 0 5
government/NGOs
Less food consumption 0 0 0 0
Shifted to non-farm 0 0 2 2
employment
Outmigration to cities 0 0 2 2
ers as 59% of them are not aware of it. Similarly, only 5% of the farmers know about
establishment of village knowledge centres. Thus, we conclude that except (i) water
regulation through irrigation department, (ii) introduction of new crop varieties and
(iii) subsidised inputs, all other government initiatives are not much popular among
farmers. This conclusion calls for formulating suitable popularisation mechanisms
by the government. Table 5.10 provides a summary of the opinions on government
initiatives, state-wise.
Chi-square test was applied to check if the opinions differ across the states and the
results show that the opinions of the farmers on government initiatives differ strongly
between the three states. This may be because the government initiatives depend on
local natural resources availability and other factors.
5.5 Analysis of the Effect of Climate Change on Rice Yield
5.5.1 Data
The objective of the study is to estimate the effects of climate variables on the yield
of rice crop using historical data. Hence, we used a panel data set of yield and
climate variables pertaining to 31 years from 1980 to 2010. The data on temperature
corresponds to the growing season of rice variety in the respective states, and the
time series data on rainfall was collected for the entire year to account for rainfall
Table 5.10 Farmers opinions on government initiatives
Govt. initiatives Punjab Haryana UP Total Chi- df p-value
square
Yes No NA Yes No NA Yes No NA Yes No NA
Water regulation 81 19 0 100 0 0 0 31 69 181 50 69 260.8 4 <0.00001
through irrigation
department
Introduction of new 79 20 1 100 0 0 99 1 0 278 21 1 41.3 4 <0.00001
crop varieties
Field demos on the 37 52 11 27 21 52 24 72 4 88 145 67 90.7 4 <0.00001
listed adaptation
practices
Organised field days 31 47 22 3 46 51 10 83 7 44 176 80 81.6 4 <0.00001
on the adaptation
practices
Subsidised inputs 10 48 42 51 41 8 100 0 0 161 89 50 180.7 4 <0.00001

(fertilizer, drip and
machinery)
Weather-based crop 3 43 54 46 33 21 0 0 100 49 76 175 175.1 4 <0.00001
insurance schemes
Establishment of 3 40 57 11 56 33 0 99 1 14 195 91 94.5 4 <0.00001
village knowledge
centres, climate cell,
etc.
NA not aware; The p-values of the chi-square test show that the opinions of farmers strongly differ between states
97
Water regulation through irrigation Introduction of new crop

department No varieties
7%
Not
aware
23% Yes
No 60%
Yes
17% 93%
Field demos on the listed Organised field days on the

Not adaptation practices adaptation practices
aware Yes Not Yes
22% aware 15%
29%
27%
No No
49% 58%
Not Subsidised inputs Weather based crop insurance

aware schemes
17% Yes
Yes Not 16%
53%
aware No
No 59%
30% 25%
Establishment of village
knowledge centres Yes
Not 5%
aware
30%
No
65%
Fig. 5.4 Opinions of farmers of northern region on various government initiatives
falling directly on the crop and interseasonal water accumulation within a year (Iski
and Devadoss 2006; Kim and Pang 2009). The summary statistics of all the variables
are discussed below:
(a) Yield (tons per ha)
The yield data (Table 5.11 and Fig. 5.5) shows an upward trend in all the three
states over years. Trend equations fitted to yield data (Table 5.11) show that all the
coefficients of trend are strongly significant (1% level). Further, Punjab consistently
maintained higher yield during all the 31 years (Fig. 5.5).
(b) Maximum Temperature (°C)
The maximum temperature had an overall average of 33.82 °C (Table 5.12). Punjab
had the highest mean maximum temperature of 34.9 °C followed by UP (33.85 °C).
5.5 Analysis of the Effect of Climate Change on Rice Yield 99
Table 5.11 Descriptive statistics of yield and trend equations (tons/ha)

N Mean Min Max Std. Trend equation R2
Dev.
Punjab 31 3.72 2.74 4.70 0.58 Y = 2.7760 + 0.0592*** t 0.8574
Haryana 31 2.75 2.22 3.65 0.33 Y = 2.4132 + 0.0208*** t 0.3279
UP 31 2.09 1.24 3.33 0.55 Y = 1.4847 + 0.0380*** t 0.3992
Overall 93 2.85 1.24 4.70 0.84 Y = 1.4637 + 0.0393*** t + 0.8305
1.6306D1 + 0.6521D2
***Significant at 1% level; D1 and D2 are dummy variables representing Punjab and Haryana states
Fig. 5.5 Trend in yield of rice in northern region of India
Table 5.12 Descriptive statistics of maximum temperature (o C)

N Mean Minimum Maximum Std. Dev.
Punjab 31 34.90 34.09 36.60 0.58
Haryana 31 32.72 31.65 35.49 0.76
UP 31 33.85 31.33 35.05 0.76
Overall 93 33.82 33.11 35.64 0.51
Also the variability in maximum temperature is least for Punjab. Its maximum tem-
perature is consistently higher than those of Haryana and UP in most of the years
(Fig. 5.6).
37
Punjab
Haryana
UP
Maximum Temperature (Centigrade)
36
35
34
33
32
31
1980 1985 1990 1995 2000 2005 2010
Year
Fig. 5.6 Trend in maximum temperature in northern region of India
Table 5.13 Descriptive statistics of minimum temperature (°C)

Punjab 31 24.14 23.20 26.36 0.66
Haryana 31 22.94 21.80 23.97 0.49
UP 31 25.10 23.27 26.28 0.63
Overall 93 24.06 23.17 25.21 0.45
(c) Minimum Temperature (°C)
The minimum temperature data shows that Uttar Pradesh has the highest average min-
imum temperature compared to other two states. Its average temperature is 25.1 °C
(Table 5.13). It is consistently higher than those of Punjab and Haryana in most of
the years. Haryana has the lowest minimum temperature in all the 31 years (Fig. 5.7).
(d) Average Temperature (°C)
Table 5.14 shows that the overall average temperature of the region was 28.94 °C.
Punjab had the maximum average temperature of 29.52 °C, and Haryana had the
lowest average temperature of 27.83 °C. The standard deviation varies between 0.62
and 0.70 °C across the three states. Figure 5.8 shows that the average temperature of
Haryana was consistently lower than those of Punjab and Uttar Pradesh.
26.5
Punjab
Haryana
26 UP
25.5
Minimum Temperature
25
24.5
24
23.5
23
22.5
22
21.5
1980 1985 1990 1995 2000 2005 2010
Year
Fig. 5.7 Trend in minimum temperature in northern region of India
Table 5.14 Descriptive statistics of average temperature (°C)

Punjab 31 29.52 28.65 31.48 0.62
Haryana 31 27.83 26.73 29.73 0.63
UP 31 29.48 27.30 30.67 0.70
Overall 93 28.94 27.13 31.48 0.97
(e) Rainfall (mm)

Annual rainfall was highest for UP followed by Haryana. Even though the average
rainfall for Punjab is the least (482 mm), it has highest annual variability in rainfall
with 216 mm (Table 5.15). Rainfall in UP is consistently higher than those of Punjab
and Haryana during most of the years (Fig. 5.9), and similarly the rainfall of Punjab
is lower than those of the other two states in most of the years. Thus, we find that
while Punjab has higher temperature than those of Haryana and UP, it has lower
rainfall compared to these two states.
31.5
Punjab
Haryana
31 UP
30.5
Average Temperature
30
29.5
29
28.5
28
27.5
27
1980 1985 1990 1995 2000 2005 2010
Year
Fig. 5.8 Trend in average temperature in northern region of India
Table 5.15 Descriptive statistics of average rainfall (mm)

Mean Minimum Maximum Std. Dev.
Punjab 482 217 1270 216
Haryana 600 271 1177 193
UP 991 671 1387 174
Overall 691 489 1085 132
5.5.2 Test of Normality of Yield Data
The normality of yield data was tested using Shapiro–Wilk W test and Lilliefors test.
The tests showed that the null hypothesis that the data is normal cannot be rejected.
The computed value of Shapiro–Wilk W test was 0.9748 with a p-value of 0.0685.
Since this value is greater than 0.05, the null hypothesis was not rejected. Similarly,
the test statistic for Lilliefors test was 1.2875 with a p-value of 0.08 which is also
greater than 0.05. So both tests did not reject the null hypothesis that yield data is
normal.
1400
Punjab
Haryana
UP
1200
1000
Annual Rainfall (mm)
800
600
400
200
1980 1985 1990 1995 2000 2005 2010
Year
Fig. 5.9 Trend in annual rainfall in northern region of India
Table 5.16 Panel root test results using Fisher’s tests

Variable Test p-value Decision Test p-value Decision
statistic for statistic for
individual individual
effects and effect with
no trend trend
Yield 10.5839 0.1021 Accept H0 14.6465 0.0232** Reject H0
Rainfall 27.6310 0.0001*** Reject H0 27.6310 0.0000*** Reject H0
Maximum 27.6310 0.0001*** Reject H0 27.6310 0.0000*** Reject H0
temperature
Minimum 27.6310 0.0001*** Reject H0 27.6310 0.0000*** Reject H0
temperature
Note *** and **denote significant at 1 and 5% levels
5.5.3 Panel Root Test: Fisher’s Tests
Since we have data on rice yield from three states for 31 years, we have a panel data
set. So it must be subjected to panel root testing before modelling. As stated in
Chap. 4, Fisher’s panel root test was applied and the results are presented in
Table 5.16.
Table 5.17 Estimation results of Just–Pope production function

Mean function Linear Quadratic
Coefficient S. error Coefficient S. error
Constant (β0 ) 3.2099** 1.5152 −8.1441 35.0335
Trend (β1 ) 0.0442*** 0.0038 0.0456*** 0.0037
MaxT (β2 ) 0.0706 0.0576 4.2178* 2.3954
MinT (β3 ) −0.1653** 0.0653 −5.3732*** 1.7349
RainT (β4 ) −0.0001 0.0002 0.0099* 0.0057
MaxT 2 (β5 ) – – −0.1520*** 0.0541
MinT 2 (β6 ) – – −0.1114* 0.0577
RainT 2 (β7 ) – – 0.0000 0.0000
MaxT MinT (β8 ) – – 0.2876*** 0.0933
MaxT RainT (β9 ) – – −0.0010*** 0.0002
MinT RainT (β10 ) – – 0.0011*** 0.0003
Punjab (β11 ) 1.3733*** 0.1396 1.5113*** 0.1582
Haryana (β12 ) 0.4389** 0.1720 0.5099*** 0.1774
Variance function
Constant 0.5333 6.2219 5.0761 5.4721
Time 0.0448** 0.0226 0.0089** 0.0210
MaxT −0.3251 0.2813 −0.5245* 0.2692
MinT 0.2715 0.3107 0.4007 0.3068
RainT 0.0011 0.0008 0.0004 0.0008
Model statistics
LL −19.1135 −9.7478
AIC 62.2270 55.4955
BIC 92.6181 101.0823
Note ***, ** and *denote significant at 1, 5 and 10% levels
The test results confirmed that all the variables in panel data do not have unit roots
and hence they are stationary.
5.5.4 Estimation of Just–Pope Production Function
As discussed in the methodology chapter, to account for both mean and variance
in yield, Just–Pope production function (Just and Pope 1978 and 79) with yield as
dependent variable and maximum and minimum temperatures, rainfall, their squares
and interaction terms and dummy variables for states were used as independent
variables. Table 5.17 gives the results of the estimated empirical models.
2.9
2.8
2.7
Yield in tonnes per ha
2.6
2.5
2.4
2.3
2.2
2.1
2
32 32.5 33 33.5 34 34.5 35 35.5 36
Maximum Temperature (oC)
Fig. 5.10 Nonlinear relationship between yield and maximum temperature
The quadratic model has the least value for AIC and so is selected. Table 5.17
shows that maximum temperature has a significant positive effect on rice yield,
whereas minimum temperature has very strong significant negative effect. Similarly,
rainfall has significant positive effect on yield. All the square terms (except that of
rainfall) and all interaction terms are significant. The significance of the two square
terms and all the three interaction terms imply that nonlinear effects of climate
variables are significant and so important (Figs. 5.10, 5.11 and 5.12). The coefficients
of the two dummies for Punjab and Haryana states are positive and strongly significant
implying that the mean yields between the states differ statistically. The coefficient
for Punjab is 1.5113 and that of Haryana is 0.5099. This means that, on the average,
difference in mean yield (per ha) between these two states is 1.0014 assuming all
other variables fixed. The positive and significant coefficient for trend implies the
significance of technological progress in rice production such as improved irrigation
methods, use of high yielding varieties and better control of pests and diseases over
the period under study.
With respect to variance function, coefficients of trend and maximum temperature
are significant. The coefficient of trend is positive implying that it is a risk-increasing
variable. This means that variability in yield increases over years. But we have
already seen that the coefficient of trend in the mean yield function is positive and
significant which implies that mean yield increases over time. So, we conclude that
over years increase in mean yield increases variability in yield also. These results
coincide with the findings of Palanisami et al. (2011), Ranganathan (2009), Isik and
Devadoss (2006) and Anderson and Hazell (1987). Since the coefficient of maximum
temperature is negative, it is a risk-decreasing variable for rice yield. So, increase
2.8
2.6
2.4
2.2
1.8
22 22.5 23 23.5 24 24.5 25 25.5 26
Minimum Temperature (oC)
Fig. 5.11 Nonlinear relationship between yield and minimum temperature
2.9
2.85
2.8
2.75
2.7
400 500 600 700 800 900 1000 1100

Rainfall(mm)
Fig. 5.12 Nonlinear relationship between yield and rainfall

Table 5.18 Elasticities of climatic variables at means

Yield function Climate variable Elasticity
Mean Maximum temperature 1.6240
Minimum temperature −2.1055
Rainfall 0.0803
Variance Maximum temperature −17.7406
Minimum temperature 9.6402
Rainfall 0.2532
in maximum temperature decreases the yield variability. Coefficients of minimum

temperature and rainfall though positive are not statistically significant.
The elasticities for all climatic variables at mean levels are presented in Table 5.18
for both mean and variance functions. The elasticity for the maximum temperature is
1.624 implying that 1% increase in maximum temperature will induce an increase of
1.62% in mean yield. Similarly, the elasticity coefficient for minimum temperature
is −2.1055 and so 1% increase in minimum temperature will reduce the mean yield
by 2.11%. The elasticity coefficient for rainfall is 0.0803 which is very small, and so
1% increase in rainfall will induce an increase of 0.08% in mean yield. When all the
three variables are simultaneously increased by 1%, we can expect a small decrease
of 0.4% in yield.
For the variance function, the elasticity coefficient of maximum temperature is −
17.74 implying a reduction in variability in yield of 17.7 for 1% increase in maximum
temperature. Similar interpretations can be made for the elasticities of minimum tem-
perature and rainfall, even though their coefficients are not significant. All together,
we can expect a decrease of 7.85% in yield when all the three variables are increased
by 1%. So we conclude that all the three variables put together are risk decreasing
in this sense.
The mean function can be used to compute the optimal levels of climate variables
which will give maximum yield. Using differential calculus, it can be shown that the
optimal levels of MaxT, MinT and RainT are obtained by solving the following set
of simultaneous equations:
∂y
= 2β5 MaxT + β8 MinT + β9 RainT = −β2
∂(MaxT )
∂y
= β8 MaxT + 2∗ β6 MinT + β10 RainT = −β3
∂(MinT )
∂y
= β9 MaxT + β10 RainT + 2β7 RainT = −β4
∂(RainT )
Solving these equations, the optimal levels are obtained as MaxT = 36.86 °C; MinT
= 26.61 °C and RainT = 639.3 mm.
Table 5.19 Climate change impacts on percentage of mean yield and yield variability
Scenario Period Percentage change in
T max T min Rainfall Mean yield Yield
variability
RCP4.5 2021–2050 1.3232 4.9525 17.5011 −6.87 28.70
2071–2100 6.5661 11.5596 11.7217 −12.73 −2.08
RCP8.5 2021–2050 0.7417 5.5802 25.1846 −8.52 47.01
2071–2100 9.7056 19.3691 44.8659 −21.42 25.90
5.5.5 Climate Change Scenarios and Rice Yield Changes
The estimated elasticities given in Table 5.18 were used to estimate the impacts of
climate change on mean yield and yield variability of yield. For this purpose, two
climate change scenarios, viz. RCP4.5 and RCP8.5 (as discussed in Chap. 3), were
considered. The results are presented in Table 5.19.
The table shows that, as per RCP4.5 and RCP8.5 scenarios, there will be substan-
tial reduction in percentages of rice yield during the periods 2021–50 and 2071–2100.
Under both scenarios, percentage reduction in the later period will be more. The per-
centage reduction in mean yield varies from 6.87 to 8.52 during the first period as
per the two scenarios, and it varies from 12.73 to 21.42% during the second period.
Yield variability as estimated by RCP8.5 scenario is more than that predicted by
RCP4.5 scenario for both periods. The first scenario predicts an increase of 28.7%
yield variability during 2021–2050, while the second scenario predicts an increase of
47% for the same period. For the second period, RCP4.5 predicts a small reduction
in yield variability of 2%, whereas RCP8.5 scenario predicts an increase in yield
variability of 25.9%.
5.6 Summary
Punjab and Uttar Pradesh farmers perceive untimely rain followed by incidence of
pests and diseases and depletion of groundwater as most important climate change
events, while, for farmers of Haryana, depletion in groundwater is the foremost
concern. Farmers ranked decline in crop yield as the most important climate change
shock. All farmers in the three states have restricted rice-growing season to one.
In addition, all the farmers in Punjab and Haryana have changed crop variety and
planting dates as mitigation strategies.
Farmers are well aware of introduction of new crop varieties, good understand-
ing of water regulation by government and on the subsidised inputs. But, not many
farmers participated in the field demos, weather-based cropping insurance schemes
5.6 Summary 109
and other government initiatives. This conclusion calls for formulating suitable pop-
ularisation mechanisms by the government.
Based on the historical data, yield of rice shows an upward trend in all the three
states over years. Punjab had the highest mean maximum temperature followed by
UP, while UP had the highest minimum temperature. Annual rainfall in UP was
higher than those of the other two states. Maximum temperature has a significant
positive effect on rice yield, whereas minimum temperature has very strong sig-
nificant negative effect. Similarly, rainfall has significant positive effect on yield.
Also, there is strong evidence of nonlinear effects of climate variables on yield. The
two climate change scenarios RCP4.5 and RCP8.5 predict substantial reduction in
rice yield during the periods 2021–50 and 2071–2100. Yield variability as estimated
by the second scenario (RCP8.5) is more than that predicted by the first scenario
(RCP4.5) for both periods.
Appendix
110
Table 5.20 Factors influencing farmers perception of climate change—Garrett ranking: Punjab
Factors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Total
Drought 1 1 1 2 3 6 4 8 9 6 2 5 6 13 14 19 100
Hailstorm 0 0 0 2 0 1 2 3 1 5 4 12 13 21 22 14 100
Flood 0 1 0 0 0 2 0 0 5 0 2 5 13 16 23 33 100
AD 0 5 3 6 6 4 7 3 6 8 10 7 18 4 6 7 100
IPD 13 19 12 19 5 8 5 2 4 1 2 1 4 2 1 2 100
IFM 3 6 7 3 13 11 13 4 3 6 8 10 5 5 1 2 100
UR 13 19 20 7 6 6 5 3 1 3 3 4 3 4 2 1 100
IW 8 15 14 14 8 5 4 2 4 4 2 4 3 6 4 3 100
TFH 5 15 12 16 15 13 5 4 5 4 0 0 1 1 3 1 100
TFL 1 8 19 13 19 9 4 7 3 4 2 3 4 0 3 1 100
SSDI 0 1 4 5 7 5 11 11 14 10 10 9 4 6 2 1 100
CCP 0 0 1 2 6 8 11 14 9 10 11 16 3 6 0 3 100
CLA 0 0 0 1 4 6 13 12 15 9 12 6 5 10 5 2 100
CFI 0 1 0 1 2 10 6 12 11 18 9 10 8 3 7 2 100
CF 1 0 0 0 2 2 8 8 9 11 23 7 10 3 7 9 100
DGW 55 9 7 9 4 4 2 7 1 1 0 1 0 0 0 0 100
Total 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 1600
Garrett score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
Appendix
Table 5.21 Rankings of factors representing climate change using Garrett ranking: Punjab
change 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 score score
event
Total score
Drought 86 76 70 130 183 348 220 416 432 270 84 195 210 390 336 266 3712 37.12 14
Hailstorm 0 0 0 130 0 58 110 156 48 225 168 468 455 630 528 196 3172 31.72 15
Flood 0 76 0 0 0 116 0 0 240 0 84 195 455 480 552 462 2660 26.6 16
AD 0 380 210 390 366 232 385 156 288 360 420 273 630 120 144 98 4452 44.52 12
IPD 1118 1444 840 1235 305 464 275 104 192 45 84 39 140 60 24 28 6397 63.97 2
IFM 258 456 490 195 793 638 715 208 144 270 336 390 175 150 24 28 5270 52.7 7
UR 1118 1444 1400 455 366 348 275 156 48 135 126 156 105 120 48 14 6314 63.14 3
IW 688 1140 980 910 488 290 220 104 192 180 84 156 105 180 96 42 5855 58.55 6
TFH 430 1140 840 1040 915 754 275 208 240 180 0 0 35 30 72 14 6173 61.73 4
TFL 86 608 1330 845 1159 522 220 364 144 180 84 117 140 0 72 14 5885 58.85 5
SSDI 0 76 280 325 427 290 605 572 672 450 420 351 140 180 48 14 4850 48.5 8
CCP 0 0 70 130 366 464 605 728 432 450 462 624 105 180 0 42 4658 46.58 9
CLA 0 0 0 65 244 348 715 624 720 405 504 234 175 300 120 28 4482 44.82 10
CFI 0 76 0 65 122 580 330 624 528 810 378 390 280 90 168 28 4469 44.69 11
CF 86 0 0 0 122 116 440 416 432 495 966 273 350 90 168 126 4080 40.8 13
DGW 4730 684 490 585 244 232 110 364 48 45 0 39 0 0 0 0 7571 75.71 1
111
112
Table 5.22 Factors influencing farmers perception of climate change—Garrett ranking: Uttar Pradesh
Climate change event Rank Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 1 0 1 4 8 6 5 2 4 3 5 6 7 12 35 1 100
Hailstorm 48 17 11 4 3 3 3 3 2 2 4 0 0 0 0 0 100
Flood 9 35 14 9 1 2 8 5 4 4 5 0 4 0 0 0 100
AD 0 0 2 0 1 2 8 17 17 10 10 7 7 11 7 1 100
IPD 12 13 5 15 10 16 17 9 3 0 0 0 0 0 0 0 100
IFM 0 0 0 0 0 0 0 0 0 0 0 1 0 5 15 79 100
UR 18 22 27 3 8 7 8 3 0 2 0 2 0 0 0 0 100
IW 9 8 12 16 5 11 10 6 5 3 2 0 1 4 6 2 100
TFH 0 0 0 1 0 2 2 10 6 11 24 21 18 4 1 0 100
TFL 0 0 0 0 0 0 2 3 4 9 9 27 26 16 4 0 100
SSDI 0 0 0 2 1 7 17 22 27 11 8 1 3 0 1 0 100
CCP 0 0 0 0 0 0 1 0 2 8 15 14 13 13 24 10 100
CLA 0 0 0 0 0 3 1 7 17 18 9 10 12 20 2 1 100
CFI 0 0 2 23 18 17 8 6 5 6 2 6 6 1 0 0 100
CF 3 5 24 19 31 15 1 0 0 0 1 0 0 1 0 0 100
DGW 0 0 2 4 14 9 9 7 4 13 6 5 3 13 5 6 100
Total 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 1600
Garrett score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
Appendix
Table 5.23 Rankings of factors representing climate change using Garrett ranking: Uttar Pradesh
change 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 score score
event
Total score
Drought 86 0 70 260 488 348 275 104 192 135 210 234 245 360 840 14 3861 38.61 13
Hailstorm 4128 1292 770 260 183 174 165 156 96 90 168 0 0 0 0 0 7482 74.82 1
Flood 774 2660 980 585 61 116 440 260 192 180 210 0 140 0 0 0 6598 65.98 3
AD 0 0 140 0 61 116 440 884 816 450 420 273 245 330 168 14 4357 43.57 10
IPD 1032 988 350 975 610 928 935 468 144 0 0 0 0 0 0 0 6430 64.3 5
IFM 0 0 0 0 0 0 0 0 0 0 0 39 0 150 360 1106 1655 16.55 16
UR 1548 1672 1890 195 488 406 440 156 0 90 0 78 0 0 0 0 6963 69.63 2
IW 774 608 840 1040 305 638 550 312 240 135 84 0 35 120 144 28 5853 58.53 6
TFH 0 0 0 65 0 116 110 520 288 495 1008 819 630 120 24 0 4195 41.95 11
TFL 0 0 0 0 0 0 110 156 192 405 378 1053 910 480 96 0 3780 37.8 14
SSDI 0 0 0 130 61 406 935 1144 1296 495 336 39 105 0 24 0 4971 49.71 8
CCP 0 0 0 0 0 0 55 0 96 360 630 546 455 390 576 140 3248 32.48 15
CLA 0 0 0 0 0 174 55 364 816 810 378 390 420 600 48 14 4069 40.69 12
CFI 0 0 140 1495 1098 986 440 312 240 270 84 234 210 30 0 0 5539 55.39 7
CF 258 380 1680 1235 1891 870 55 0 0 0 42 0 0 30 0 0 6441 64.41 4
DGW 0 0 140 260 854 522 495 364 192 585 252 195 105 390 120 84 4558 45.58 9
113
Chapter 6
in Eastern Region of India

6.1 Introduction
The East Region for the present study covers four states, viz. Bihar, Odisha, Chhat-
tisgarh and West Bengal, and the districts selected from these States were Jehanabad,
Bhadrak, Raipur and Bankura, respectively. As stated in the methodology chapter,
the present study is based on (i) primary data: farm survey among 100 farmers from
Bihar, 500 farmers from Odisha, 70 farmers from Chhattisgarh and 151 farmers from
West Bengal and (ii) secondary data: time series data (1982–2010) on productivity
(yield) of rice and climate variables from the three districts belonging to Bihar, Odisha
and West Bengal (since Chhattisgarh is a newly formed State, it was excluded). The
location of the selected districts in eastern region is shown in Fig. 6.1.
From the total of 821 farmers surveyed, 790 were males and 31 were females. Out of
them, all the 100 farmers in Bihar and 51 farmers in West Bengal have irrigated crop-
based farming. Also, 200 farmers in Odisha and 100 farmers in West Bengal follow
crop with supplemental irrigation typology. Thus, in West Bengal, both irrigation-
based and crop with supplemental irrigation cropping typologies are prevalent. All the
farmers in Chhattisgarh and 100 farmers in Odisha have dry-land-based typologies.
The average age of the farmers in the four States, viz. Bihar, Odisha, Chhattisgarh
and West Bengal is 45, 49, 44 and 43 years, respectively. About 12% of the farmers
in the region have no formal education, 32% have studied up to 5th standard, 46%
up to primary level and the rest, 11% have studied up to secondary level (Fig. 6.2)
and 56% of the farmers from Odisha have education up to primary level. Thus, the
education level of the farmers in the region is low.

116 6 Climate Change Impacts on Rice Yield in Eastern Region …
60 56
49
50 47
46
42
40 35
33 32
30 No formal education
30 26
Upto 5th
19 Primary level
20
11 11 11 13 12 Secondary & above
10 11
10 6
0
0
Bihar Odissa Chattisgarh West Bengal Total
Fig. 6.2 Educational status of sample farmers

Farmers in the study area possess land and wells. They use different sources of
irrigation like tanks, canals and wells, and they invest in digging borewells. In Bihar
and Odisha, they raise rice crop only during the kharif season. Farmers in Chhattisgarh
raise rice crop in kharif and rabi seasons, and in West Bengal, farmers raise rice in
all three seasons, viz. kharif, rabi and summer. The farmers in West Bengal also raise
other crops. Overall, farmers from the four states derive income mainly from rice
during kharif season. Though farmers in West Bengal could get income from rice
crop in kharif, they could get income from other crops also during all the seasons.
(i) Land owned: Farmers in Bihar own wet, irrigated and dry lands, whereas farm-
ers in Odisha own only wet and irrigated land. They do not have dry lands.
Chhattisgarh farmers possess only irrigated lands. Farmers belonging to West
Bengal own irrigated and dry lands. The summary statistics is given in Table 6.1.
The average size of land owned by farmers in the four states are, respectively,
0.77, 1.06, 2.34 and 0.85 ha. The mean area under irrigation varies from 0.94 to
2.34 ha across the four states with the highest area corresponding to Chhattisgarh.
But, Chhattisgarh has the highest standard deviation of 2.31 ha of irrigated land. The
average area under wetland in Bihar is 0.69 ha while 1.0 ha in Odisha. The average
area under dry land is 0.63 ha in Bihar and 0.45 ha in West Bengal.
(ii) Wells owned: No farmer possesses open or borewells in Odisha and West
Bengal, and farmers in Bihar possess no open wells whereas only four farmers
in Chhattisgarh possess open wells. Table 6.2 gives the frequency distribution
of number of borewells owned by the farmers in Bihar and Chhattisgarh states.
Table 6.1 Summary statistics of land owned (hectares)

State Land type Mean S.D. Min Max Count
Bihar Wet 0.69 0.41 0.40 2.02 68
Irrigated 0.94 0.51 0.40 3.64 100
Dry 0.63 0.31 0.40 1.62 84
Overall 0.77 0.45 0.40 3.64 252
Odisha Wet 1.00 0.57 0.40 3.24 193
Irrigated 1.11 0.99 0.20 8.09 300
Overall 1.06 0.85 0.20 8.09 493
Chhattisgarh Irrigated 2.34 2.31 0.40 14.16 70
West Bengal Irrigated 1.00 1.25 0.03 12.02 151
Dry 0.45 0.55 0.04 3.64 59
Overall 0.85 1.13 0.03 12.02 210
Table 6.2 Distribution of number of borewells owned across states

No. of wells Number of farmers owning
Bihar Chhattisgarh Total
Borewells Open wells Borewells
0 18 66 43 127
1 61 4 27 92
2 12 – – 12
3 6 – – 6
4 2 – – 2
>4 1 – – 1
Table 6.3 Sources of irrigation in West Bengal

Source of irrigation No. of farmers
Canal 1
Tank 89
well 3
Well (Village) 7
Well (Village) & canal 47
Well (Own) 1
Well (Own) & canal 1
Well (Village and own) & canal 2
Total 151
It is evident that borewells are common in these two states. The sources of
irrigation in West Bengal are shown in Table 6.3.
(iii) Sources of irrigation. Farmers have three sources of irrigation, viz. tanks,
canals and wells. All farmers in Bihar depend on wells only (Fig. 6.3).
(iv) Investment in irrigation: No farmer in all the four states used water-saving
technologies like drip and sprinkler irrigation.
(v) Area (ha) under rice in the region.
Rice is grown in all the four states during kharif season. The average area under
rice per farmer during this season ranged between 0.79 and 1.07 ha. In Chhattisgarh,
rice is grown during rabi season also (Tables 6.4 and 6.5).
(b) Net Income from rice and other crops
The average net income per ha has a wide variation among the states. During kharif
season, it varies between Rs. 79,479 for Chhattisgarh and Rs. 5822 for Bihar. Farmers
in all the states except Chhattisgarh derive income from other crops also and West
Bengal farmer earns an average income of Rs. 46,493.
160
140
120
89 Tanks
100
0
Canals
80
0
0 Wells
60 11
100 39
Wells & Canals
40
0
20 51
31
0
0
Bihar Chaƫsgarh West Bengal
Fig. 6.3 Frequency distribution of sources of irrigation
Table 6.4 Average area (ha) per farmer under rice across the states in eastern region
State Season Rice Other crops
Bihar Kharif 0.79 (100) –
Odisha Kharif 1.06 (500) –
Rabi – 0.23 (40)
Chhattisgarh Kharif 2.34 (70) –
Rabi 0.63 (11) –
West Bengal Kharif 1.07 (151) 0.13 (51)
Rabi – 0.42 (107)
Summer 1.38 (51) 0.15 (59)
Figures in brackets denote the number of farmers who raised the particular crop in the particular
season

Farmers’ perception of the climate change, its effects on agriculture and various
adaptation strategies/techniques followed by the farmers are assessed based on the
detailed survey. The results are discussed below.
Table 6.5 Net income from rice and other crops

State Net income (Rs/ha)
Rice Wheat Other crops
Kharif Rabi Summer Rabi
Bihar 5822 (100) – – 8194 (100) 2862 (100)
Odisha 17,725 (500) – – – 31,870 (39)
Chhattisgarh 79,479 (70) 82,881(11) – – –
West Bengal 28,533 (151) – 13,911 (51) – 46,493 (111)
Overall 23,528 (821) 82,881 (11) 13,911 (51) 8194 (100) 26,759 (250)
season
according to their perceptions of climate change. To find what the farmers perceive
as the most significant effects of climate change, Garrett’s ranking technique as
described in Chap. 4 was used. The results are given in Tables 6.6 and 6.7. Table 6.7
shows that crop failure is the most important climate change event (Rank 1) perceived
by Rice farmers in eastern region. It earned an average Garrett Score of 69.54.
As given in Table 6.6, it was given Rank 1 by 500 farmers. The second and third
important climate change events are, respectively, depletion of groundwater and
drought. Similar analysis was done separately for each state, viz. Bihar, Odisha,
Chhattisgarh and West Bengal. The results are presented in Appendix 6.1.
climate change events (listed in Table 6.6) by the farmers. The calculated value of
Friedman’s F statistics was 6214.3 while the table value (corresponding to 15 degrees
of freedom at 5% level of significance) is 24.996. This shows that the farmers did not
consistently rank the events and the farmers’ perceptions of climate change effects
differ across the sample chosen. The same conclusion was arrived at by applying
Kendall’s coefficient of concordance. Similar analysis was done for individual states,
viz. Bihar, Odisha, Chhattisgarh and West Bengal. The study results confirm that there
is no consistency in the rankings of the events by the farmers. Table 6.8 gives the
rankings of the climate change events state-wise.
Table 6.8 shows that the rankings of the climate change events are not uniform
across the states. So, to check for agreement in rankings, Kendall’s Coefficient of
Concordance was computed and tested for its significance. The computed value of
Kendall’s Coefficient, W was 0.4493 (with number of objects n = 16 and number
of judges m = 3) with a χ 2 value of 26.9559 (p-value = 0.0291). Since the p-value
is less than 0.05, we reject the null hypothesis and conclude that the states are not
concordant in ranking the climate change events. That is, the states differ in ranking
the climate change events. The statewise information on the factors influencing and
the ranking of the factors are given in Appendix (Tables 6.21, 6.22, 6.23, 6.24, 6.25,
6.26, 6.27, 6.28).
Table 6.6 Factors influencing farmers perception of climate change—Garrett ranking: eastern region
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 173 7 200 4 303 3 14 16 16 8 8 16 27 26 0 0 821
Hailstorm 0 0 0 36 55 36 24 0 0 1 92 23 29 18 6 501 821
Flood 0 1 9 16 14 25 20 11 4 0 200 70 0 0 451 0 821
AD 41 19 9 24 21 21 35 19 15 114 9 119 71 204 100 0 821
IPD 33 19 7 88 7 511 5 13 12 6 22 17 25 21 25 10 821
IFM 5 5 5 3 15 18 10 11 6 9 22 39 146 407 114 6 821
UR 29 171 66 117 13 11 111 0 301 0 0 2 0 0 0 0 821
IW 7 31 125 148 24 19 307 121 21 12 4 0 2 0 0 0 821
TFH 27 37 44 33 97 21 117 409 16 16 4 0 0 0 0 0 821
TFL 0 3 5 11 9 83 7 14 17 220 101 0 200 0 0 151 821
SSDI 2 8 306 16 225 24 101 32 34 28 28 13 4 0 0 0 821

CCP 0 1 0 303 1 1 3 6 205 16 27 35 46 62 92 23 821
CLA 1 1 11 9 13 18 18 32 24 147 230 226 16 5 0 70 821
CFI 0 0 0 2 1 4 7 14 18 178 22 238 237 43 22 35 821
CF 500 3 5 0 6 10 20 95 29 24 37 16 13 27 11 25 821
DGW 3 515 29 12 16 17 21 28 103 41 16 6 6 8 0 0 821
Total 821 821 821 822 820 822 820 821 821 820 822 820 822 821 821 821 13,136
Garrett score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
121
122
Table 6.7 Rankings of factors representing climate change using Garrett ranking: eastern region
change 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
event
Total score
Drought 14,878 532 14,000 260 18,483 174 770 832 768 360 336 624 945 780 0 0 53,742 65.4592 3
Hailstorm 0 0 0 2340 3355 2088 1320 0 0 45 3864 897 1015 540 144 7014 22,622 27.5542 16
Flood 0 76 630 1040 854 1450 1100 572 192 0 8400 2730 0 0 10,824 0 27,868 33.94397 14
AD 3526 1444 630 1560 1281 1218 1925 988 720 5130 378 4641 2485 6120 2400 0 34,446 41.95615 10
IPD 2838 1444 490 5720 427 29,638 275 676 576 270 924 663 875 630 600 140 46,186 56.25579 8
IFM 430 380 350 195 915 1044 550 572 288 405 924 1521 5110 12,210 2736 84 27,714 33.75639 15
UR 2494 12,996 4620 7605 793 638 6105 0 14,448 0 0 78 0 0 0 0 49,777 60.62972 5
IW 602 2356 8750 9620 1464 1102 16,885 6292 1008 540 168 0 70 0 0 0 48,857 59.50914 6
TFH 2322 2812 3080 2145 5917 1218 6435 21,268 768 720 168 0 0 0 0 0 46,853 57.06821 7
TFL 0 228 350 715 549 4814 385 728 816 9900 4242 0 7000 0 0 2114 31,841 38.78319 12
SSDI 172 608 21,420 1040 13,725 1392 5555 1664 1632 1260 1176 507 140 0 0 0 50,291 61.25579 4
CCP 0 76 0 19,695 61 58 165 312 9840 720 1134 1365 1610 1860 2208 322 39,426 48.02192 9
CLA 86 76 770 585 793 1044 990 1664 1152 6615 9660 8814 560 150 0 980 33,939 41.33861 11
CFI 0 0 0 130 61 232 385 728 864 8010 924 9282 8295 1290 528 490 31,219 38.02558 13
CF 43,000 228 350 0 366 580 1100 4940 1392 1080 1554 624 455 810 264 350 57,093 69.5408 1
DGW 258 39,140 2030 780 976 986 1155 1456 4944 1845 672 234 210 240 0 0 54,926 66.90134 2
6 Climate Change Impacts on Rice Yield in Eastern Region …
Table 6.8 Ranking of climate change events state-wise

Climate change event Bihar Odisha CGH WB Eastern region
Drought 1 4 1 12 3
Hailstorm 11 16 11 6 16
Flood 4 14 12 15 14
AD 5 13 13 5 10
IPD 14 6 4 4 8
IFM 12 15 14 8 15
UR 2 8 2 2 5
IW 7 7 3 3 6
TFH 6 9 5 1 7
TFL 9 11 6 16 12
SSDI 8 3 7 7 4
CCP 15 5 15 14 9
CLA 10 10 16 11 11
CFI 16 12 10 13 13
CF 13 1 8 9 1
DGW 3 2 9 10 2
They were asked to rank the seven effects of climatic shocks, as described in method-
ology chapter. Farmers ranked declining crop yield as the most important shock as
out of 821 farmers, 647 of them ranked this factor as number one (Tables 6.9 and
6.10). The second and third ranks are given to the loss of income and food secu-
rity/shortage, respectively. This is understandable because the decline in crop yield
will imply loss of income and food shortage.
Farmers apply several mitigation strategies to tide over the negative impacts of climate
change. Table 6.11 provides a summary. It shows that 83% of the farmers in the region
have restricted rice growing season to one. About 88% of the farmers (54% in total)
in Odisha have changed crop variety and planting dates. These two strategies were
not followed by farmers in other states. Only 22% of the farmers have left their land
fallow. Also, 82% of the farmers in the region borrowed money from relatives/others
and only 38% received assistance from the government. All other strategies are
followed by less than 20% of the farmers.

Factors Rank Total no. of farmers
1 2 3 4 5 6 7
DCY 647 88 61 14 0 11 0 821
LA 0 23 37 223 228 135 175 821
LI 50 475 131 89 51 23 2 821
FIS 4 118 229 61 284 108 17 821
DL 24 15 32 161 165 149 275 821
DC 51 42 245 230 72 66 115 821
DH 43 60 88 41 22 331 236 821
Total 819 821 823 819 822 823 820 5747
Garrett score 79 66 57 50 43 34 21

Factors Rank Rank
1 2 3 4 5 6 7 Total score Average
score
Total score
DCY 51,113 5808 3477 700 0 374 0 61,472 74.87454 1
LA 0 1518 2109 11,150 9804 4590 3675 32,846 40.00731 5
LI 3950 31,350 7467 4450 2193 782 42 50,234 61.18636 2
FIS 316 7788 13,053 3050 12,212 3672 357 40,448 49.26675 3
DL 1896 990 1824 8050 7095 5066 5775 30,696 37.38855 7
DC 4029 2772 13,965 11,500 3096 2244 2415 40,021 48.74665 4
DH 3397 3960 5016 2050 946 11,254 4956 31,579 38.46407 6

Strategies Bihar Odisha Chhattisgarh WB Total Percentage
Did nothing 100 0 0 0 100 12
Changed crop 0 441 0 0 441 54
variety
Changed planting 0 441 0 0 441 54
date
Left land fallow 100 14 0 64 178 22
Sold part of land 0 15 0 7 22 3
Leased out part of 0 153 0 14 167 20
land
Restricted number 100 488 0 95 683 83
of seasons for rice
growing to 1
Sold livestock 0 1 0 87 88 11
(continued)

Strategies Bihar Odisha Chhattisgarh WB Total Percentage
Shifted to other 0 0 0 47 47 6
crops
Borrowed money 100 493 0 82 675 82
from relatives/others
Received assistance 0 139 70 99 308 38
from
government/NGOs
Less food 0 0 0 66 66 8
consumption
Shifted to non-farm 19 1 11 90 121 15
employment
Outmigration to 0 44 0 6 50 6
cities
From time to time, both Central and State governments implement several mitigations
and long-term measures to offset the negative effects of climate change on agriculture.
With a view to know the awareness of these measures among the farmers, they were
questioned on six government initiatives (as described in the methodology chapter).
They were asked to state their replies as either ‘yes’, or ‘no’ or ‘not aware’. Figure 6.4
gives a pictorial representation of the awareness of these initiatives. About 95% of the
farmers are well aware of subsidized inputs (like fertilizers, drip and machinery), 73%
know about the introduction of new crop varieties, 70% have knowledge on water-
based crop insurance schemes and 62% of the farmers know about water regulation
by the government through irrigation department. Only 16% of the farmers know
about the establishment of village knowledge centres and 14% only are aware of
field demos on the listed adaptation practices. The last two government initiatives
are very important for the farmers to know about latest technologies and so suitable
mechanisms should be formed by the government to popularize these two initiatives.
Table 6.12 provides a summary of the opinions on government initiatives, state-wise.
among the three states. This may be due to the government initiatives depending on
Water regulaƟon through IntroducƟon of new crop

irrigaƟon department varieƟes
Not aware
Not aware 13%
20% No
14% Yes
No Yes 73%
18% 62%
Field demos on the listed

adaptaƟon pracƟces Subsidised inputs
No
Not aware Yes 5%
13% 14%
Yes
No 95%
73%
Weather based crop Establishment of village

No
insurance schemes 28% knowledge centres
Not aware
2% Not aware
9% Yes
16%
Yes No
70% 75%
Fig. 6.4 Opinions of farmers of eastern region on various government initiatives
6.5.1 Data
corresponds to the growing season of rice variety in the respective states and the time
series data on rainfall was collected for the entire year to account for rainfall falling
directly on the crop and interseasonal water accumulation within a year (Iski and
Table 6.12 Farmers’ opinions on government initiatives
Govt. initiatives Bihar Odisha Chhattisgarh West Bengal Total Chi-square df p-value
Yes No NA Yes No NA Yes No NA Yes No NA Yes No NA
Water regulation 0 0 100 388 112 0 70 0 0 54 36 61 512 148 161 800.16 6 <0.00001
through irrigation
department
Introduction of new 0 0 100 498 2 0 70 0 0 29 118 4 597 120 104 1451.6 6 <0.00001
crop varieties
Field demos on the 0 100 0 1 399 100 70 0 0 47 96 8 118 595 108 584.63 6 <0.00001
listed adaptation
practices
Organized field days 0 100 0 0 400 100 70 0 0 42 103 6 112 603 106 602.01 6 <0.00001
on the adaptation
practices
Subsidized inputs 100 0 0 499 1 0 70 0 0 114 37 0 783 38 0 165.58 6 <0.00001

machinery)
Weather-based crop 0 100 0 500 0 0 70 0 0 4 128 19 574 228 19 848.52 6 <0.00001
insurance schemes
Establishment of 0 100 0 133 367 0 0 70 0 1 81 69 134 618 69 412.38 6 <0.00001
village knowledge
centres, climate cell,
etc.
127

Dev.
Bihar 29 1.2144 0.681 1.600 0.244 Y = 1.0183 + 0.21
0.0131** T
Odisha 29 1.3094 0.737 1.795 0.278 Y = 1.0513 + 0.28
0.0172*** T
West Bengal 29 2.1569 1.018 3.346 0.475 Y = 1.4333 + 0.75
0.0482*** T
Overall 87 1.560 0.681 3.346 0.547 – –
** and *** denotes significane at 5% and 1% level
3.5
Bihar
Odissa
WestBengal
3
Yield (tonnes per ha)
2.5
1.5
0.5
1985 1990 1995 2000 2005 2010
Year
Fig. 6.5 Trend in yield of rice in eastern region of India
Devadoss 2006; Kim and Pang 2009). Since Chhattisgarh State was formed during
2000 only, this State was excluded from panel data analysis and only three states,
viz. Bihar, Odisha and West Bengal were included. The summary statistics of all the
variables are discussed below.
The yield data (Table 6.13; Fig. 6.5) shows an upward trend in all the three states
over years, and yield for West Bengal was greater than that of the other two states
in all the 29 years.. Trend equations fitted to yield data (Table 6.13) show that the
coefficient of trend for all the three states is statistically significant.
The maximum temperature had an overall average of 32.19 °C. Odisha had the
highest mean maximum temperature of 32.5 °C followed by Bihar (32.3 °C). Also,
the variability in maximum temperature is least for Bihar and highest for West Bengal
(Table 6.14; Fig. 6.6).
The minimum temperature data shows that Odisha has the highest average minimum
temperature compared to the other two states. Its average minimum temperature
is 22.02 °C. It is consistently higher than those of the other two states in most of
Table 6.14 Descriptive statistics of maximum temperature (°C)

Bihar 29 32.325 31.814 32.903 0.298
Odisha 29 32.483 31.750 34.250 0.634
West Bengal 29 32.106 29.750 33.670 0.858
Overall 87 32.190 29.750 34.250 1.121
34.5
Bihar
Odissa
34
WestBengal
33.5
33
32.5
32
31.5
31
30.5
30
29.5
1985 1990 1995 2000 2005 2010
Year
Fig. 6.6 Trend in maximum temperature in eastern region of India


Bihar 29 19.936 19.305 20.481 0.363
Odisha 29 22.026 19.250 23.750 1.022
West Bengal 29 21.138 19.333 22.167 0.768
Overall 87 21.033 19.250 23.750 1.148
24
Bihar
Odissa
23.5
WestBengal
Minimum Temperature (Centigrade)
23
22.5
22
21.5
21
20.5
20
19.5
19
1985 1990 1995 2000 2005 2010
Year
Fig. 6.7 Trend in minimum temperature in eastern region of India
the years. Bihar had the lowest minimum temperature with an average of 19.93 °C
(Table 6.15; Fig. 6.7).
(d) Rainfall (mm)
Average annual rainfall was the highest for Odisha (1488 mm) followed by West
Bengal (1410 mm) (Table 6.16). Also, variability in rainfall was the highest for
Odisha with a standard deviation of 310 mm. Both Bihar and West Bengal had the
same variability in rainfall, but the average rainfall of Bihar was least (885.79 mm).
Rainfall for Bihar was consistently lower than that of the other two states in most of
the years (Fig. 6.8).
Table 6.16 Descriptive statistics of rainfall (mm)

Bihar 29 885.79 518.10 1472.5 226.87
Odisha 29 1488.0 922.60 2052.6 309.54
West Bengal 29 1401.3 917.0 1954.0 227.91
Overall 87 1258.4 518.10 2052.6 369.25
2200
Bihar
Odissa
2000 WestBengal
1800
1600
Rainfall (mm)
1400
1200
1000
800
600
400
1985 1990 1995 2000 2005 2010
Year
Fig. 6.8 Trend in annual rainfall in eastern region of India
The normality of yield data was tested and the tests showed that the data slightly
deviates from normality and hence, the outliers were removed to have a normally
distributed data.
Table 6.17 Panel root test results using Fisher’s tests

Variable Test statistic p-value Decision Test statistic p-value Decision
for individual for individual
effects and no effect with
trend trend
Yield 20.6380 0.0021 Reject 26.2447 0.0002 Reject
H0 H0
Maximum 18.6992 0.0047 Reject 19.2068 0.0038 Reject
temperature H0 H0
Minimum 10.9396 0.0093 Reject 17.3331 0.0081 Reject
temperature H0 H0
Rainfall 27.631 0.0001 Reject 24.2215 0.0005 Reject
H0 H0
The data on rice yield and climate variables for three States for 29 years is available
and was subjected to panel root testing before modelling. As stated in the method-
ology Chap. 3, Fisher’s panel root test was applied.
The test results confirmed (Table 6.17) that all the variables in panel data do not
have unit roots and hence, they are stationary.
As discussed in the methodology chapter, to account for both mean and variance in
yield, Just–Pope production function (Just and Pope 1978 & 79) with yield as the
The quadratic model has the least value for AIC and hence selected. Table 6.18
whereas rainfall has a significant negative effect. The square term corresponding to
maximum temperature and its interaction term with rainfall are significant. The sig-
nificance of these terms implies that nonlinear effects of maximum temperature are
significant and so important (Fig. 6.9). The coefficient of the two dummies for Bihar
and Odisha are negative and strongly significant. The positive and significant coeffi-
cient for trend implies the significance of technological progress in rice production
such as improved irrigation methods, use of high-yielding varieties and better control
of pests and diseases over the period under study.
In the case of variance function, coefficient of maximum temperature alone is
significant and also positive. So, it is risk-increasing variable. This means that vari-

Constant (β0 ) 0.4945 2.3423 −73.7162 91.0495
Trend (β1 ) 0.0248*** 0.0044 0.0282*** 0.0047
MaxT (β2 ) −0.0040 0.0572 3.2601** 1.3023
MinT (β3 ) 0.0410 0.0550 2.8695 3.2449
RainT (β4 ) 0.0002 0.0001 −0.0119* 0.0072
MaxT 2 (β5 ) – – −0.0164*** 0.0043
MinT 2 (β6 ) – – 0.0227 0.0272
RainT 2 (β7 ) – – 0.0000 0.0000
MaxT MinT (β8 ) – – −0.1230 0.0987
MaxT RainT (β9 ) – – 0.0003** 0.0001
MinT RainT (β10 ) – – 0.0001 0.0001
Bihar (β11 ) −0.7763*** 0.1305 −0.8840*** 0.1419
Odisha (β12 ) −0.8466*** 0.1196 −0.8832*** 0.1008
Variance function
Constant −45.9696*** 8.4349 −55.3047 10.4527
Time −0.0045 0.0280 −0.0032 0.0303
MaxT 1.2445*** 0.1968 1.5594*** 0.2878
MinT 0.1063 0.2502 0.0637 0.2463
RainT 0.0007 0.0006 0.0007 0.0007
Model statistics
LL 1.175168 7.7028
AIC 20.650 20.5943
** and *** denotes significane at 5% and 1% level
ability in yield increases with an increase in maximum temperature. Other variables

are not statistically significant.
for both mean and variance functions. The elasticity of the three climate variables is
all less than 1, and so they are inelastic. The elasticity of the maximum temperature
for the mean function is −0.4629 implying that 1% increase in maximum temperature
will induce a decrease of 0.46% in mean yield. Similarly, interpretations can be given
to the other two elasticities. When all the three variables are simultaneously increased
by 1%, we can expect a very small decrease of 0.04% in yield.
For the variance function, the elasticity coefficient of maximum temperature is
−50.1963 implying an increase in variability in yield of 5.02 for 1% increase in
maximum temperature. Similar meanings can be given for the elasticities of minimum
temperature and rainfall. All together, we can expect an increase of 7.2% in yield
1.4
1.3
1.2
1.1
0.9
0.8
0.7
25 26 27 28 29 30 31 32 33 34 35
Fig. 6.9 Nonlinear relationship between yield and maximum temperatures
Table 6.19 Elasticities of Yield function Climate variable Elasticity

climatic variables at means
Mean Maximum temperature −0.4629
Rainfall 0.1827
Variance Maximum temperature 5.0196
Rainfall 0.8393
variability when all the three variables are increased by 1% simultaneously. So, we
conclude that all the three variables, put together are risk increasing in this sense.
∂y
∂(MaxT )
∂y
= β8 MaxT + 2 ∗ β6 MinT + β10 RainT = −β3
∂(MinT )
∂y
∂(RainT )
Solving these equations, the optimal levels are obtained as MaxT = 31.9 °C;
MinT = 20.4 °C and RainT = 1043 mm.
climate change scenarios, viz. RCP4.5 and RCP8.5 (as discussed in Chap. 3) were
considered (Table 6.20).
Table 6.20 shows that, as per RCP4.5 and RCP8.5 scenarios, there will be a
marginal increase in percentages of rice yield during the periods 2021–50 and
2071–2100. Under both scenarios, percentage of increase in the later period will
be more. The percentage increase in mean yield varies from 1.40 to 1.63 during the
first period as per the two scenarios, and it varies from 0.96 to 2.36% during the
second period as per the two scenarios.
RCP4.5 scenario for both periods. The first scenario predicts an increase of 35%
yield variability during 2021–2050 while the second scenario predicts an increase of
41% for the same period. For the second period, RCP4.5 predicts a higher increase
variability of 139%.
Thus, we conclude that climate change may not have adverse effect on the mean
rice yield in eastern region but yield variability may be substantial.
Scenario Period Percentage change in Percentage change in
T max T min Rainfall Mean rice yield Yield variability
RCP4.5 2021–2050 3.36 7.10 9.82 1.40 34.63
2071–2100 9.79 14.42 17.09 0.96 82.79
RCP8.5 2021–2050 3.97 8.86 11.00 1.63 41.04
2071–2100 15.80 26.33 29.28 2.36 139.14
6.6 Summary
Eastern region farmers perceive crop failure followed by the depletion of groundwater
and drought as the most important climate change events. As in the case of northern
region, farmers of this region also ranked decline in crop yield as the most important
climate change shock. As far as mitigation strategies, more than 80% of the farmers
in the region have restricted rice growing season to one and they borrowed money
from their relatives/friends.
About 95% of the farmers are well aware of subsidized inputs as they have good
knowledge about the introduction of new crop varieties and weather-based crop insur-
ance schemes. Only the small number of farmers knows about the establishment of
village knowledge centres and aware of field demos on the listed adaptation prac-
tices. These two government initiatives are very important for the farmers to update
their knowledge about the latest technologies and so suitable steps should be taken
by the government to popularize these two initiatives.
Based on historical data, yield of rice generally shows an upward trend in all the
states over the years, although there was a decline in yield in some years. Odisha
had the highest maximum temperature, minimum temperature and annual rainfall
compared to the other two states. Maximum temperature has a significant positive
effect on rice yield, whereas rainfall has significant negative effect. The square term
corresponding to maximum temperature and its interaction term with rainfall is sig-
nificant implying nonlinear effects of climatic variables on yield. The two climate
change scenarios RCP4.5 and RCP8.5 predict a marginal increase in percentages
of rice yield during the periods 2021–50 and 2071–2100. Also, yield variability as
estimated by the second scenario is more than that predicted by the first scenario for
both periods.
Appendix
.
Appendix
Table 6.21 Factors influencing farmers’ perception of climate change–Garrett ranking: Bihar
Factors Rank Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 99 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 100
Hailstorm 0 0 0 0 0 0 0 0 0 1 22 23 29 18 6 1 100
Flood 0 1 9 16 14 25 20 11 4 0 0 0 0 0 0 0 100
AD 0 4 8 17 13 13 20 14 11 0 0 0 0 0 0 0 100
IPD 0 0 0 0 0 0 0 0 0 0 13 13 18 21 25 10 100
IFM 0 0 0 0 0 0 0 0 0 3 11 23 26 17 14 6 100
UR 0 70 20 6 4 0 0 0 0 0 0 0 0 0 0 0 100
IW 1 1 10 12 5 10 7 18 19 12 4 0 1 0 0 0 100
TFH 0 0 8 10 17 10 13 9 15 16 2 0 0 0 0 0 100
TFL 0 3 5 11 9 13 7 14 17 20 1 0 0 0 0 0 100
SSDI 0 3 6 8 15 5 16 13 13 18 3 0 0 0 0 0 100
CCP 0 1 0 0 0 0 0 0 0 3 14 9 7 21 22 23 100
CLA 0 1 9 8 10 11 7 13 14 21 4 2 0 0 0 0 100
CFI 0 0 0 0 0 0 0 0 0 0 8 16 8 11 22 35 100
CF 0 0 0 0 0 1 0 0 3 6 17 14 11 12 11 25 100
DGW 0 15 25 12 13 12 10 8 4 0 1 0 0 0 0 0 100
Total 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 1600
Garrett score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
137
138
Table 6.22 Rankings of factors representing climate change using Garrett ranking: Bihar
Factors Rank Total Average Rank
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Total score
Drought 8514 76 0 0 0 0 0 0 0 0 0 0 0 0 0 0 8590 85.9 1
Hailstorm 0 0 0 0 0 0 0 0 0 45 924 897 1015 540 144 14 3579 35.79 11
Flood 0 76 630 1040 854 1450 1100 572 192 0 0 0 0 0 0 0 5914 59.14 4
AD 0 304 560 1105 793 754 1100 728 528 0 0 0 0 0 0 0 5872 58.72 5
IPD 0 0 0 0 0 0 0 0 0 0 546 507 630 630 600 140 3053 30.53 14
IFM 0 0 0 0 0 0 0 0 0 135 462 897 910 510 336 84 3334 33.34 12
UR 0 5320 1400 390 244 0 0 0 0 0 0 0 0 0 0 0 7354 73.54 2
IW 86 76 700 780 305 580 385 936 912 540 168 0 35 0 0 0 5503 55.03 7
TFH 0 0 560 650 1037 580 715 468 720 720 84 0 0 0 0 0 5534 55.34 6
TFL 0 228 350 715 549 754 385 728 816 900 42 0 0 0 0 0 5467 54.67 9
SSDI 0 228 420 520 915 290 880 676 624 810 126 0 0 0 0 0 5489 54.89 8
CCP 0 76 0 0 0 0 0 0 0 135 588 351 245 630 528 322 2875 28.75 15
CLA 0 76 630 520 610 638 385 676 672 945 168 78 0 0 0 0 5398 53.98 10
CFI 0 0 0 0 0 0 0 0 0 0 336 624 280 330 528 490 2588 25.88 16
CF 0 0 0 0 0 58 0 0 144 270 714 546 385 360 264 350 3091 30.91 13
DGW 0 1140 1750 780 793 696 550 416 192 0 42 0 0 0 0 0 6359 63.59 3
Appendix
Table 6.23 Factors influencing farmers perception of climate change—Garrett ranking: Odisha
Factors Rank Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 0 0 200 0 300 0 0 0 0 0 0 0 0 0 0 0 500
Hailstorm 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 500 500
Flood 0 0 0 0 0 0 0 0 0 0 200 0 0 0 300 0 500
AD 0 0 0 0 0 0 0 0 0 100 0 100 0 200 100 0 500
IPD 0 0 0 0 0 500 0 0 0 0 0 0 0 0 0 0 500
IFM 0 0 0 0 0 0 0 0 0 0 0 0 100 300 100 0 500
UR 0 0 0 100 0 0 100 0 300 0 0 0 0 0 0 0 500
IW 0 0 0 100 0 0 300 100 0 0 0 0 0 0 0 0 500
TFH 0 0 0 0 0 0 100 400 0 0 0 0 0 0 0 0 500
TFL 0 0 0 0 0 0 0 0 0 200 100 0 200 0 0 0 500
SSDI 0 0 300 0 200 0 0 0 0 0 0 0 0 0 0 0 500
CCP 0 0 0 300 0 0 0 0 200 0 0 0 0 0 0 0 500
CLA 0 0 0 0 0 0 0 0 0 100 200 200 0 0 0 0 500
CFI 0 0 0 0 0 0 0 0 0 100 0 200 200 0 0 0 500
CF 500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 500
DGW 0 500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 500
Total 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 8000
Garrett score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
139
140
Table 6.24 Rankings of factors representing climate change using Garrett ranking: Odisha
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Total score
Drought 0 0 14,000 0 18,300 0 0 0 0 0 0 0 0 0 0 0 32,300 64.6 4
Hailstorm 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7000 7000 14 16
Flood 0 0 0 0 0 0 0 0 0 0 8400 0 0 0 7200 0 15,600 31.2 14
AD 0 0 0 0 0 0 0 0 0 4500 0 3900 0 6000 2400 0 16,800 33.6 13
IPD 0 0 0 0 0 29,000 0 0 0 0 0 0 0 0 0 0 29,000 58 6
IFM 0 0 0 0 0 0 0 0 0 0 0 0 3500 9000 2400 0 14,900 29.8 15
UR 0 0 0 6500 0 0 5500 0 14,400 0 0 0 0 0 0 0 26,400 52.8 8
IW 0 0 0 6500 0 0 16,500 5200 0 0 0 0 0 0 0 0 28,200 56.4 7
TFH 0 0 0 0 0 0 5500 20,800 0 0 0 0 0 0 0 0 26,300 52.6 9
TFL 0 0 0 0 0 0 0 0 0 9000 4200 0 7000 0 0 0 20,200 40.4 11
SSDI 0 0 21,000 0 12,200 0 0 0 0 0 0 0 0 0 0 0 33,200 66.4 3
CCP 0 0 0 19,500 0 0 0 0 9600 0 0 0 0 0 0 0 29,100 58.2 5
CLA 0 0 0 0 0 0 0 0 0 4500 8400 7800 0 0 0 0 20,700 41.4 10
CFI 0 0 0 0 0 0 0 0 0 4500 0 7800 7000 0 0 0 19,300 38.6 12
CF 43,000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 43,000 86 1
DGW 0 38,000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 38,000 76 2
Appendix
Table 6.25 Factors influencing farmers perception of climate change—Garrett ranking: Chhattisgarh
Factors Rank Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 70 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 70
Hailstorm 0 0 0 0 0 0 0 0 0 0 70 0 0 0 0 0 70
Flood 0 0 0 0 0 0 0 0 0 0 0 70 0 0 0 0 70
AD 0 0 0 0 0 0 0 0 0 0 0 0 70 0 0 0 70
IPD 0 0 0 70 0 0 0 0 0 0 0 0 0 0 0 0 70
IFM 0 0 0 0 0 0 0 0 0 0 0 0 0 70 0 0 70
UR 0 70 0 0 0 0 0 0 0 0 0 0 0 0 0 0 70
IW 0 0 70 0 0 0 0 0 0 0 0 0 0 0 0 0 70
TFH 0 0 0 0 70 0 0 0 0 0 0 0 0 0 0 0 70
TFL 0 0 0 0 0 70 0 0 0 0 0 0 0 0 0 0 70
SSDI 0 0 0 0 0 0 70 0 0 0 0 0 0 0 0 0 70
CCP 0 0 0 0 0 0 0 0 0 0 0 0 0 0 70 0 70
CLA 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 70 70
CFI 0 0 0 0 0 0 0 0 0 70 0 0 0 0 0 0 70
CF 0 0 0 0 0 0 0 70 0 0 0 0 0 0 0 0 70
DGW 0 0 0 0 0 0 0 0 70 0 0 0 0 0 0 0 70
Total 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 1120
Garret-score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
141
142
Table 6.26 Rankings of factors representing climate change using Garrett ranking: Chhattisgarh
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Total score
Drought 6020 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6020 86 1
Hailstorm 0 0 0 0 0 0 0 0 0 0 2940 0 0 0 0 0 2940 42 11
Flood 0 0 0 0 0 0 0 0 0 0 0 2730 0 0 0 0 2730 39 12
AD 0 0 0 0 0 0 0 0 0 0 0 0 2450 0 0 0 2450 35 13
IPD 0 0 0 4550 0 0 0 0 0 0 0 0 0 0 0 0 4550 65 4
IFM 0 0 0 0 0 0 0 0 0 0 0 0 0 2100 0 0 2100 30 14
UR 0 5320 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5320 76 2
IW 0 0 4900 0 0 0 0 0 0 0 0 0 0 0 0 0 4900 70 3
TFH 0 0 0 0 4270 0 0 0 0 0 0 0 0 0 0 0 4270 61 5
TFL 0 0 0 0 0 4060 0 0 0 0 0 0 0 0 0 0 4060 58 6
SSDI 0 0 0 0 0 0 3850 0 0 0 0 0 0 0 0 0 3850 55 7
CCP 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1680 0 1680 24 15
CLA 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 980 980 14 16
CFI 0 0 0 0 0 0 0 0 0 3150 0 0 0 0 0 0 3150 45 10
CF 0 0 0 0 0 0 0 3640 0 0 0 0 0 0 0 0 3640 52 8
DGW 0 0 0 0 0 0 0 0 3360 0 0 0 0 0 0 0 3360 48 9
Appendix
Table 6.27 Factors influencing farmers’ perception of climate change—Garrett ranking: West Bengal
Factors Rank Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 4 6 0 4 3 3 14 16 16 8 8 16 27 26 0 0 151
Hailstorm 0 0 0 36 55 36 24 0 0 0 0 0 0 0 0 0 151
Flood 0 0 0 0 0 0 0 0 0 0 0 0 0 0 151 0 151
AD 41 15 1 7 8 8 15 5 4 14 9 19 1 4 0 0 151
IPD 33 19 7 18 7 11 5 13 12 6 9 4 7 0 0 0 151
IFM 5 5 5 3 15 18 10 11 6 6 11 16 20 20 0 0 151
UR 29 31 46 11 9 11 11 0 1 0 0 2 0 0 0 0 151
IW 6 30 45 36 19 9 0 3 2 0 0 0 1 0 0 0 151
TFH 27 37 36 23 10 11 4 0 1 0 2 0 0 0 0 0 151
TFL 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 151 151
SSDI 2 5 0 8 10 19 15 19 21 10 25 13 4 0 0 0 151
CCP 0 0 0 3 1 1 3 6 5 13 13 26 39 41 0 0 151
CLA 1 0 2 1 3 7 11 19 10 26 26 24 16 5 0 0 151
CFI 0 0 0 2 1 4 7 14 18 8 14 22 29 32 0 0 151
CF 0 3 5 0 6 9 20 25 26 18 20 2 2 15 0 0 151
DGW 3 0 4 0 3 5 11 20 29 41 15 6 6 8 0 0 151
Total 151 151 151 152 150 152 150 151 151 150 152 150 152 151 151 151 2416
Garrett score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
143
144
Table 6.28 Rankings of factors representing climate change using Garrett ranking: West Bengal
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Total score
Drought 344 456 0 260 183 174 770 832 768 360 336 624 945 780 0 0 6832 45.25 12
Hailstorm 0 0 0 2340 3355 2088 1320 0 0 0 0 0 0 0 0 0 9103 60.28 6
Flood 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3624 0 3624 24.00 15
AD 3526 1140 70 455 488 464 825 260 192 630 378 741 35 120 0 0 9324 61.75 5
IPD 2838 1444 490 1170 427 638 275 676 576 270 378 156 245 0 0 0 9583 63.46 4
IFM 430 380 350 195 915 1044 550 572 288 270 462 624 700 600 0 0 7380 48.87 8
UR 2494 2356 3220 715 549 638 605 0 48 0 0 78 0 0 0 0 10703 70.88 2
IW 516 2280 3150 2340 1159 522 0 156 96 0 0 0 35 0 0 0 10254 67.91 3
TFH 2322 2812 2520 1495 610 638 220 0 48 0 84 0 0 0 0 0 10749 71.19 1
TFL 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2114 2114 14.00 16
SSDI 172 380 0 520 610 1102 825 988 1008 450 1050 507 140 0 0 0 7752 51.34 7
CCP 0 0 0 195 61 58 165 312 240 585 546 1014 1365 1230 0 0 5771 38.22 14
CLA 86 0 140 65 183 406 605 988 480 1170 1092 936 560 150 0 0 6861 45.44 11
CFI 0 0 0 130 61 232 385 728 864 360 588 858 1015 960 0 0 6181 40.93 13
CF 0 228 350 0 366 522 1100 1300 1248 810 840 78 70 450 0 0 7362 48.75 9
DGW 258 0 280 0 183 290 605 1040 1392 1845 630 234 210 240 0 0 7207 47.73 10
Chapter 7
in Western Region of India

7.1 Introduction
The western region for the present study covers the Gujarat state. As stated in the
methodology chapter, the present study is based on (i) primary data: farm survey
among 103 farmers belonging to Navsari district and (ii) secondary data: time series
data (1980–2010) on productivity (yield) of rice and climate variables from the
district. The location of the study region is given in Fig. 7.1.
Fig. 7.1 Study region

146 7 Climate Change Impacts on Rice Yield in Western Region of India
37
40 33
35
Percentage of farmers 30
25
17
20
14
15
10
5
0
No formal Upto 5th Primary level Secondary &
educaƟon above
Fig. 7.2 Educational status of surveyed farmers—western region
From the 103 farmers surveyed 99 were males and 4 were females. Out of the 103
farmers, 82 farmers have dry-land-based farming. The average age of the farmers is
49 years. About 37% of the farmers have primary-level education, 33% have studied
secondary level and above, 17% have studied up to 5th and the rest, 14% have no
formal education (Fig. 7.2).
Farmers in the study area have owned land and wells for irrigation where drip,
sprinkler and other conventional types of irrigation methods are used. Farmers in
Gujarat derive income from rice during kharif season and from other crops during
all the three seasons.
(a) Physical assets
(i) Land owned: Among the sample farmers, 98 farmers owned irrigated lands
and 38 farmers owned both irrigated and dry lands. Table 7.1 gives the summary
statistics of the type of land owned by the farmers.
Table 7.1 Summary Land type Mean S.D. Min. Max. Count
statistics of land owned
(hectares) Irrigated 0.96 0.74 0.20 3.64 98
Dry 0.73 0.83 0.20 4.86 38
Overall 0.90 0.77 0.20 4.86 136
Table 7.2 Distribution of No. of wells Number of farmers

number of wells owned
Open wells Borewells
0 91 38
1 12 61
2 0 3
3 0 1
Fig. 7.3 Frequency 1 4

distribution of sources of
irrigation
67
Tanks Canals Wells
The mean area under irrigation is 0.96 and 0.73 ha under dry lands, and the
average land area is 0.90 ha. Irrigated landholding has a standard deviation of
0.74 ha, and the minimum and maximum size of the overall landholdings are,
respectively, 0.20 and 4.86 ha.
(ii) Wells owned: Borewells are most common in the study area. Table 7.2 gives
the frequency distribution of number of borewells owned by the farmers.
About 91% of the farmers do not have open wells and about 38% do not have
borewells. Further, 61 farmers (59%) have one borewell each and 12% farmers
possess one open well.
(iii) Sources of irrigation: Farmers have three sources of irrigation, viz. tank, canal
and well.
Well is the main source of irrigation for 67 farmers (65%) as shown in Fig. 7.3.
(iv) Investment in irrigation: Farmers’ investment is related to digging wells and
installation of the microirrigation systems.
(v) Area under rice: From the total 103 farmers, 96 of them have cultivated rice
during the kharif season. During this season, on the average, every farmer
cultivates rice in 0.75 ha. Other crops are grown covering an area of 0.8, 0.75,
0.57 ha in kharif, rabi and summer seasons, respectively.

With a view to have an in-depth understanding of farmers perception on the climate

lowed by the farmers, detailed survey was conducted and the results are discussed
below.
Farmers were asked to rank the climate change events as listed in the methodology
chapter according to their perceptions of climate change. To find what the farmers
perceive as the most significant effects of climate change, Garrett Ranking technique
was applied. The results are given in Tables 7.3 and 7.4. Table 7.4 shows that irregular
weather is the most important climate change event (Rank 1) perceived by rice farmers
in western region. It earned an average Garrett Score of 79.34. As given in Table 7.3,
it was given Ranks 1, 2 and 3 by 61, 25 and 12 farmers, respectively. The second and
third important climate change events are, respectively, high temperature fluctuation
and untimely rain.
climate change events (listed in Table 7.4) by the farmers. The methodology as
was 884.58 while the table value (corresponding to 14 degrees of freedom at 5%
the events; thus, the farmers’ perceptions of climate change effects differ across the
sample chosen. The same conclusion was arrived at by applying Kendall’s coefficient
of concordance also.
They were asked to rank seven effects of climatic shocks, as described in the method-
ology chapter. But, they ranked only five climatic effect shocks. Farmers ranked loss
of income as the most important shock (Tables 7.5 and 7.6). Out of 103 farmers, 95 of
them ranked loss of income as the most important shock. The second and third ranks
are given to the decline in health and food insecurity/shortage. This is understandable
because the decline in income will reflect in lesser spending on consumption leading
to decline in health and food insecurity.
Table 7.3 Factors influencing farmers perception of climate change—Garrett ranking: western region
Factors Rank Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Hailstorm 0 0 0 0 2 0 0 0 0 0 0 0 2 24 75 103
Flood 0 0 1 9 7 2 0 0 0 0 0 3 38 39 4 103
AD 0 0 2 1 9 6 12 13 7 11 7 18 14 3 0 103
IPD 0 5 2 14 20 22 17 12 8 1 2 0 0 0 0 103
IFM 0 2 3 2 7 2 4 13 13 10 25 14 4 3 1 103
UR 10 24 29 8 6 8 4 6 3 1 3 0 1 0 0 103
IW 61 25 12 2 1 0 1 0 0 1 0 0 0 0 0 103
TFH 26 33 33 8 0 0 0 0 1 0 0 0 1 1 0 103
TFL 4 4 8 8 5 2 4 1 2 2 1 3 16 25 18 103
SSDI 0 1 0 19 10 33 11 11 8 4 3 1 0 0 2 103
CCP 0 2 4 7 8 7 10 9 19 20 9 8 0 0 0 103
CLA 0 0 0 1 2 6 6 5 5 13 26 26 10 2 1 103
CFI 0 1 4 1 1 1 7 8 15 9 17 24 10 5 0 103
CF 0 0 0 2 4 5 21 14 16 26 8 5 1 1 0 103
DGW 2 6 5 21 21 9 6 11 6 5 2 1 6 0 2 103
7.4 Climate Change Events, Shocks and Their Effects as Perceived by Farmers
Total 103 103 103 103 103 103 103 103 103 103 103 103 103 103 103 1545
Garrett score 85 75 69 64 60 56 53 50 47 43 40 36 31 25 15
149
150
Table 7.4 Rankings of factors representing climate change using Garrett ranking: western region
Factors Rank Total score Average Rank
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 score
Total score
Hailstorm 0 0 0 0 120 0 0 0 0 0 0 0 62 600 1125 1907 18.51 15
Flood 0 0 69 576 420 112 0 0 0 0 0 108 1178 975 60 3498 33.96 14
AD 0 0 138 64 540 336 636 650 329 473 280 648 434 75 0 4603 44.69 10
IPD 0 375 138 896 1200 1232 901 600 376 43 80 0 0 0 0 5841 56.71 4
IFM 0 150 207 128 420 112 212 650 611 430 1000 504 124 75 15 4638 45.03 9
UR 850 1800 2001 512 360 448 212 300 141 43 120 0 31 0 0 6818 66.19 3
IW 5185 1875 828 128 60 0 53 0 0 43 0 0 0 0 0 8172 79.34 1
TFH 2210 2475 2277 512 0 0 0 0 47 0 0 0 31 25 0 7577 73.56 2
TFL 340 300 552 512 300 112 212 50 94 86 40 108 496 625 270 4097 39.78 13
SSDI 0 75 0 1216 600 1848 583 550 376 172 120 36 0 0 30 5606 54.43 6
CCP 0 150 276 448 480 392 530 450 893 860 360 288 0 0 0 5127 49.78 7
CLA 0 0 0 64 120 336 318 250 235 559 1040 936 310 50 15 4233 41.10 12
CFI 0 75 276 64 60 56 371 400 705 387 680 864 310 125 0 4373 42.46 11
CF 0 0 0 128 240 280 1113 700 752 1118 320 180 31 25 0 4887 47.45 8
DGW 170 450 345 1344 1260 504 318 550 282 215 80 36 186 0 30 5770 56.02 5
7 Climate Change Impacts on Rice Yield in Western Region of India
7.4 Climate Change Events, Shocks and Their Effects as Perceived by Farmers 151
Table 7.5 Rankings of Factors Rank Total

effects of climatic shocks by
farmers 1 2 3 4 5
LI 95 8 0 0 0 103
FIS 4 26 14 32 27 103
DL 1 16 35 30 21 103
DC 2 20 32 21 28 103
DH 1 33 22 20 27 103
Total 103 103 103 103 103 515
Garrett score 75 60 50 40 25

Factors Rank Total score Average score Rank
1 2 3 4 5
Total score
LI 7125 480 0 0 0 7605 73.8 1
FIS 300 1560 700 1280 675 4515 43.8 3
DL 75 960 1750 1200 525 4510 43.8 4
DC 150 1200 1600 840 700 4490 43.6 5
DH 75 1980 1100 800 675 4630 45.0 2
From time to time, both Central and State governments implement several mitigations
and long-term measures to offset the negative effects of climate change on agriculture.
With a view to know the awareness of these measures among the farmers, they were
questioned on six government initiatives (as described in the methodology chapter).
They were asked to state their replies as either ‘yes’, or ‘no’ or ‘not aware’. Figure 7.4
gives a pictorial representation of the awareness of these initiatives. About 91%
of the farmers are well aware of subsidized inputs (like fertilizers, drip irrigation
and machinery), 57% know about field demos on the adaptation practices. Farmers
have no knowledge about all other initiatives like weather-based crop insurance
schemes and establishment of village knowledge centres. To sum up, farmers had
poor knowledge of government initiatives, and so suitable mechanisms should be
formed by the government to popularize government initiatives these two initiatives.
Table 7.7 provides a summary of the opinions of farmers on government initiatives.
Field demos on the listed Organised field days on the

adaptaƟon pracƟces adaptaƟon pracƟces
Not Yes
aware 3%
30% Not No
Yes aware 27%
57% 70%
No
13%
Subsidised inputs Weather based crop

No insurance schemes
1% Not
aware
Yes 8% Not
No
91% aware
51%
49%
Establishment of village knowledge centres
No
32%
Not aware
68%
Fig. 7.4 Opinions of farmers of western region on various government initiatives

Table 7.7 Farmers’ opinions on government initiatives

Govt. initiatives Yes No Not
aware
Field demos on the listed adaptation practices 59 13 31
Organized field days on the adaptation practices 3 28 72
Subsidized inputs (fertilizer, drip and machinery) 94 1 8
Weather-based crop insurance schemes 0 53 50
Establishment of village knowledge centres, climate cell, etc. 0 33 70
7.5.1 Data
One important objective of the study is to estimate the effects of climate variables
on the yield of rice crop using historical data. Hence, we used a time series data set
of yield and climate variables pertaining to 31 years from 1980 to 2010. The data
on temperature corresponds to the rice growing season and the time series data on
rainfall was collected for the entire year to account for rainfall falling directly on the
crop and interseasonal water accumulation within a year (Iski and Devadoss 2006;
Kim and Pang 2009). The summary statistics of all the variables are discussed below:
The yield data (Table 7.8; Fig. 7.5) shows a general upward trend over the years.
Trend equation fitted to yield data shows that the coefficient of trend is strongly
significant (1% level).
The maximum temperature from 1980 to 2010 had an overall average of 32.08 °C
with a standard deviation of 0.546 °C. It shows a slightly increasing trend over the
years (Fig. 7.6) (Table 7.9).
The minimum temperature had an average of 20.83 °C with a standard deviation of
1.11 °C. Compared with maximum temperature, it had greater variability. Further,
the minimum temperature had a clear increasing trend over the years (Fig. 7.7)
(Table 7.10).
(d) Rainfall (mm)
During the study period, 1980–2010, the rainfall in the region had an average of
1593 mm with a standard deviation of 472 mm. It had a wide range between 961
and 2439 mm. Further, the rainfall did not show any trend over the years (Fig. 7.8)
(Table 7.11).
Table 7.8 Summary Mean Std. Min. Max. Trend equation R2

statistics on rice yield and
trend equation (tons/ha) 1.46 0.31 0.81 1.96 Y = 1.12 + 0.0227*** 0.45
*** denotes significance at 1% level
Fig. 7.5 Trend in yield of rice per ha in Gujarat
33.5
33
32.5
32
31.5
31
30.5
1980 1985 1990 1995 2000 2005 2010
Year
Fig. 7.6 Trend in maximum temperature in Gujarat

Table 7.9 Descriptive Mean Minimum Maximum Std. dev.

statistics of maximum
temperature in Gujarat (°C) 32.08 30.76 33.34 0.546
22.5
22
21.5
21
20.5
20
19.5
19
18.5
1980 1985 1990 1995 2000 2005 2010
Year
Fig. 7.7 Trend in minimum temperature in Gujarat

statistics of minimum
temperature in Gujarat (°C) 20.83 18.59 22.49 1.11

statistics of rainfall (mm) in
Gujarat 1593 961 2439 472
The normality of yield data was tested using Shapiro–Wilk W test and Lilliefors test.
The computed value of Shapiro–Wilk W test was 0.9316 with a p-value of 0.0584.
greater than 0.05. So, both tests did not reject the null hypothesis that yield data is
normal.
2600
2400
2200
2000
Rainfall (mm)
1800
1600
1400
1200
1000
800
1980 1985 1990 1995 2000 2005 2010
Year
Fig. 7.8 Trend in rainfall in Gujarat
in yield, Just-Pope production function (Just and Pope 1978, 1979) was employed.
The dependent variable was yield, and independent variables were maximum and
minimum temperatures and rainfall (introduction of square and interaction terms of
maximum temperature, minimum temperature and rainfall resulted in convergence
problems for estimation and so they were omitted). Table 7.12 gives the results of
the estimated empirical models.
Table 7.12 shows that trend and all the climate variables are strongly significant
for the mean function implying that climate change effects are important. For the
variance function, minimum temperature and rainfall have significant negative effects
and so they are risk-decreasing variables. That is, when the values of these variables
are increased, variability in yield decreases.
for both mean function and variance function. The elasticity for the maximum tem-
perature is −1.8082 implying that 1% increase in maximum temperature will induce
a decrease of 1.81% in mean yield. Similarly, the elasticity coefficient for minimum
temperature is 0.5978 and so 1% increase in minimum temperature will increase the
mean yield by a small percentage of 0.6%. The elasticity coefficient for rainfall is
Table 7.12 Estimation Coefficient S. error p-value

results of Just–Pope
production function Mean function
Constant (β0 ) 2.5794 1.8944 0.173
Trend (β1 ) 0.0146** 0.0058 0.011
MaxT (β2 ) −0.0827** 0.0399 0.038
MinT (β3 ) 0.0429** 0.0196 0.0287
Rain (β4 ) 0.0003*** 0.0001 0
Variance function
Constant −1.7338 19.3450 0.929
Trend 0.1356 0.0922 0.142
MaxT 0.8925 0.5541 0.107
MinT −1.4207* 0.7386 0.054
Rain −0.0025*** 0.0009 0.004
Model statistics
Log-likelihood 16.931
AIC −13.862
*, **, *** denotes significance at 10%, 5% and 1% level
Table 7.13 Elasticities of Yield function Climate variable Elasticity

Rainfall 0.2797
Variance Maximum temperature 28.4503
Minimum temperature −28.9105
Rainfall −3.9189
0.2797 which is very small and so 1% increase in rainfall will induce a small increase
of 0.3% in mean yield. When all the three variables are simultaneously increased by
1%, we can expect a small decrease of 0.93% in yield.
−28.4503 implying an increase in variability in yield of 28.5% for 1% increase in
temperature and rainfall. All together, we can expect a decrease of 4.4% in yield
variability when all the three variables are increased by 1%. So, we conclude that all
the three variables put together are risk decreasing in this sense.
T max T min Rainfall Mean yield Yield variability
RCP4.5 2021–2050 5.83 9.65 23.72 1.9 −206.0
2071–2100 8.15 15.40 4.58 −5.8 −224.3
RCP8.5 2021–2050 5.44 10.26 38.57 5.6 −285.4
2071–2100 15.26 27.11 98.82 14.7 −729.9
climate change scenarios, viz. RCP4.5 and RCP8.5 (as discussed in Chap. 3) were
considered and the results are given in Table 7.14.
The table shows that there will be an increase in percentages of all the three
climatic variables during 2021–2050 and 2071–2100 under both scenarios RCP4.5
and RCP8.5. There will be an increase of 5–6% in maximum temperature in the
first period. During the second period, the increase will be between 8 and 15%. The
predicted increase in minimum temperature is higher. It varies from about 10% for
the first period to 27% for the second period. RCP4.5 scenario predicts an increase
of 24% in rainfall during the first period, whereas it predicts only 5% increase for the
second period. The RCP8.5 scenario predicts very high increase in rainfall varying
between 39 and 99% in the two periods.
The table shows that as per RCP4.5 and RCP8.5 scenarios, there will be a marginal
increase in percentages of mean rice yield, viz. 1.9 and 5.6% during the period
2021–2050. Further, during this period, there will a drastic reduction in variability
of yield varying from 206 to 285%. For the period 2071–2100, according to RCP4.5
scenario, there will be a reduction in mean yield of 5.8% while RCP8.5 scenario
predicts an increase in mean yield of 14.7%. There will be a reduction of 224 and
730% variability in yield as per the two scenarios during 2071–2100.
7.6 Summary
Irregular weather is the most important climate change event (Rank 1) perceived by
rice farmers in western region, followed by temperature fluctuation and untimely
rain. Farmers ranked loss of income as the most important shock, followed by a loss
of assets.
Farmers in the region have very good awareness of subsidized inputs and field
demos on the adaptation practices. But, they do not have knowledge about all other
initiatives like weather-based crop insurance schemes and establishment of village
7.6 Summary 159
knowledge centres, thus, warranting suitable measures by the government to popu-

larize government schemes.
Based on historical data, the yield of rice shows a marginal upward trend in the
region over the years, although there was a decline in yield in some years. Both
maximum and minimum temperatures exhibited an increasing trend over the years.
Annual rainfall had a wide range with no trend over the years. Maximum temperature
has a significant negative effect on rice yield, whereas minimum temperature and
rainfall have significant negative effects. There is no nonlinear effect of the climate
variables on rice yield.
The two climate change scenarios, viz. RCP4.5 and RCP8.5 predict a marginal
increase in percentages of mean rice yield during the period 2021–50. But during this
period, there will a drastic reduction in variability of yield. For the period 2071–2100,
the two scenarios predict inconsistent changes in mean yield.
Chapter 8
in Central Region of India

8.1 Introduction
The Central Region for the present study covers two states: Madhya Pradesh and
Maharashtra. The districts selected from the two States were: Balaghat and Bhandara
respectively. As stated in the methodology chapter, the present study is based on
(i) primary data: farm survey among 70 farmers from Madhya Pradesh and 100
farmers from Maharashtra of the two districts and (ii) Secondary data: time series
data (1980–2010) on productivity (yield) of rice and climate variables from the two
districts. Figure 8.1 gives the map of the study area.
Among the 170 farmers surveyed, 166 were males and 4 were females. All the 70
farmers in Madhya Pradesh, have dry land-based crop typology while in Maharashtra,
31 farmers have irrigated crop based farming, 64 farmers have irrigated crop with
supplemental irrigation and 5 farmers have dry land irrigation. The average age
of the farmers in the two States are respectively, 46 and 51 years and farmers in
Maharashtra are relatively older than those in Madhya Pradesh. In general, about
6% of the farmers have no formal education, 21% have studied up to 5th, 55% have
studied up to primary level and the rest, 18% have studied up to secondary level and
above (Fig. 8.2). About 59% of the farmers in Maharashtra are having education up
to primary level whereas about 43% farmers in Madhya Pradesh have education level
of secondary and above.

162 8 Climate Change Impacts on Rice Yield in Central …
59
60 55
50
50
43
40
30 Madhya Pradesh
30
21 Maharastra
18
20 Total
11
10 6 7
0 0
0
No formal Upto 5th Primary level Secondary &
educaƟon above
Fig. 8.2 Educational status of surveyed farmers

Farmers in the study region had irrigated lands supported by wells as the major source
of irrigation. In addition to their traditional gravity based irrigation, they also used
drip and sprinkler systems. Most of the farmers in these states derive income from
rice and other crops mainly during Kharif season although few are raising other crops
during Rabi season.
(i) Land owned: All the farmers own irrigated lands. Table 8.1 gives the summary
statistics of the land size.
The mean area under irrigation varies from 1.41 to 3.37 ha across the two states
and the overall mean is 2.22 ha. Land holdings in Madhya Pradesh has relatively
higher standard deviation (2.30 ha) and the minimum and maximum holdings are
respectively 0.81 and 16.19 ha in Madhya Pradesh and 0.20 and 7.28 ha in Maha-
rashtra.
(ii) Wells owned: Farmers in Madhya Pradesh possess only open wells whereas
those from Maharashtra possess both open and bore-wells. Table 8.2 gives the
frequency distribution of number of wells owned by the farmers.
Out of 70 farmers in Madhya Pradesh, only 22 (31%) own one or two open wells
and the remaining 69% don’t own any well at all. About 42% of the farmers in
Maharashtra possess, either open or bore well. Overall about 75% of the farmers in
the region don’t own open wells and 46% don’t own bore wells.
Table 8.1 Summary statistics of irrigated land owned (hectares)

State Mean S.D. Min Max
Madhya Pradesh 3.37 2.30 0.81 16.19
Maharashtra 1.41 1.42 0.20 7.28
Overall 2.22 2.07 0.20 16.19
Table 8.2 Distribution of number of wells owned by farmers

No. of wells Number of farmers
Madhya Pradesh Maharashtra Total
Open wells Open wells Bore wells Open wells Bore wells
0 48 79 79 127 79
1 17 15 19 32 19
2 5 5 1 10 1
3 – – 1 – 1
>3 – 1 – 1 –
Fig. 8.3 Frequency

distribution of sources of
100
irrigation
90 38
80
70
Wells
60
50 Canals
40 27 62 Tanks
30
0
20
19
10
0
0
Madhya Pradesh Maharastra
Table 8.3 Rice area (ha) cultivated by farmers in the region

States Rice Other crops
Kharif Rabi Summer Kharif Rabi Summer
Madhya Pradesh 3.37 (70) 0 (0) 0 (0) 0 (0) 1.33 (27) 0 (0)
Maharashtra 1.21 (81) 1.21 (2) 0.92 (13) 1.72 (20) 0 (0) 0 (0)
season
(iii) Sources of irrigation: Like in other regions, farmers in this region also bene-
fitted from three sources of irrigation, viz., canals, wells and tanks. Farmers of
Madhya Pradesh mostly use only wells and tanks whereas Maharashtra farmers
mostly use wells and canals (Fig. 8.3).
(iv) Investment in irrigation: Field surveys in the selected districts in the states
indicated that the main investment type is well investment followed by invest-
ment in micro irrigation systems like drip and sprinkler irrigation.
(v) Area under rice: All farmers in Madhya Pradesh have cultivated rice during
Kharif season only and 81% of the farmers in Maharashtra have grown rice
during this season. During this season, average area per farmer under rice in
the two states are respectively 3.37 and 1.21 ha. In Madhya Pradesh, out of
70 farmers surveyed, 46 farmers have irrigated rice with an average area of
2.2 ha. During Rabi and summer seasons only very small number of farmers
in Maharashtra have grown rice. Some farmers in Madhya Pradesh grow other
crops during Rabi season. In Maharashtra other crops are grown only during
Kharif season (Table 8.3).

The detailed survey conducted in the region helped to get an idea of farmers percep-
tion on the climate change, its effects on agriculture and various adaptation strate-
gies/techniques followed by the farmers and the details are discussed below.
according to their perceptions of climate change. To find what the farmers perceive
as the most significant effects of climate change, Garrett’s ranking technique as
described in Chap. 4 was used. The results are given in Tables 8.4 and 8.5. Table 8.5
shows that the drought is the most important climate change event (Rank 1) perceived
by Rice farmers in Central Region with an average Garret Score of 68.1. As given in
Table 8.4 it was given Rank 1 by 45 farmers and rank 2 by 35 farmers. The second and
third important climate change events are increase in pest damage and untimely rains
respectively. Similar analysis was done separately for the two states in the region
and the results are presented in Appendix (Tables 8.18, 8.19, 8.20, 8.21).
the events, thus, the farmers’ perceptions of climate change effects differ across the
sample chosen. The same conclusion was arrived at by applied Kendall’s coefficient
of concordance also. Similar analysis was done for individual states in the region.
The results show that there is no consistency in the rankings of the events by the
farmers.
They were asked to rank 7 effects of climatic shocks, as described in methodology
chapter. Farmers ranked loss of income as the most important shock (Tables 8.6 and
8.7). Out of 170 farmers, 61 of them have given rank 1 and 77 of them have given
rank 2 to this effect. The second and third ranks are given to decline in crop yield
and loss of assets respectively.
166
Table 8.4 Factors influencing farmers perception of climate change—Garrett ranking: Central Region
Factors Rank Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 45 35 20 12 12 4 1 1 1 38 1 0 0 0 0 0 170
Hailstorm 17 2 4 2 20 0 0 0 1 53 0 13 20 10 15 13 170
Flood 29 10 3 3 0 1 13 14 9 2 5 0 21 59 0 1 170
AD 0 9 25 14 2 2 13 12 20 6 0 11 7 42 0 7 170
IPD 20 65 20 3 14 2 6 7 20 4 0 2 4 0 1 2 170
IFM 9 5 26 23 10 1 0 0 0 6 1 13 13 35 13 15 170
UR 9 7 24 50 57 10 10 0 0 0 0 3 0 0 0 0 170
IW 5 1 13 19 9 4 4 2 90 12 10 1 0 0 0 0 170
TFH 0 0 0 0 1 105 26 10 10 4 4 7 0 0 0 3 170
TFL 0 0 5 6 7 2 105 22 8 4 8 0 1 1 0 1 170
SSDI 0 0 0 1 1 3 6 7 0 14 97 20 10 10 1 0 170
CCP 0 0 0 18 18 7 1 5 0 0 10 90 9 4 1 7 170
CLA 0 0 0 0 2 8 3 12 7 7 31 0 0 0 100 0 170
CFI 0 0 0 0 4 3 1 3 0 6 0 39 3 9 5 97 170
CF 0 0 1 1 0 14 8 6 3 0 7 10 104 6 9 1 170
DGW 0 0 0 1 10 8 6 111 26 7 0 0 0 0 0 1 170
Total 134 134 141 153 167 174 203 212 195 163 174 209 192 176 145 148 2720
Garret-score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
8 Climate Change Impacts on Rice Yield in Central …
Table 8.5 Rankings of factors representing climate change using Garrett ranking: Central Region
Factors Rank
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Total Average Rank
Total score score score
Drought 3870 2660 1400 780 732 232 55 52 48 1710 42 0 0 0 0 0 11581 68.1 1
Hailstorm 1462 152 280 130 1220 0 0 0 48 2385 0 507 700 300 360 182 7726 45.4 11
Flood 2494 760 210 195 0 58 715 728 432 90 210 0 735 1770 0 14 8411 49.5 8
AD 0 684 1750 910 122 116 715 624 960 270 0 429 245 1260 0 98 8183 48.1 9
IPD 1720 4940 1400 195 854 116 330 364 960 180 0 78 140 0 24 28 11329 66.6 2
IFM 774 380 1820 1495 610 58 0 0 0 270 42 507 455 1050 312 210 7983 47.0 10
UR 774 532 1680 3250 3477 580 550 0 0 0 0 117 0 0 0 0 10960 64.5 3
IW 430 76 910 1235 549 232 220 104 4320 540 420 39 0 0 0 0 9075 53.4 6
TFH 0 0 0 0 61 6090 1430 520 480 180 168 273 0 0 0 42 9244 54.4 4
TFL 0 0 350 390 427 116 5775 1144 384 180 336 0 35 30 0 14 9181 54.0 5
SSDI 0 0 0 65 61 174 330 364 0 630 4074 780 350 300 24 0 7152 42.1 13
CCP 0 0 0 1170 1098 406 55 260 0 0 420 3510 315 120 24 98 7476 44.0 12
CLA 0 0 0 0 122 464 165 624 336 315 1302 0 0 0 2400 0 5728 33.7 15
CFI 0 0 0 0 244 174 55 156 0 270 0 1521 105 270 120 1358 4273 25.1 16
CF 0 0 70 65 0 812 440 312 144 0 294 390 3640 180 216 14 6577 38.7 14
DGW 0 0 0 65 610 464 330 5772 1248 315 0 0 0 0 0 14 8818 51.9 7
167

Factors Rank Total
1 2 3 4 5 6 7
DCY 88 37 0 0 0 0 45 170
LA 21 18 32 24 56 2 17 170
LI 61 77 7 9 16 0 0 170
FIS 0 20 18 89 11 28 4 170
DL 0 9 79 28 38 2 14 170
DC 0 9 20 16 46 69 10 170
DH 0 0 14 4 3 69 80 170
Total 170 170 170 170 170 170 170 1190
Garret-score 79 66 57 50 43 34 21
Table 8.7 Garret ranking of effects of climatic shocks

1 2 3 4 5 6 7 score score
Total score
DCY 6952 2442 0 0 0 0 945 10,339 60.82 2
LA 1659 1188 1824 1200 2408 68 357 8704 51.20 3
LI 4819 5082 399 450 688 0 0 11,438 67.28 1
FIS 0 1320 1026 4450 473 952 84 8305 48.85 5
DL 0 594 4503 1400 1634 68 294 8493 49.96 4
DC 0 594 1140 800 1978 2346 210 7068 41.58 6
DH 0 0 798 200 129 2346 1680 5153 30.31 7
Farmers in both the strategies followed several mitigation strategies to tide over
the negative impacts of climate change (Table 8.8). It shows that out of 70 farmers
in Madhya Pradesh, 65 of them (93%) have raised crop only in one season, 68
farmers (97%) have sold livestock and 44 farmers (63%) have leased part of land and
50% of the farmers have shifted to other crops. Also, all the 70 farmers (100%) in
Madhya Pradesh borrowed money and received assistance from government. The low
income from crops have forced 53 farmers (75%) to go for less food consumption, 61
farmers (81%) have shifted to non-farm employment and 64 (91%) of farmers have
out migrated to cities. In the case of Maharashtra, out of 100 farmers, 50 farmers
(50%) shifted to cities, 32 farmers (32%) left land fallow, 30 farmers (30%) received
support from Government, 25 farmers (25%) shifted to non-farm employment. The
above discussion brings forth the severity of the effect of climate change on the
livelihood of the farmers.

Strategy Madhya Maharashtra Total
Pradesh
Did nothing 18 10 28
Left land fallow 52 32 84
Sold part of land 35 0 35
Leased out part of land 44 0 44
Restricted number of seasons for rice growing to 1 65 11 76
Sold livestock (cows, buffalos) 68 0 68
Shifted to other crops 35 15 50
Borrowed money from relatives/others 70 10 80
Received assistance from government/NGOs 70 30 100
Less food consumption 53 20 73
Shifted to non-farm employment 61 25 86
Out migration to cities 64 50 114
From time to time both Central and State governments implement several mitigation
and long term measures to offset the negative effects of climate change on agriculture.
With a view to know the awareness of these measures among the farmers, they
were questioned on seven government initiatives (as described in the methodology
chapter). They were asked to state their replies as either ‘yes’, or ‘no’ or ‘not aware’.
Figure 8.4 gives a pictorial representation of the awareness on these initiatives by
the farmers belonging to the entire region.
In general, the government initiatives are not popular among farmers. Majority of
the farmers said “No” or “not aware”. However, among the farmers who are aware of
the programs, about 38% of the farmers know the introduction of new crop varieties,
27% have good understanding of water regulation by government through irrigation
department and only 11% of the farmers have understanding on the subsidised inputs
(like fertilizers, drip irrigation system and machinery) and establishment of village
knowledge centres. Only 9% of the farmers participated in the field demos of the
listed adaptation practices. Further, weather based cropping insurance schemes are
not popular among the farmers as only 1% of them is aware of it. Thus, it could be
concluded that except (i) introduction of new crop varieties, and (ii) water regulation
by government through irrigation department, all other government initiatives are
not much popular among farmers. This conclusion calls for formulating suitable
popularisation mechanisms by the government. Table 8.9 provides a summary of the
opinions on government initiatives.
Water regulation through irrigation Introduction of new crop

department varieties
Not
Not Yes Yes
aware
aware 27% 14% 38%
15%
No
48%
No
58%
Field demos on the listed Yes Not Organised field days on the
adaptation practices 9% aware adaptation practices
Not
44% Yes
aware
1%
41%
No No
50% 55%
Subsidised inputs Yes Weather based crop insurance

Not 11% schemes Yes
aware 1%
Not
31%
aware
52%
No No
58% 47%
Establishment of village knowledge centres

Yes
11%
Not aware
51% No
38%
Fig. 8.4 Opinions of farmers of Central region on various government initiatives

Table 8.9 Farmers opinions on government initiatives
Govt. initiations Madhya Pradesh Maharashtra Total Chi- Df p-value
Yes No Not Yes No Not Yes No Not square
aware aware aware
Water regulation 45 20 5 0 79 21 45 99 26 87.44 2 <0.00001
through irrigation
department
Introduction of new 64 2 4 0 80 20 64 82 24 148.18 2 <0.00001
crop varieties
Field demos on the 0 18 52 16 66 18 16 84 70 56.41 2 <0.00001
listed adaptation
practices
Organised field days 0 15 55 2 78 20 2 93 75 57.51 2 <0.00001
on the adaptation
practices
Subsidized inputs 4 20 46 15 79 6 19 99 52 69.16 2 <0.00001

machinery)
Weather based crop 2 0 68 0 80 20 2 80 88 106.19 2 <0.00001
insurance schemes
Establishment of 1 0 69 18 64 18 19 64 87 107.15 2 <0.00001
village knowledge
centres, climate cell
etc.
171
between the two states. This may be because the government initiatives depend on
8.5.1 Data
corresponds to the growing season of rice variety in the respective states and the time
series data on rainfall was collected for the entire year to account for rainfall falling
directly on the crop and inter-seasonal water accumulation within a year (Iski and
Devadoss 2006; Kim and Pang 2009). The summary statistics of all the variables are
discussed below:
The yield data (Table 8.10; Fig. 8.5) shows an upward trend in Madhya Pradesh
as the coefficient of time is positive and strongly significant. But the yield data for
Maharashtra do not exhibit such trend even though it had higher yield (per ha) than
that of Madhya Pradesh during most of the years (Fig. 8.5).
The maximum temperature had an overall average of 32.14 °C. Maharashtra had
the highest mean maximum temperature of 32.85 °C. Further for both states, the data
do not exhibit any trend and the coefficients of time are not significant (Fig. 8.6;
Table 8.11).
The minimum temperature had an overall average of 19.51 °C. Madhya Pradesh
had the lowest mean minimum temperature of 18.61 °C. Also, the minimum tem-
perature of Maharashtra was higher than that of Madhya Pradesh for all the 31 years

State N Mean Min Max Std. Trend equation R2
Dev.
Madhya Pradesh 31 2.11 0.96 3.82 0.91 Y 0.736 + 0.736
0.086*** t
Maharashtra 31 3.02 1.24 4.34 0.774 Y 3.105 − 0.004
0.0051 t
Overall 62 2.57 0.96 4.34 0.957 – –
***Significant at 1% level; D1 and D2 are dummy variables representing Punjab and Haryana states
1.8
MP
MHA
1.6
1.4
Yield (tonnes per ha)
1.2
0.8
0.6
0.4
0.2
1980 1985 1990 1995 2000 2005 2010
Year
Fig. 8.5 Trend in yield of rice in Central Region States of India
33.5
MP
MHA
Maximum Temperature(Centigrade)
33
32.5
32
31.5
31
30.5
1980 1985 1990 1995 2000 2005 2010
Year
Fig. 8.6 Trend in maximum temperature in Central Region of India


Dev.
Madhya Pradesh 31 31.43 30.73 32.17 0.065 Y 31.33 + 0.03
0.0064T
Maharashtra 31 32.85 31.63 33.46 0.417 Y 32.87 − 0.001
0.001T
Overall 62 32.14 30.73 33.46 0.811 – –
21.5
MP
MHA
21
20.5
20
19.5
19
18.5
18
17.5
1980 1985 1990 1995 2000 2005 2010
Year
Fig. 8.7 Trend in minimum temperature in Central Region of India

Dev.
Madhya Pradesh 31 18.61 17.94 19.37 0.368 Y 18.536 + 0.014
0.0048T
Maharashtra 31 20.40 19.67 21.41 0.417 Y 20.099 + 0.179
0.019**T
Overall 62 19.51 17.94 21.41 0.986 – –
** denotes significance at 5% level
(Fig. 8.7). Further, the data of Maharashtra showed an increasing trend as the coef-
ficients of time was statistically significant (Table 8.12) .

Dev.
Madhya Pradesh 31 1531 883 2636 377 Y 1555.6 − 0.001
1.55T
Maharashtra 31 1224 771 1968 261 Y 1215.7 + 0.0003
0.496T
Overall 62 1377 771 2636 357 – –
2800
MP
2600 MHA
2400
2200
2000
Rainfall (mm)
1800
1600
1400
1200
1000
800
600
1980 1985 1990 1995 2000 2005 2010
Year
Fig. 8.8 Trend in rainfall in Central Region of India
(d) Rainfall (mm)

The region had an overall mean rainfall of 1377 mm with a standard deviation
of 357 mm (Table 8.13). Annual rainfall of Madhya Pradesh was higher than that
of Maharashtra during most of the years (Fig. 8.8). The data for both states did
not exhibit any trend over time as the coefficients of trend were statistically non
significant.
The normality of yield data was tested using Shapiro-Wilk W test and Lilliefors test.
The computed value of Shapiro-Wilk W test was 0.9450 with a p-value of 0.0769.
Table 8.14 Panel root test results using Fisher’s type Tests
statistic statistic
for for
no trend trend
Yield 9.3555 0.0528* Reject H0 9.8813 0.0425** Reject H0
tempera-
ture
tempera-
ture
greater than 0.05. So, both tests did not reject the null hypothesis that yield data is
normal.
8.5.3 Panel Root Test: Fisher’s Type Tests
Since data on rice yield for the 2 States for 30 years is available as a panel data set, it
was subjected to panel root testing before modelling. As stated in the methodology
Chap. 4, Fisher’s type panel root test was applied and the results are presented in
Table 8.14.
8.5.4 Estimation of Just-Pope Production Function
in yield, Just-Pope Production function (Just and Pope 1978 & 79) with yield as
The linear model was selected based on AIC criterion. It has the least AIC value.
For the mean function maximum temperature has a strong negative effect on rice
yield. It shows that every 1 °C increase in maximum temperature will reduce the
Table 8.15 Estimation results of Just-Pope production function

Constant (β0 ) 5.9449 3.0164 2.3181 106.2396
Trend (β1 ) 0.0299*** 0.0035 0.0301 0.0036
MaxT (β2 ) −0.2065*** 0.0747 2.9486 7.6797
MinT (β3 ) 0.0879 0.0816 −4.1877 4.8272
RainT (β4 ) −0.000079 0.000086 −0.0094 0.0098
MaxT 2 (β5 ) – – −0.1190 0.1406
MinT 2 (β6 ) – – −0.0730 0.0966
RainT 2 (β7 ) – – 0.0000 0.0000
MaxT MinT (β8 ) – – 0.2170 0.1409
MaxT RainT (β9 ) – – 0.0003 0.0004
MinT RainT (β10 ) – – 0.0001 0.0003
MP (β11 ) −0.5866*** 0.2081 −0.6413 0.2456
Variance
Constant −8.9320 15.9732 −17.1926 15.1368
Time 0.0207 0.0241 0.0340 0.0301
MaxT −1.0253 0.7755 −0.5949 0.6432
MinT 1.9960*** 0.6367 1.6823 0.4858
RainT −0.000051 0.0008478 0.0001 0.0008
LL −4.9068 −1.8315
AIC 31.8136 37.6630
BIC 55.2120 73.8243
***, **, *Significant at 1, 5 and 10% levels
average yield per ha by 0.21 tonnes. Trend has a positive coefficient and it significant
at 1% level. Maximum temperature and rainfall are statistically non-significant. The
coefficient for Maharashtra (dummy variable) is negative and significant implying
that, on an average, difference in mean yield (per ha) between these two states is
0.5866 assuming all other variables fixed. So, Maharashtra has lower productivity
compared to that of Madhya Pradesh. The minimum temperature had strong positive
effect on yield variability. Its coefficient in the variance function is 1.996 which
shows that yield variability increases by 1.996 for every 1 °C increase in minimum
temperature. Thus, minimum temperature is risk increase variable for rice yield.
perature is −6.38 implying that 1% increase in maximum temperature will induce a
decrease of 6.38% in mean yield. Similarly, the elasticity coefficient for minimum
Table 8.16 Elasticities of climatic variables at means

Yield function Climate variable Elasticity
Rainfall −0.104689
Rainfall −0.0688501
T max T min Rainfall Mean yield Yield
variability
RCP4.5 2021–2050 4.1615 8.2665 18.2199 −14.85 183.55
2071–2100 8.6017 16.1215 5.5231 −28.92 344.13
RCP8.5 2021–2050 4.5133 10.0345 34.9925 −15.94 239.55
2071–2100 14.29309 29.36223 74.89477 −50.71 667.27
mean yield by 1.65%. The elasticity coefficient for rainfall is −0.10 which is very
small and so 1% increase in rainfall will induce a small decrease of 0.10% in mean
yield. When all the three variables are simultaneously increased by 1%, we can expect
a substantial decrease of 4.8% in yield.
For the variance function, the elasticity coefficient of minimum temperature is
38.84 implying an increase in variability in yield of 38.8 for 1% increase in mini-
mum temperature. Similar interpretations can be given for the elasticities of maxi-
mum temperature and rainfall, even though their coefficients are not significant. All
together, we can expect an increase of 6.0% in yield when all the three variables are
increased by 1%. So, we conclude that all the three variables, put together are risk
increasing in this sense.
The estimated elasticities given in Table 8.16 can further be used to estimate the
impacts of climate change on mean yield and yield variability. For this purpose, two
climate change scenarios, viz., RCP4.5 and RCP8.5 (as discussed in Chap. 3) were
considered (Table 8.17).
Table 8.17 shows that, as per RCP4.5 and RCP8.5 scenarios, there will be substan-
tial reduction in percentages of rice yield during the periods 2021–50 and 2071–2100.
Under both scenarios, percentage reduction in the later period will be more. The per-
centage reduction in mean yield varies from 14.85 to 15.94 during the first period
as per the two scenarios and as per RCP8.5 scenario it varies from 28.92 to 50.71%
during the second period.
yield variability during 2021–2050 while the second scenario predicts an increase
of 240% for the same period. For the second period, RCP4.5 predicts an increase
of 344% in yield variability whereas RCP8.5 scenario predicts an increase in yield
variability of 667%.
8.6 Summary
Farmers in the region perceive drought followed by increase in pest damage and
untimely rain as the most important climate change events. Farmers ranked loss of
income as the most important climate change shock followed by decline in crop yield
and loss of assets. Farmers in both the states followed different mitigation strategies.
Farmers, in general, are not aware of government initiatives like field demos and
weather based cropping insurance schemes as only less than 40% of them know
about them. Hence it is very important that state governments need to take suitable
actions to popularise these initiatives among farmers.
Historical data on rice yield shows an upward trend in Madhya Pradesh, but the
data for Maharashtra had no such trend over years. Maharashtra had the higher max-
imum and minimum temperatures than that of Madhya Pradesh. Annual rainfall of
Madhya Pradesh was higher than that of Maharashtra during most of the years. Max-
imum temperature had a strong negative effect on mean rice yield while minimum
temperature had strong positive effect on yield variability. So the minimum temper-
ature is risk increasing variable for rice yield. There is no nonlinear effect of climate
variables on rice yield which is in contrast to the results obtained in other regions.
The two climate change scenarios viz., RCP4.5 and RCP8.5, predict substantial
reduction in percentages of rice yield during the periods 2021–2050 and 2071–2100.
RCP4.5 scenario for both periods.
Appendix
180
Table 8.18 Factors influencing farmers perception of climate change—Garrett ranking: Madhya Pradesh
Factors Rank Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 21 12 7 11 12 4 1 1 1 0 0 0 0 0 0 0 70
Hailstorm 0 0 0 1 0 0 0 0 0 0 0 13 20 8 15 13 70
Flood 0 0 0 0 0 1 13 14 9 0 5 0 21 6 0 1 70
AD 0 0 0 0 0 2 13 12 9 0 0 0 7 21 0 6 70
IPD 0 16 0 0 11 2 6 7 18 4 0 1 4 0 1 0 70
IFM 0 0 0 0 0 0 0 0 0 5 0 13 13 12 13 14 70
UR 8 6 15 12 10 9 10 0 0 0 0 0 0 0 0 0 70
IW 5 0 13 14 1 4 4 2 4 12 10 1 0 0 0 0 70
TFH 0 0 0 0 0 18 25 10 10 4 3 0 0 0 0 0 70
TFL 0 0 5 6 7 2 8 22 8 4 8 0 0 0 0 0 70
SSDI 0 0 0 0 0 2 6 7 0 14 0 20 10 10 1 0 70
CCP 0 0 0 7 14 0 1 5 0 0 10 12 9 4 1 7 70
CLA 0 0 0 0 2 8 3 12 7 7 31 0 0 0 0 0 70
CFI 0 0 0 0 0 0 0 2 0 6 0 39 3 9 5 6 70
CF 0 0 1 1 0 14 7 6 3 0 7 10 5 6 9 1 70
DGW 0 0 0 1 10 8 6 12 26 7 0 0 0 0 0 0 70
Total 34 34 41 53 67 74 103 112 95 63 74 109 92 76 45 48 1120
Garret- 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
score
Appendix
Table 8.19 Rankings of factors representing climate change using Garrett ranking: Madhya Pradesh
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 score score
Total score
Drought 1806 912 490 715 732 232 55 52 48 0 0 0 0 0 0 0 5042 72.0 1
Hailstorm 0 0 0 65 0 0 0 0 0 0 0 507 700 240 360 182 2054 29.3 16
Flood 0 0 0 0 0 58 715 728 432 0 210 0 735 180 0 14 3072 43.9 9
AD 0 0 0 0 0 116 715 624 432 0 0 0 245 630 0 84 2846 40.7 13
IPD 0 1216 0 0 671 116 330 364 864 180 0 39 140 0 24 0 3944 56.3 4
IFM 0 0 0 0 0 0 0 0 0 225 0 507 455 360 312 196 2055 29.4 15
UR 688 456 1050 780 610 522 550 0 0 0 0 0 0 0 0 0 4656 66.5 2
IW 430 0 910 910 61 232 220 104 192 540 420 39 0 0 0 0 4058 58.0 3
TFH 0 0 0 0 0 1044 1375 520 480 180 126 0 0 0 0 0 3725 53.2 6
TFL 0 0 350 390 427 116 440 1144 384 180 336 0 0 0 0 0 3767 53.8 5
SSDI 0 0 0 0 0 116 330 364 0 630 0 780 350 300 24 0 2894 41.3 12
CCP 0 0 0 455 854 0 55 260 0 0 420 468 315 120 24 98 3069 43.8 10
CLA 0 0 0 0 122 464 165 624 336 315 1302 0 0 0 0 0 3328 47.5 8
CFI 0 0 0 0 0 0 0 104 0 270 0 1521 105 270 120 84 2474 35.3 14
CF 0 0 70 65 0 812 385 312 144 0 294 390 175 180 216 14 3057 43.7 11
DGW 0 0 0 65 610 464 330 624 1248 315 0 0 0 0 0 0 3656 52.2 7
181
182
Table 8.20 Factors influencing farmers perception of climate change—Garrett ranking: Maharashtra
Factors Rank Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 24 23 13 1 0 0 0 0 0 38 1 0 0 0 0 0 100
Hailstorm 17 2 4 1 20 0 0 0 1 53 0 0 0 2 0 0 100
Flood 29 10 3 3 0 0 0 0 0 2 0 0 0 53 0 0 100
AD 0 9 25 14 2 0 0 0 11 6 0 11 0 21 0 1 100
IPD 20 49 20 3 3 0 0 0 2 0 0 1 0 0 0 2 100
IFM 9 5 26 23 10 1 0 0 0 1 1 0 0 23 0 1 100
UR 1 1 9 38 47 1 0 0 0 0 0 3 0 0 0 0 100
IW 0 1 0 5 8 0 0 0 86 0 0 0 0 0 0 0 100
TFH 0 0 0 0 1 87 1 0 0 0 1 7 0 0 0 3 100
TFL 0 0 0 0 0 0 97 0 0 0 0 0 1 1 0 1 100
SSDI 0 0 0 1 1 1 0 0 0 0 97 0 0 0 0 0 100
CCP 0 0 0 11 4 7 0 0 0 0 0 78 0 0 0 0 100
CLA 0 0 0 0 0 0 0 0 0 0 0 0 0 0 100 0 100
CFI 0 0 0 0 4 3 1 1 0 0 0 0 0 0 0 91 100
CF 0 0 0 0 0 0 1 0 0 0 0 0 99 0 0 0 100
DGW 0 0 0 0 0 0 0 99 0 0 0 0 0 0 0 1 100
Total 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 1600
Garret-score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
Appendix
Table 8.21 Rankings of factors representing climate change using Garrett ranking: Maharashtra
Factors Rank Average Rank
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Total score
score
Total Score
Drought 2064 1748 910 65 0 0 0 0 0 1710 42 0 0 0 0 0 6539 65.4 2
Hailstorm 1462 152 280 65 1220 0 0 0 48 2385 0 0 0 60 0 0 5672 56.7 5
Flood 2494 760 210 195 0 0 0 0 0 90 0 0 0 1590 0 0 5339 53.4 8
AD 0 684 1750 910 122 0 0 0 528 270 0 429 0 630 0 14 5337 53.4 9
IPD 1720 3724 1400 195 183 0 0 0 96 0 0 39 0 0 0 28 7385 73.9 1
IFM 774 380 1820 1495 610 58 0 0 0 45 42 0 0 690 0 14 5928 59.3 4
UR 86 76 630 2470 2867 58 0 0 0 0 0 117 0 0 0 0 6304 63.0 3
IW 0 76 0 325 488 0 0 0 4128 0 0 0 0 0 0 0 5017 50.2 11
TFH 0 0 0 0 61 5046 55 0 0 0 42 273 0 0 0 42 5519 55.2 6
TFL 0 0 0 0 0 0 5335 0 0 0 0 0 35 30 0 14 5414 54.1 7
SSDI 0 0 0 65 61 58 0 0 0 0 4074 0 0 0 0 0 4258 42.6 13
CCP 0 0 0 715 244 406 0 0 0 0 0 3042 0 0 0 0 4407 44.1 12
CLA 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2400 0 2400 24.0 15
CFI 0 0 0 0 244 174 55 52 0 0 0 0 0 0 0 1274 1799 18.0 16
CF 0 0 0 0 0 0 55 0 0 0 0 0 3465 0 0 0 3520 35.2 14
DGW 0 0 0 0 0 0 0 5148 0 0 0 0 0 0 0 14 5162 51.6 10
183
Chapter 9
in Southern Region of India

9.1 Introduction
The southern region for the present study covered three states: Tamil Nadu (TN),
Andhra Pradesh (AP) and Karnataka (KA), and in each state, one district where rice
is predominantly grown was selected, viz. Thanjavur, West Godavari and Shimoga,
respectively. As stated in the methodology chapter, the present study is based on (i)
primary data: farm survey with 313 farmers constituting 112 farmers from Tamil
Nadu, 101 from Andhra Pradesh and 100 from Karnataka and (ii) secondary data:
time series data (1981–2010) on productivity (yield) of rice and climate variables
from the three districts. The location map of the study region is given in Fig. 9.1.

186 9 Climate Change Impacts on Rice Yield in Southern Region of India
Out of the 313 farmers surveyed, 302 farmers were males. The typology of crop
irrigation differs among the states (Table 9.1). In Tamil Nadu, crop with supplemental
irrigation is followed by all the farmers while 100% of the farmers surveyed in Andhra
Pradesh and 78% in Karnataka have irrigated crop-based typology.
Overall, irrigated crop-based typology is followed by 57% of the farmers and 37%
follow crop with supplemental irrigation and the rest 6% follow a combination of
typologies as shown in Fig. 9.2 and Table 9.1.
The average age of the farmers in the three states was nearly equal and is respec-
tively 48.5, 47.7 and 51.2 years. About 27% of the farmers had no formal education,
29% have studied up to fifth standard, 32% up to primary level and only 12% have
the education of secondary and above (Fig. 9.3).
Among the states, Tamil Nadu had the highest percentage (30) of farmers who
had done education up to the level of secondary and above while Andhra Pradesh
had the highest percentage (55) of farmers with no formal education. Karnataka had
the highest percentage (60) of farmers who had done up to primary level.
Table 9.1 Cropping typologies of farmers in the three states of southern region
State/typology Irrigated Crop with Dry (1) and (1) and (1), (2) Total
crop- Supplemental land- (2) (3) and (3)
based Irrigation (2) based
(1) (3)
Tamil Nadu 0 112 0 0 0 0 112
Andhra Pradesh 101 0 0 0 0 0 101
Karnataka 78 3 0 13 4 2 100
Total 179 115 0 13 4 2 313
Others
6%
Crop with
Supplemental Irrigated
Irrigation Crop Based
37% 57%
Fig. 9.2 Typologies of households in southern region

60
60 55
51
50
40
32 30 TN
27 29
30 25 AP
24
KA
20 1716
13 12 Total
10 5 4
0
0
No formal Upto 5th Primary Secondary
education Standard Level and above
Education level
Fig. 9.3 Educational status of surveyed farmers
Farmers in the study area have wells as the main source of irrigation to their lands.
Irrigation is provided to the crops by adopting different irrigation methods like drip,
sprinkler and improved gravity irrigation. Mostly, rice is grown under assured water
supply situations like canals and tanks where one rice crop is grown during Kharif
season using gravity irrigation methods. In the rabi season and in those areas with
well irrigation sources, farmers grow irrigated dry crops using both gravity and
micro-irrigation methods.
(i) Land owned: Farmers in Tamil Nadu and Karnataka own wet, irrigated and dry
lands, whereas farmers in Andhra Pradesh own only irrigated land. Table 9.2
gives the summary statistics of the land ownership pattern:
Table 9.2 Land owned by farmers in the three states

State Average land area owned (ha)
Wet Irrigated Dry
Tamil Nadu 0.58 0.83 0.47
Andhra Pradesh 0.00 1.95 0.00
Karnataka 1.50 1.47 0.02
Total 0.68 1.40 0.17
Table 9.3 Distribution of number of borewells owned across states. (no. of farmers)
State No. of borewells owned Total
1 2 3 4
Tamil Nadu 32 68 11 1 112
Andhra Pradesh 101 0 0 0 101
Karnataka 99 1 0 0 100
Total 232 69 11 1 313
Overall, the region is predominated by irrigated land with an average area of 1.4 ha
per farmer.
(ii) Wells owned: No farmer possesses open well and they own only borewells due to
the declining groundwater table in the region. The following table gives the fre-
quency distribution of a number of borewells owned by the farmers (Table 9.3).
The table shows that most of the farmers (74%) own at least one borewell and
22% of farmers own 2 borewells.
(iii) Sources of irrigation: Farmers have three sources of irrigation, viz. canal, tanks
and well. In the case of the sample farmers studied, all the sample farmers in
Tamil Nadu use both canal and well for irrigation, whereas all the farmers in
Andhra Pradesh and Karnataka depend on the canal for irrigation mostly for
growing rice in the kharif season.
(iv) Investment in irrigation: Most of the irrigation investments in the regions are
for digging and drilling new wells. Farmers in several locations also use drip
and sprinkler irrigation for irrigating non-rice crops. Several farmers also made
investments in farm ponds to recharge groundwater and also for supplemental
irrigation of the irrigated dry crops.
(v) Area under rice: Farmers in all the three states have cultivated rice only during
the kharif season and some in rabi season. During kharif season, the average
area under rice in the three states were respectively 0.58, 1.95 and 0.72 ha.
Similarly, in rabi season, the average rice area in the three states were 0.83,
1.95 and 0.64 ha. Other crops are grown during the summer season by Tamil
Nadu and Andhra Pradesh farmers with an average area of 0.27 and 0.31 ha
while Karnataka farmers grew other crops only during kharif season and the
average area is 0.73 ha.
9.4 Climate Change Events, Shocks ... 189

In order to have an in-depth understanding of farmers’ perception on the climate

lowed by the farmers, a detailed survey was conducted and the results are discussed
below.
9.4.1 Perception of Farmers on Climate Change
Farmers were asked to rank the 16 events as listed in the methodology section on
their perceptions of climate change. To find what the farmers perceive as the most
significant effects of climate change, Garrett’s ranking technique as described in
Chapter 4 was used and the results are given in Tables 9.4 and 9.5. Table 9.5 shows
that incidence of pests and diseases is the most important climate change event (Rank
1) perceived by rice farmers in southern region. It earned an average Garrett Score
of 69.2 (Table 9.5). As given in Table 9.4, it was given Rank 1 by 93 farmers. The
second and third important climate change events are respectively untimely rain and
irregular weather. A similar analysis was done for individual states in the region and
are given in Appendix (Tables 9.18, 9.19, 9.20, 9.21, 9.22, 9.23).
described in Chapter 4 was followed. The calculated value of Friedman’s F statistics
the events; thus, the farmers’ perceptions of climate change effects differ across the
sample chosen. The same conclusion was arrived at by applied Kendall’s coefficient
of concordance also. Similar analysis was done for individual states viz. Tamil Nadu,
Andhra Pradesh and Karnataka. The study results show that there is no consistency in
the rankings of the events by the farmers. Table 9.6 gives the rankings of the climate
change events state-wise and also region-wise.
The above table shows that the rankings of the climate change events are not
uniform across the states. So to check for agreement in rankings, the Kendall’s coef-
ficient of concordance was computed and tested for its significance. The computed
value of Kendall’s Coefficient, W was 0.4451 (with number of objects n = 16 and
number of judges m = 3) with a χ 2 value of 20.0294 (p-value = 0.1708). Since the
p-value is more than 0.10, we do not reject the null hypothesis and conclude that the
states are concordant in ranking the climate change events.
190
Table 9.4 Factors influencing farmers perception of climate change—Garrett ranking: southern region
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 38 8 5 15 23 32 17 42 2 2 0 20 32 31 22 24 313
Hailstorm 4 6 8 37 24 23 20 1 0 0 0 19 44 44 64 19 313
Flood 23 8 56 1 0 3 15 18 19 14 2 1 6 15 95 37 313
AD 17 43 18 6 10 13 18 9 50 36 20 19 7 13 19 15 313
IPD 93 46 25 52 4 10 35 29 8 1 2 2 5 0 1 0 313
IFM 3 4 2 3 31 8 9 6 13 20 43 89 55 17 3 7 313
UR 35 26 45 45 84 25 10 9 13 5 6 4 3 3 0 0 313
IW 9 37 25 54 26 66 22 26 8 6 15 1 12 6 0 0 313
TFH 35 41 24 13 22 19 16 22 13 27 12 13 34 20 2 0 313
TFL 10 0 1 15 4 13 13 11 24 25 33 21 12 40 34 57 313
SSDI 11 11 13 29 19 28 9 5 19 14 45 16 19 33 22 20 313
CCP 0 0 0 4 13 19 55 18 33 40 19 24 22 20 12 34 313
CLA 0 22 11 7 14 11 19 37 12 39 51 16 23 7 10 34 313
CFI 7 28 27 6 21 7 19 41 12 8 20 27 26 35 26 3 313
CF 8 33 24 25 17 35 16 27 56 16 36 7 6 5 1 1 313
DGW 20 0 29 1 1 1 20 12 31 60 9 34 7 24 2 62 313
Total 313 313 313 313 313 313 313 313 313 313 313 313 313 313 313 313 5008
Garrett score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
9 Climate Change Impacts on Rice Yield in Southern Region of India
Table 9.5 Rankings of factors representing climate change using Garrett ranking: southern region
Climate Rank Total score Average score Rank
change Event 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Total score
Drought 3268 608 350 975 1403 1856 935 2184 96 90 0 780 1120 930 528 336 15,459 49.4 7
Hailstorm 344 456 560 2405 1464 1334 1100 52 0 0 0 741 1540 1320 1536 266 13,118 41.9 15
Flood 1978 608 3920 65 0 174 825 936 912 630 84 39 210 450 2280 518 13,629 43.5 11
9.4 Climate Change Events, Shocks ...
AD 1462 3268 1260 390 610 754 990 468 2400 1620 840 741 245 390 456 210 16,104 51.5 6
IPD 7998 3496 1750 3380 244 580 1925 1508 384 45 84 78 175 0 24 0 21,671 69.2 1
IFM 258 304 140 195 1891 464 495 312 624 900 1806 3471 1925 510 72 98 13,465 43.0 12
UR 3010 1976 3150 2925 5124 1450 550 468 624 225 252 156 105 90 0 0 20,105 64.2 2
IW 774 2812 1750 3510 1586 3828 1210 1352 384 270 630 39 420 180 0 0 18,745 59.9 3
TFH 3010 3116 1680 845 1342 1102 880 1144 624 1215 504 507 1190 600 48 0 17,807 56.9 4
TFL 860 0 70 975 244 754 715 572 1152 1125 1386 819 420 1200 816 798 11,906 38.0 16
SSDI 946 836 910 1885 1159 1624 495 260 912 630 1890 624 665 990 528 280 14,634 46.8 9
CCP 0 0 0 260 793 1102 3025 936 1584 1800 798 936 770 600 288 476 13,368 42.7 14
CLA 0 1672 770 455 854 638 1045 1924 576 1755 2142 624 805 210 240 476 14,186 45.3 10
CFI 602 2128 1890 390 1281 406 1045 2132 576 360 840 1053 910 1050 624 42 15,329 49.0 8
CF 688 2508 1680 1625 1037 2030 880 1404 2688 720 1512 273 210 150 24 14 17,443 55.7 5
DGW 1720 0 2030 65 61 58 1100 624 1488 2700 378 1326 245 720 48 868 13,431 42.9 13
191
Table 9.6 Ranking of climate change events state-wise

Climate change event Tamil Nadu Andhra Karnataka Southern
Pradesh region
Drought 10 4 15 7
Hailstorm 9 10 13 15
Flood 14 2 16 11
AD 6 8 9 6
IPD 4 1 1 1
IFM 11 11 12 12
UR 2 3 5 2
IW 3 5 6 3
TFH 1 12 7 4
TFL 15 13 11 16
SSDI 12 15 3 9
CCP 16 7 8 14
CLA 7 16 10 10
CFI 13 9 2 8
CF 8 6 4 5
DGW 5 14 14 13
They were asked to rank seven effects of climatic shocks, as described in methodology
chapter. Farmers ranked declined in crop yield as the most important shock (Tables 9.7
and 9.8). Out of 313 farmers, 196 of them have ranked 1 to this effect. The second
and third ranks are given to the loss of income and loss of assets, respectively. This
is understandable because the decline in crop yield will reflect in income and asset
generation.
9.5.1 Data
of rice crop using historical data. A panel data set of yield and climate variables
pertaining to 30 years from 1981 to 2010 was used in the analysis. The data on
temperature corresponds to the growing season of rice variety in the respective states
and the time series data on rainfall was collected for the entire year to account for

Factors Rank Total
1 2 3 4 5 6 7
DCY 196 57 16 35 4 5 0 313
LA 35 18 110 38 14 33 65 313
LI 69 194 29 6 2 11 2 313
FIS 4 11 29 65 130 40 34 313
DL 7 14 28 57 90 91 26 313
DC 1 9 84 71 46 45 57 313
DH 1 10 17 41 27 88 129 313
Total 313 313 313 313 313 313 313 2191
Garrett score 79 66 57 50 43 34 21

1 2 3 4 5 6 7 score
Total score
DCY 15,484 3762 912 1750 172 170 0 22,250 71.1 1
LA 2765 1188 6270 1900 602 1122 1365 15,212 48.6 3
LI 5451 12,804 1653 300 86 374 42 20,710 66.2 2
FIS 316 726 1653 3250 5590 1360 714 13,609 43.5 5
DL 553 924 1596 2850 3870 3094 546 13,433 42.9 6
DC 79 594 4788 3550 1978 1530 1197 13,716 43.8 4
DH 79 660 969 2050 1161 2992 2709 10,620 33.9 7
rainfall falling directly on the crop and interseasonal water accumulation within a
year (Iski and Devadoss 2006; Kim and Pang 2009). The summary statistics of all
the variables are discussed below:
The yield data (Table 9.9; Fig. 9.4) shows an upward trend in all the three states over
the years. Trend equations fitted to yield data show that the coefficient of trend for
all the three states are statistically significant.
The maximum temperature had an overall average of 29.91 °C. Tamil Nadu had
the highest mean maximum temperature of 32.97 °C followed by Andhra Pradesh
(32.24 °C). Also, the variability in maximum temperature was least for Karnataka
(0.17 °C) (Table 9.10). Tamil Nadu’s maximum temperature is consistently higher
than those of Andhra Pradesh and Karnataka. Similarly, Karnataka’s temperature is
consistently lower than those of the other two states with an average difference of
approximately 8 °C in all the years (Fig. 9.5).
Table 9.9 Descriptive statistics of rice yield and trend equations (tons/ha)
Dev.
Tamil Nadu 30 2.848 1.855 3.579 0.467 Y = 2.518 + 0.16
0.0213** t
Andhra Pradesh 30 3.041 1.757 4.095 0.529 Y = 2.4162 + 0.45
0.0403*** t
Karnataka 30 2.345 1.772 3.868 0.422 Y = 1.8499 + 0.44
0.0319*** t
Overall 90 2.744 1.757 4.095 0.554
** and *** denotes significance at 5% and 1% level
Fig. 9.4 Trend in yield of rice in southern region of India

Tamil Nadu 30 32.97 32.20 34.3 0.47
Andhra Pradesh 30 32.24 31.39 32.96 0.37
Karnataka 30 24.52 24.22 24.87 0.17
Overall 90 29.91 29.27 30.71 3.86

The temperature data shows that Tamil Nadu had the highest average minimum
temperature (26.11 °C) compared to the other two states. Its average minimum tem-
perature is 25.23 °C (Table 9.11). It is consistently higher than those of the other two
states. Karnataka had the lowest minimum temperature with an average of 17.8 °C
in all the 30 years. On the average, there was a difference of approximately 5 °C in
its minimum temperature with those of Tamil Nadu and Andhra Pradesh (Fig. 9.6).
Fig. 9.5 Trend in maximum temperature in southern region of India

Tamil Nadu 30 25.23 24.59 26.11 0.43
Andhra Pradesh 30 22.95 21.94 23.72 0.37
Karnataka 30 17.79 17.48 18.13 0.17
Overall 90 21.99 17.48 26.11 3.15
Fig. 9.6 Trend in minimum temperature in southern region of India
(d) Rainfall (mm)

Karnataka had the highest mean annual rainfall of 2479 mm followed by Andhra
Pradesh (1149 mm). Its variability in rainfall was also highest with 593 mm. Com-
pared with other two states, Tamil Nadu consistently received lesser rainfall in most
of the years while rainfall at Karnataka was consistently higher than those of the
other two states in all the years (Fig. 9.7). Thus we find that Tamil Nadu had higher
temperature than Andhra Pradesh and Karnataka, whereas Karnataka had higher
rainfall compared to other two states (Table 9.12).
Fig. 9.7 Trend in annual rainfall in southern region of India

Tamil Nadu 30 1064 610 1509 246
Andhra Pradesh 30 1149 603 2185 341
Karnataka 30 2479 1226 3639 593
Overall 90 1564 603 3639 773
Table 9.13 Tests normality of yield data

Name of the test Test statistic p-value Decision
Doornik–Hansen test 2.713 0.258 Don’t reject Ho
Shapiro–Wilk W 0.980 0.190 Don’t reject Ho
Lilliefors test 0.055 0.720 Don’t reject Ho
Jarque–Bera test 2.334 0.311 Don’t reject Ho
The normality of yield data was tested using four tests. The results of the tests are
given in Table 9.13. For all the tests, the null hypothesis was the data that is normally
distributed. Hence based on the four tests, the null hypothesis can’t be rejected.
Since data on rice yield for the three states for 30 years formed a panel data set, it
was subjected to panel root testing before modelling. As stated in the methodology
chapter, Fisher’s panel root test was applied (Table 9.14).
Table 9.14 Panel root test results using fisher’s tests

statistic for statistic for
no trend trend
Yield 14.1114 0.0284** Reject H0 17.5767 0.0074*** Reject H0
temperature
Tempera-
ture
As discussed in the methodology chapter, to account for both mean and variance in
yield, Just–Pope production function (Just and Pope 1978 & 79) with yield as the
The quadratic model has the least value for AIC and so is selected. Table 9.15
whereas minimum temperature has a very strong significant negative effect. Similarly,
rainfall has a significant positive effect on yield. The square term corresponding to
maximum temperature and two interaction terms are significant. The significance of
these terms implies that nonlinear effects of climate variables are significant and so
important (Figs. 9.8, 9.9 and 9.10). The coefficient of the dummy variable for Andhra
Pradesh is positive and strongly significant implying that the mean yields between the
states differ statistically. The positive and significant coefficient for trend implies the
significance of technological progress in rice production such as improved irrigation
methods, use of high yielding varieties, better control of pests and diseases.
With respect to variance function, coefficients of trend and all the three climate
variables are significant. The coefficient of trend is positive implying that it is a
risk-increasing variable. This means that variability in yield increases over the years.
But we have already seen that the coefficient of a trend in the mean yield function
is positive and significant which implies that mean yield increases over time. So, we
conclude that over years increase in mean yield increases variability in yield also.

Constant (β0 ) −0.8181 2.3554 46.9815 22.9983
Trend (β1 ) 0.0342 0.0042 0.0388*** 0.0039
MaxT (β2 ) −0.0352 0.2669 10.8095*** 4.0932
MinT (β3 ) 0.2051 0.3101 −10.4427** 4.0121
RainT (β4 ) 0.0000 0.0001 0.0028* 0.0016
MaxT 2 (β5 ) – – −0.4626** 0.2122
MinT 2 (β6 ) – – −0.3063 0.2272
RainT 2 (β7 ) – – 0.0000 0.0000
MaxT MinT (β8 ) – – 0.7881** 0.4304
MaxT RainT (β9 ) – – −0.0002*** 0.0001
MinT RainT (β10 ) – – 0.0002 0.0001
TN (β11 ) −0.6040 0.8050 −1.0438 1.3089
AP (β12 ) −0.2727 0.7583 0.1012** 0.0130
Variance function
Constant 1.2537 4.9896 −0.0831 4.1591
Time 0.0263 0.0250 0.0306* 0.0233
MaxT 0.3751 0.2349 0.2967* 0.2310
MinT −0.6369 0.2210 −0.4959** 0.2345
RainT −0.0009 0.0006 −0.0007** 0.0005
Model statistics
LL −17.1199 −5.6029
AIC 58.2399 47.2058
BIC 88.2376 92.2023
These results coincide with the findings of Palanisami et al. (2011), Ranganathan
(2009), Isik and Devadoss (2006) and Anderson and Hazell (1987). Since the coef-
ficient of maximum temperature is positive, it is a risk-increasing variable for rice
yield. So, the increase in maximum temperature increases the yield variability. Coef-
ficients of minimum temperature and rainfall are negative and statistically significant.
So, these two are risk-decreasing variables.
perature is −2.894 implying that 1% increase in maximum temperature will induce
a decrease of 2.89% in mean yield. Similarly, the elasticity coefficient for minimum
mean yield by 3.35%. The elasticity coefficient for rainfall is −0.1420 which is very
Fig. 9.8 Nonlinear 3.5

relationship between yield Maximum Temperature
and maximum temperature

3

2.5
1.5
1
28 28.2 28.4 28.6 28.8 29 29.2 29.4 29.6 29.8 30

relationship between yield Minimum Temperature
and minimum temperature 3
2.5
1.5
0.5
0
22 22.5 23 23.5 24 24.5 25
Minimum Temperature (oC)
small and so 1% increase in rainfall will induce a decrease of 0.14% in mean yield.
When all the three variables are simultaneously increased by 1%, we can expect a
small increase of 0.31% in yield.
−5.3073 implying a reduction in variability in yield of 5.31 for 1% increase in
temperature and rainfall. All together, we can expect a decrease of 5.19% in yield
when all the three variables are increased by 1%. So, we conclude that all the three
variables put together are risk-decreasing variables.

relationship between yield 3.4
Rainfall
and rainfall
3.35

3.3
3.25
3.2
3.15
3.1
3.05
2.95
600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600
Rainfall(mm))
Table 9.16 Elasticities of Yield Function Climate variable Elasticity

Rainfall −0.1420
Rainfall −1.8431
∂y
∂(MaxT )
∂y
= β8 MaxT + 2 ∗ β6 MinT + β10 RainT = −β3
∂(MinT )
∂y
∂(RainT )
Solving these equations, the optimal levels are obtained as MaxT = 29.96 °C;
MinT = 20.86 °C and RainT = 1028 mm.
The estimated elasticities given in Table 9.17 can be used to estimate the impacts of
T max T min Rainfall mean Yield variability
yield
RCP45 2021−2050 3.66 6.57 1.42 11.2 −9.13
2071–2100 7.44 12.66 1.86 20.57 −18.08
RCP85 2021–2050 4.33 7.69 6.71 12.23 −20.28
2071–2100 12.92 23.25 42.33 34.40 −101.00
climate change scenarios, viz. RCP45 and RCP85 (as discussed in Chap. 3) were
considered. Table 9.17 presents the results.
The table shows that, as per RCP45 and RCP85 scenarios, the climate change
will have a positive effect on rice yield in the region. There will be an increase
in percentages of rice yield during the periods 2021–50 and 2071–2100 and the
percentage increase in the later period (2071–2100) will be more. The percentage
increase in mean yield varies from 11.2 to 12.23 during the first period and it varies
from 20.57 to 34.40% during the second period.
Yield variability is expected to reduce in both periods under the two scenarios.
The predicted reduction in variability is more with respect to the RCP85 scenario.
The first scenario predicts a decrease in variability of 9.1% during 2021–2050 while
the second scenario predicts a decrease of 20.3% for the same period. For the second
period, RCP45 predicts a reduction of 18% in yield variability, whereas RCP85
scenario predicts 101% reduction in variability. Thus, we find that climate change
will induce an increase in yield accompanied by a substantial reduction in yield
variability in southern region.
9.6 Summary
Incidence of pests and diseases is the most important climate change event (Rank 1)
perceived by rice farmers in southern region, followed by untimely rain and irregular
weather. Farmers ranked decline in crop yield as the most important shock, followed
by loss of income and loss of assets.
Historical rice yield data shows an upward trend in all the three states over years.
Tamil Nadu registered greater maximum and minimum temperature compared with
other two states throughout the study period. But Karnataka had the highest mean
annual rainfall followed by Andhra Pradesh and Tamil Nadu. The analysis of rela-
tionship between yield and climate variables shows that maximum temperature had
a significant positive effect on rice yield, whereas minimum temperature has very
strong significant negative effect. Similarly, rainfall has a significant positive effect
on yield. Also, there is strong evidence of nonlinear effects of climate variables on
yield.
The two climate change scenarios, viz. RCP45 and RCP85, predict an increase
in percentages of rice yield during the periods 2021–50 and 2071–2100 while yield
variability is expected to reduce in both periods under the two scenarios. The per-
centage increase in mean yield varies from 11.2 to 12.23 during the first period and
from 20.57 to 34.40 during the second period as per the two scenarios.
Appendix
Appendix
Table 9.18 Factors influencing farmers’ perception of climate change—Garrett ranking: Tamil Nadu
Factors Rank Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 2 6 0 0 2 24 15 24 1 1 0 11 7 19 0 0 112
Hailstorm 0 0 0 11 18 18 20 0 0 0 0 13 32 0 0 0 112
Flood 0 0 0 0 0 0 0 18 16 11 0 0 0 1 53 13 112
AD 17 2 1 6 2 5 12 5 22 25 10 3 0 2 0 0 112
IPD 14 3 5 12 2 8 34 18 7 1 2 2 4 0 0 0 112
IFM 0 3 2 0 14 7 7 3 3 3 20 33 10 7 0 0 112
UR 14 10 25 8 37 16 2 0 0 0 0 0 0 0 0 0 112
IW 5 25 14 38 6 0 0 0 1 0 11 0 12 0 0 0 112
TFH 34 38 18 8 7 5 1 0 0 0 1 0 0 0 0 0 112
TFL 0 0 0 12 0 11 3 4 6 7 0 2 0 0 13 54 112
SSDI 1 2 0 3 2 14 4 4 12 6 11 7 1 22 13 10 112
CCP 0 0 0 1 5 0 0 1 0 9 4 17 14 16 11 34 112
CLA 0 22 10 5 10 0 4 9 4 16 7 11 10 4 0 0 112
CFI 0 0 0 2 0 2 1 3 4 2 12 11 20 32 22 1 112
CF 7 1 8 6 6 2 6 16 16 11 28 0 2 3 0 0 112
DGW 18 0 29 0 1 0 3 7 20 20 6 2 0 6 0 0 112
Total 112 112 112 112 112 112 112 112 112 112 112 112 112 112 112 112 1792
Garrett score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
203
204
Table 9.19 Rankings of factors representing climate change using Garrett ranking: Tamil Nadu
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 score
Total score
Drought 172 456 0 0 122 1392 825 1248 48 45 0 429 245 570 0 0 5552 49.6 10
Hailstorm 0 0 0 715 1098 1044 1100 0 0 0 0 507 1120 0 0 0 5584 49.9 9
Flood 0 0 0 0 0 0 0 936 768 495 0 0 0 30 1272 182 3683 32.9 14
AD 1462 152 70 390 122 290 660 260 1056 1125 420 117 0 60 0 0 6184 55.2 6
IPD 1204 228 350 780 122 464 1870 936 336 45 84 78 140 0 0 0 6637 59.3 4
IFM 0 228 140 0 854 406 385 156 144 135 840 1287 350 210 0 0 5135 45.8 11
UR 1204 760 1750 520 2257 928 110 0 0 0 0 0 0 0 0 0 7529 67.2 2
IW 430 1900 980 2470 366 0 0 0 48 0 462 0 420 0 0 0 7076 63.2 3
TFH 2924 2888 1260 520 427 290 55 0 0 0 42 0 0 0 0 0 8406 75.1 1
TFL 0 0 0 780 0 638 165 208 288 315 0 78 0 0 312 756 3540 31.6 15
SSDI 86 152 0 195 122 812 220 208 576 270 462 273 35 660 312 140 4523 40.4 12
CCP 0 0 0 65 305 0 0 52 0 405 168 663 490 480 264 476 3368 30.1 16
CLA 0 1672 700 325 610 0 220 468 192 720 294 429 350 120 0 0 6100 54.5 7
CFI 0 0 0 130 0 116 55 156 192 90 504 429 700 960 528 14 3874 34.6 13
CF 602 76 560 390 366 116 330 832 768 495 1176 0 70 90 0 0 5871 52.4 8
DGW 1548 0 2030 0 61 0 165 364 960 900 252 78 0 180 0 0 6538 58.4 5
Appendix
Table 9.20 Factors influencing farmers perception of climate change—Garrett ranking: Andhra Pradesh
Factors Rank Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 33 0 4 10 21 8 2 18 1 1 0 1 1 0 0 1 101
Hailstorm 1 6 6 25 6 5 0 1 0 0 0 0 0 1 33 17 101
Flood 23 5 53 1 0 3 15 0 0 0 0 0 0 0 0 1 101
AD 0 32 1 0 2 3 3 0 17 2 0 1 1 6 18 15 101
IPD 21 25 16 35 2 0 0 0 0 0 0 0 1 0 1 0 101
IFM 0 1 0 0 17 1 1 0 0 8 2 43 18 5 3 2 101
UR 17 2 10 25 42 2 0 0 0 1 1 1 0 0 0 0 101
IW 2 10 5 2 7 63 5 3 3 0 0 0 0 1 0 0 101
TFH 1 0 1 1 1 1 9 14 4 14 0 2 32 19 2 0 101
TFL 0 0 1 0 1 2 5 5 6 3 23 3 3 32 17 0 101
SSDI 0 1 0 1 0 0 1 0 3 4 34 9 18 11 9 10 101
CCP 0 0 0 0 0 0 37 7 13 21 7 4 7 4 1 0 101
CLA 0 0 0 0 0 1 1 13 4 4 22 2 8 2 10 34 101
CFI 0 1 0 0 1 1 18 38 6 6 2 14 6 2 4 2 101
CF 1 18 4 0 1 10 3 2 38 4 8 4 4 2 1 1 101
DGW 2 0 0 1 0 1 1 0 6 33 2 17 2 16 2 18 101
Total 101 101 101 101 101 101 101 101 101 101 101 101 101 101 101 101 1616
Garrett score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
205
206
Table 9.21 Rankings of factors representing climate change using Garrett ranking: Andhra Pradesh
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 score
Total score
Drought 2838 0 280 650 1281 464 110 936 48 45 0 39 35 0 0 14 6740 66.7 4
Hailstorm 86 456 420 1625 366 290 0 52 0 0 0 0 0 30 792 238 4355 43.1 10
Flood 1978 380 3710 65 0 174 825 0 0 0 0 0 0 0 0 14 7146 70.8 2
AD 0 2432 70 0 122 174 165 0 816 90 0 39 35 180 432 210 4765 47.2 8
IPD 1806 1900 1120 2275 122 0 0 0 0 0 0 0 35 0 24 0 7282 72.1 1
IFM 0 76 0 0 1037 58 55 0 0 360 84 1677 630 150 72 28 4227 41.9 11
UR 1462 152 700 1625 2562 116 0 0 0 45 42 39 0 0 0 0 6743 66.8 3
IW 172 760 350 130 427 3654 275 156 144 0 0 0 0 30 0 0 6098 60.4 5
TFH 86 0 70 65 61 58 495 728 192 630 0 78 1120 570 48 0 4201 41.6 12
TFL 0 0 70 0 61 116 275 260 288 135 966 117 105 960 408 0 3761 37.2 13
SSDI 0 76 0 65 0 0 55 0 144 180 1428 351 630 330 216 140 3615 35.8 15
CCP 0 0 0 0 0 0 2035 364 624 945 294 156 245 120 24 0 4807 47.6 7
CLA 0 0 0 0 0 58 55 676 192 180 924 78 280 60 240 476 3219 31.9 16
CFI 0 76 0 0 61 58 990 1976 288 270 84 546 210 60 96 28 4743 47.0 9
CF 86 1368 280 0 61 580 165 104 1824 180 336 156 140 60 24 14 5378 53.2 6
DGW 172 0 0 65 0 58 55 0 288 1485 84 663 70 480 48 252 3720 36.8 14
Appendix
Table 9.22 Factors influencing farmers perception of climate change—Garrett ranking: Karnataka
Factors Rank Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drought 3 2 1 5 0 0 0 0 0 0 0 8 24 12 22 23 100
Hailstorm 3 0 2 1 0 0 0 0 0 0 0 6 12 43 31 2 100
Flood 0 3 3 0 0 0 0 0 3 3 2 1 6 14 42 23 100
AD 0 9 16 0 6 5 3 4 11 9 10 15 6 5 1 0 100
IPD 58 18 4 5 0 2 1 11 1 0 0 0 0 0 0 0 100
IFM 3 0 0 3 0 0 1 3 10 9 21 13 27 5 0 5 100
UR 4 14 10 12 5 7 8 9 13 4 5 3 3 3 0 0 100
IW 2 2 6 14 13 3 17 23 4 6 4 1 0 5 0 0 100
TFH 0 3 5 4 14 13 6 8 9 13 11 11 2 1 0 0 100
TFL 10 0 0 3 3 0 5 2 12 15 10 16 9 8 4 3 100
SSDI 10 8 13 25 17 14 4 1 4 4 0 0 0 0 0 0 100
CCP 0 0 0 3 8 19 18 10 20 10 8 3 1 0 0 0 100
CLA 0 0 1 2 4 10 14 15 4 19 22 3 5 1 0 0 100
CFI 7 27 27 4 20 4 0 0 2 0 6 2 0 1 0 0 100
CF 0 14 12 19 10 23 7 9 2 1 0 3 0 0 0 0 100
DGW 0 0 0 0 0 0 16 5 5 7 1 15 5 2 0 44 100
Total 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 1600
Garret score 86 76 70 65 61 58 55 52 48 45 42 39 35 30 24 14
207
208
Table 9.23 Rankings of factors representing climate change using Garrett ranking: Karnataka
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 score
Total score
Drought 2838 0 280 650 1281 464 110 936 48 45 0 39 35 0 0 14 6740 66.7 4
Hailstorm 86 456 420 1625 366 290 0 52 0 0 0 0 0 30 792 238 4355 43.1 10
Flood 1978 380 3710 65 0 174 825 0 0 0 0 0 0 0 0 14 7146 70.8 2
AD 0 2432 70 0 122 174 165 0 816 90 0 39 35 180 432 210 4765 47.2 8
IPD 1806 1900 1120 2275 122 0 0 0 0 0 0 0 35 0 24 0 7282 72.1 1
IFM 0 76 0 0 1037 58 55 0 0 360 84 1677 630 150 72 28 4227 41.9 11
UR 1462 152 700 1625 2562 116 0 0 0 45 42 39 0 0 0 0 6743 66.8 3
IW 172 760 350 130 427 3654 275 156 144 0 0 0 0 30 0 0 6098 60.4 5
TFH 86 0 70 65 61 58 495 728 192 630 0 78 1120 570 48 0 4201 41.6 12
TFL 0 0 70 0 61 116 275 260 288 135 966 117 105 960 408 0 3761 37.2 13
SSDI 0 76 0 65 0 0 55 0 144 180 1428 351 630 330 216 140 3615 35.8 15
CCP 0 0 0 0 0 0 2035 364 624 945 294 156 245 120 24 0 4807 47.6 7
CLA 0 0 0 0 0 58 55 676 192 180 924 78 280 60 240 476 3219 31.9 16
CFI 0 76 0 0 61 58 990 1976 288 270 84 546 210 60 96 28 4743 47.0 9
CF 86 1368 280 0 61 580 165 104 1824 180 336 156 140 60 24 14 5378 53.2 6
DGW 172 0 0 65 0 58 55 0 288 1485 84 663 70 480 48 252 3720 36.8 14
Chapter 10
Economics of Adoption of Rice Production
and Management Technologies

The major factor that may help to mitigate the climate change impacts in the agricul-
ture sector will be introduction of technologies or practices that will help in providing
the assured income. List of such technologies or practices adopted by the farmers
in rice crop cultivation was obtained from the field surveys from different states and
analysed. The key technologies or practices popular among the farmers and useful
for upscaling in the states in different regions are discussed in the following sections.
Only the states with proven use of the technologies or practices have been included
in the detailed analysis.
10.1 Rice Technology Adoption and Profitability
10.1.1 Northern Region
There are many technologies available to farmers for improving their rice cultivation
and mitigate the impacts due to climate change. The technologies such as direct
sown rice, machine transplanting, system of rice intensification, alternate wetting and
drying of rice are some of the practices which are followed up and found profitable
(Palanisami et al. 2014). Since details on the spread up of these practices are not
available for all the regions, a detailed analysis was done by asking the farmers about
the availability, access and adoption of those technologies and also the constraints
in adopting them.
The authors would like to thank Dr. K.R. Karunakaran, Prof of Agrl. Economics, Tamilnadu
Agrl. University for providing substantial inputs in preparing this chapter without which, the
chapter would not have taken the shape that it has now.

210 10 Economics of Adoption of Rice Production and Management Technologies
120
100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
100
89 89
80
62 62
60 53
44 44 43 43
38 38
40
31
20 14 14
2 2 3
0
0
Access
Access
Access
Access
Access
Access
Access
Access
Access
Access
Access
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
DSR MSRI MT IMP AMOM SITG DI AWD IMD IPM MH
Fig. 10.1 Popularity of rice technologies among farmers in northern region (%). Note DSR =
direct sown rice; MSRI = modified system of rice intensification; MT = machine transplanting;
IMP = improved management practices; AMOM = application of more organic manure; SITG =
supplemental irrigation through groundwater; DI = drip irrigation; AWD = alternate wetting and
drying; IMD = irrigation at minimum depth; IPM = integrated pest management; MH = machine
harvesting
The access, availability and adoption level of 11 rice technologies observed in

the region were analysed and presented in Fig. 10.1. Among the 11 rice technolo-
gies, improved management practices (IMP), application of more organic manure
(AMOM), machine harvesting (MH), integrated pest management (IPM) and sup-
plemental irrigation through groundwater (SITG) are the most popular technologies
adopted by all the sample farmers. The least adopted technologies are machine trans-
planting (MT), direct sown rice (DSR), drip irrigation (DI) and irrigation at minimum
depth (IMD).
10.1.1.1 Costs and Benefits of Technology Adoption—Punjab
In order to get an idea on the economics of technology adoption at state level, the
farm-level cost of cultivation (COC), cost of technology, yield, gross return and net
return were computed and are presented in Table 10.1
The table shows that an increase in net income over the conventional method (CM)
was higher with alternate wetting and drying (AWD) (Rs. 19,142/ha) followed by
MSRI (Rs. 17,882/ha) and improved management practices (IMP) (Rs. 15,342/ha).
The table also highlights that in terms of cost of production of rice per quintal (qtl), it
is invariably less under all the technologies/practices compared to the conventional
method, thus providing incentives to the farmers for technology adoption. Nonethe-
less, the cost of technology is increased from Rs. 2000 to 7000/ha by adopting the
10.1 Rice Technology Adoption and Profitability 211
Table 10.1 Economics of rice technologies—Punjab

Technology Cost of Cost of Yield Gross Net Increase Cost of
cultiva- Tech. (t/ha) Return Return in NI produc-
tion (Rs./ha) (Rs./ha) (NI) over CM tion
(COC) (Rs./ha) (Rs./ha) (COP)
(Rs./ha) (Rs./qtl)
CM 35,392 4.9 62,750 27,358 722
DSR 33,200 −2192 4.8 62,300 29,100 1742 692
MSRI 40,500 5108 6.6 85,740 45,240 17,882 614
IMP 38,600 3208 6.2 81,300 42,700 15,342 623
AMOM 41,400 6008 6.1 79,650 38,250 10,892 679
SITG 42,640 7248 6.4 83,100 40,460 13,102 666
AWD 37,400 2008 6.5 83,900 46,500 19,142 575
Note CM conventional method; DSR direct sown rice; MSRI modified system of rice intensification;
MT machine transplanting; IMP improved management practices; AMOM application of more
organic manure; SITG supplemental irrigation through groundwater; AWD alternate wetting and
drying
% adopƟon & addl. net return over CM
120 7.0
6.6 6.4 6.5
6.2 6.1 6.0
100
5.0
Rice Yield t/ha

80 4.8
4.0
60
3.0
40
65 70 2.0
56
20 40 48
1.0
6 2 63 100 100 100 12
0 0.0
DSR MSRI IMP AMOM SITG AWD
% of addl net return over CM % adopt. Yield
Fig. 10.2 Technology-wise % adoption, % addl.net income over CM and rice yield in Punjab
technologies. But, the cost of technology for adopting DSR is reduced (Rs. 2192/ha)
due to direct seeding by withdrawing labour for transplantation. Figure 10.2 illus-
trates that in Punjab, adoption percentage is more for improved management practices
(IMP), application of more organic manure (AMOM) and supplemental irrigation
through groundwater (SITG) with 100%. Among the technologies, though all the
farmers have adopted these technologies, the highest productivity was found to be
in MSRI (6.6 t/ha) followed by AWD (6.5 t/ha), SITG (6.4 t/ha), IMP (6.2 t/ha)
and AMOM (6.1 t/ha). However, AWD realized 6.5 t/ha yield gaining 70% higher
additional returns over CM which is presently followed by 12% of the rice farmers.
10.1.1.2 Costs and Benefits of Technology Adoption—Haryana
Table 10.2 gives summary of the cost and benefits for different practices followed by
farmers in Haryana. In Haryana also, AWD and MSRI are the two most profitable
technologies among seven technologies followed by the rice farmers.
In the case of Haryana, an increase in net income over the CM was higher with
AWD (Rs. 13,661/ha) followed by MSRI (Rs. 8393/ha). All other technologies are
yielding only marginal increase over CM, and this might be due to comparatively
lesser yield than in Punjab. In the case of cost of production of rice per quintal, in most
cases, it is marginally higher than CM, thus implying that some of the technologies
are not financially attractive to adopt and this might be the reason for their low level
of adoption (see Fig. 10.3).
The IMP, AMOM and SITG practices have 100% adoption, but the DI and IMD
are adopted by 2–6% only due to the higher cost. The yield benefits are also less
Table 10.2 Economics of rice technologies—Haryana

cultiva- Tech. (t/ha) return return in NI produc-
(Rs./ha) (Rs./qtl)
CM 31,681 4.8 59,520 27,839 660
MSRI 38,168 6487 6 74,400 36,232 8393 636
IMP 40,791 9110 5.5 68,660 27,869 30 742
AMOM 36,705 5024 5.4 66,960 30,255 2416 680
SITG 39,850 8169 5.7 70,680 30,830 2991 699
AWD 34,140 2459 6.1 75,640 41,500 13,661 560
DI 43,540 11,859 5.9 73,340 29,800 1961 738
IMD 32,980 1299 4.9 60,960 27,980 141 673
% adopƟon & addl. net return over
120 7.0
100 6.0 6.1 5.9 6.0

5.5 5.7
5.4
Rice Yield t/ha
4.9 5.0
80
4.0
60 49
CM
100 100 100 3.0

40
2.0
20 43 11 49
9 7 1.0
30 0 2 1 6
0 0.0
MSRI IMP AMOM SITG AWD DI IMD
Fig. 10.3 Technology-wise % adoption, % addl.net return over CM and rice yield in Haryana
Table 10.3 Economics of rice technologies in UP

(Rs./ha) (Rs./qtl)
CM 29,138 4.8 60,240 31,102 607
DSR 31,200 2062 5 63,250 32,050 948 624
IMP 36,780 7642 5.4 68,310 31,530 428 681
AMOM 32,705 3567 5.1 64,515 31,810 708 641
SITG 31,850 2712 5.3 67,045 35,195 4093 601
MSRI 33,450 4312 5.6 68,750 35,300 4198 567
MH 32,980 3842 5.1 64,120 31,140 38 673
for IMD. Though DI has higher yield (5.9 t/ha) and water use efficiency, the cost of
production (Rs. 738/qtl) is comparatively high, resulting in low adoption level.
10.1.1.3 Costs and Benefits of Technology Adoption—Uttar Pradesh
In case of Uttar Pradesh (UP), increase in net income over the conventional method
(CM) was higher with MSRI (Rs. 4198/ha) followed by SITG (Rs. 4093/ha). As in
Haryana, all other technologies are yielding only marginal increase over the CM due
to comparatively lesser yield. The average yield increase under the technologies is
only about 9% over the CM compared to about 14% increase in average cost of culti-
vation. In most cases, the cost of production per quintal is marginally higher than the
CM, thus warranting efforts to minimize the cost of cultivation with more emphasis
on cost-effective adoption mechanisms including field level training. Interestingly,
the cost of production with MSRI is less than the CM due to higher yield (5.6 t/ha)
(Table 10.3). However, the adoption rate under MSRI and DSR is less, which is about
2% (Fig. 10.4). The cost of technology in DSR is also higher in the state due to the
increased cost of herbicides and fertilizers.
10.1.2 Eastern Region
Among the ten rice technologies identified in the region, MH, IMP, SITG and IPM
are the most popular technologies. Though the availability of these technologies
to the farmers is reported to be high, viz. 45, 81, 48 and 23%, the access to these
technologies by the farmers was 40, 81, 32 and 40%, respectively (Fig. 10.5). In the
case of adoption, it was only 34, 64, 26 and 15%, whereas MSRI and MT were adopted

120 7.0
100 5.9 6.0

5.4 5.3
5.0 5.1 4.9 5.0
80
Rice Yield t/ha

4.0
60
100 3.0
40
2.0
61
20 13 13
35 1.0
3 2 1 2 18 2 0
0 0.0
DSR IMP AMOM SITG MSRI MH
Fig. 10.4 Technology-wise % adoption, % addl.net Income over CM and rice yield in Uttar Pradesh
90
81 81
80
70 64
60 54 54
50 45 48 46 45
42 40
40 34
32 30
30 27 26
25 23 23
20 17
14 15
10
10 5 4 6
2 2 2 1
0
Access
Access
Access
Access
Access
Access
Access
Access
Access
Access
Available
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Available
Available
Available
Available
Available
Available
Available
Available
Available
DSR SRI MSRI MT IMP AMOM SITG AWD IPM MH
Fig. 10.5 Popularity of rice technologies among farmers in eastern region (%)
by 1.5 and 2% farmers, respectively, indicating the need for initiating upscaling
measures.
10.1.2.1 Costs and Benefits of Technology Adoption—Odisha
Table 10.4 shows that in Odisha, the average cost of cultivation varies from Rs.
30,650/ha under CM to Rs. 34,250/ha with DSR. The net returns (NI) are higher
for IMP and MSRI practices. Cost of production in all the adopted technologies is
slightly lower than the CM. Among these technologies, the cost of production under
MSRI is comparatively lower (Rs. 551/qtl) followed by IMP (Rs. 584/qtl) and MH
and DSR both at Rs. 591/qtl.
Table 10.4 Economics of rice technologies in Odisha

(Rs./ha) (Rs./qtl)
CM 30,650 4.8 52,800 22,150 634
IMP 33,311 2661 5.7 65,550 32,239 10,089 584
DSR 34,250 3600 5.8 63,800 29,550 7400 591
MSRI 32,490 1840 5.9 64,900 32,410 10,260 551
MT 33,540 2890 5.5 60,500 26,960 4810 610
MH 33,100 2450 5.6 61,600 28,500 6350 591
80 6.0
70 5.9 5.9
5.9
60
5.8 5.8
Rice Yield t/ha

50
5.8
40 5.7 5.7
64 68 5.7
30
5.6 5.6
20
5.6
10 19
31 23 32 2 15 45.5 20 5.5
0 5.5
IMP DSR MSRI MT MH
Fig. 10.6 Technology-wise % adoption, % addl.net return over CM and rice yield in Odisha
MH and IMP are adopted by more than half of the farmers. The DSR, MSRI and
MT, which are very important to address the climate change impacts and improve
water use efficiency, are adopted by only 19, 2, and 4%, respectively (Fig. 10.6). The
yields are also higher in the MSRI (5.9 t/ha) and DSR (5.8 t/ha) when compared to
the other adoption practices.
10.1.2.2 Costs and Benefits of Technology Adoption—West Bengal
The details of the cost and return of technology adoption in West Bengal are presented
in Table 10.5. The average cost of cultivation varies from Rs. 27,258/ha under CM to
Rs. 33,800/ha with AMOM. The rice yield is uniformly higher with the technologies,
showing the impact of technology adoption in boosting the rice yield. Increase in net
income over the CM was higher with SRI, MSRI, AWD and IPM. In terms of cost
of production, most of the technologies yielded varying results from CM implying
Table 10.5 Economics of rice technologies—West Bengal

(Rs./ha) (Rs./qtl)
CM 27,258 4.9 54,880 27,622 556
SRI 33,420 6162 6.2 70,200 36,780 9158 539
IMP 31,456 4198 5.7 64,500 33,044 5422 552
MSRI 32,490 5232 6.1 68,650 36,160 8538 533
AMOM 33,800 6542 5.4 61,620 27,820 198 626
SITG 33,438 6180 5.8 64,580 31,142 3520 577
AWD 32,450 5192 6.0 67,200 34,750 7128 541
IPM 31,840 4582 5.9 66,080 34,240 6618 540
% adopƟon & addl. net return
70 6.4
60 6.2 6.2
Rice Yield t/ha

50 6.1
6.0 6.0
over CM
40 5.9
5.8 5.8
30 64.00
5.7 52.00 5.6
20
5.4 11.00 29.00 5.4
10
28 3.00 17 26 2.00 1 2.00 11 22 21
0 5.2
SRI IMP MSRI AMOM SITG AWD IPM
Fig. 10.7 Technology-wise % adoption, % addl.net return over CM and rice yield in West Bengal
that technology-induced cost components need to be examined in detail so that cost

minimization efforts can be targeted.
IMP and SITG are adopted more than 50% in West Bengal. The SRI, MSRI and
AWD practices which are high yielding (6.2, 6.1 and 6.0 t/ha, respectively) in the
state are less adopted possibly due to lack of proper awareness (Fig. 10.7).
10.1.3 Western Region
The available, access and adoption level of different rice technologies in Gujarat are
presented in Fig. 10.8. Applying more organic manure (AMOM), MSRI and SITG
are the three prominent technologies widely available to and accessed by the Gujarat
rice farmers. However, the actual adoption rates were 48, 35 and 28%, respectively.
120
100 100 100
100
80
80
60 50 50 50 48
40 35
28
20
20 10
2 5 2
0
Access
Access
Available
Adoption
Access
Available
Adoption
Adoption
Access
Available
Available
Adoption
Access
Adoption
Available
SRI MSRI MT AMOM SITG
Fig. 10.8 Popularity of rice technologies among farmers in western (%)
Table 10.6 Economics of rice technologies in Gujarat

(Rs./ha) (Rs./qtl)
CM 30,075 5.1 62,317 32,242 590
SRI 37,680 7605 6.4 77,658 39,978 7736 589
MSRI 36,530 6455 6.2 75,890 39,360 7118 589
MT 30,870 795 5.7 71,475 40,605 8363 542
AMOM 34,890 4815 5.6 69,364 34,474 2232 623
SITG 37,450 7375 6.1 75,640 38,190 5948 614
MT is another important labour-saving rice technology which is adopted by only 2%

of the rice farmers in Gujarat.
The costs and benefits of technology adoption are presented in Table 10.6. The
average cost of cultivation varies from Rs. 30,075/ha under CM to Rs. 37,680/ha
with SRI. In the case of increased net income over the conventional methods, all
the adopted technologies resulted in higher net income though the magnitude varies
across the technologies. Among the technologies, the cost of production under MT
is comparatively lower (Rs. 542/qtl) than other technologies. However, the cost of
production has not much varied among the technologies which are almost similar to
the cost under CM.
90
83
80 78
73
70
61
60 55
50 50 50 50
50 45
44
40
40 37
31
30
23
20 17.5
12 12
10 7.5
3
1
0
Access
Access
Access
Access
Access
Access
Access
Available
Available
Available
Available
Available
Available
Available
AdopƟon
AdopƟon
AdopƟon
AdopƟon
AdopƟon
AdopƟon
AdopƟon
DSR SRI IMP AMOM SITG DI MH
Fig. 10.9 Popularity of rice technologies among farmers in central region (%)
10.1.4 Central Region
The types of technologies available, access and adoption level in the central region
were studied in detail and are presented in Fig. 10.9. Among the identified rice
technologies, seven technologies were available and reported. Among them, IMP,
AMOM and SITG are the three popular technologies whose adoption rates were 83,
31 and 17.5%, respectively. It is important to examine why some of the technologies
popular in other regions such as SRI, MSRI are not being followed in the region.
Cross-learning of the technology adoption process will help identify such adoption
constraints.
10.1.4.1 Costs and Benefits of Technology Adoption—Madhya Pradesh
The cost and additional return realized under four major rice technologies in MP
were analysed, and the results are presented in Table 10.7.
The table shows that the average cost of cultivation varies from Rs. 31,803/ha
under CM to Rs. 37,890/ha with SRI (i.e. 19% higher). In terms of increased net
income, except DSR, all other technologies showed a similar increase. The cost of
technology is negative for DSR, implying that compared to conventional (transplant-
ing) method, DSR resulted in lesser cost, and hence, negative and in practice we can
say DSR involves no additional cost towards technology and this is also reflected
in the reduction in the total cost of cultivation compared to conventional method
Table 10.7 Economics of rice technologies in Madhya Pradesh

(Rs./ha) (Rs./qtl)
CM 31,803 4.8 54,000 22,197 663
DSR 29,100 −2703 4.9 54,860 25,760 3563 594
IMP 32,780 977 5.4 62,300 29,520 7323 607
SITG 35,650 3847 5.9 65,480 29,830 7633 604
SRI 37,890 6087 6.2 69,560 31,670 9473 611
70 7.0
60 5.9 6.2 6.0
Rice Yield t/ha

5.4 5.0
50 4.9
over CM
40 4.0
30 65 3.0
20 2.0
27 1.0
10 15
16 4 33 34 43
0 0.0
DSR IMP SITG SRI
Fig. 10.10 Technology-wise % adoption, % addl.net return over CM and rice yield in Madhya
Pradesh
(Kakumanu et al. 2018). The cost of production under all the adopted technologies
is lower than the conventional method, and among them, DSR is seemingly much
lower (Rs. 594/qtl) due to less labour use and overcoming transplanting operations.
However, IMP adoption rate is higher (65%) followed by SITG (27%) and SRI
(15%) with higher rice productivity in SRI (6.2 t/h) and SITG (5.9 t/ha) (Fig. 10.10).
This implies the need for initiating more awareness programs and field demonstra-
tions to upscale these technologies as their cost of production is also less compared
to CM.
10.1.5 Southern Region
Among the rice technologies, machine harvesting (MH), improved management

practices (IMP), supplemental irrigation (SIGT) and integrated pest management
(IPM) are more popular. The percentage adoption level of these technologies varied
from 31 to 71 (Fig. 10.11). It is interesting to observe that though these technologies
120
99 96
100 93 94
86 85 84
79 82
78 76
80 71
67
57 55
60 51 48 48
40 31
29
20 18
20 13
4
0
Access
Access
Access
Access
Access
Access
Access
Access
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Adoption
Available
Available
Available
Available
Available
Available
Available
Available
DSR MSRI MT IMP SITG AWD IPM MH
Fig. 10.11 Popularity of rice technologies among farmers in southern region (%)
are known for their higher yield, their lower level of adoption indicates the need for
examining the constraints in their adoption.
10.1.5.1 Costs and Benefits of Technology Adoption—Andhra Pradesh
In the case of economics of technology adoption, the average cost of cultivation

varies from Rs. 37,830/ha under CM to Rs. 45,983/ha with SITG in Andhra Pradesh
(Table 10.8). In all the cases, the cost of cultivation has been higher by 3% (for DSR)
to 21% (for SITG) indicating the importance of the cost in technology introduction
and adoption. In case of DSR, a recent study shows that there is a decrease in cost
of cultivation compared to CM (Kakumanu et al. 2018). In addition, the study also
explained the benefits of DSR and the role in mitigating climate change impacts.
In the case of increased net income, MSRI, AWD, IMP and IPM have compar-
atively higher income than SSD and MH. The cost of production under AWD is
comparatively lower (Rs. 653/qtl) than SSD (Rs. 826/qtl). However, as observed
in other regions, the cost of production with technology and without technology is
not much significant, and this might due to different levels of increase in cost of
cultivation and net income.
AWD and MT are also practised in the state but to a very less extent (Fig. 10.12).
It is mainly due to the non-availability of machinery for timely transplantation and
lack of on- and off-systems in canal operation. IMP and IPM are also ighly practised
in the state as there is a need to control increasing pest and diseases to the changing
climate.
Table 10.8 Economics of rice technologies—Andhra Pradesh

(Rs./ha) (Rs./qtl)
CM 37,830 4.8 59,040 21,210 788
MSRI 43,284 5454 6.6 75,640 32,356 11,146 656
AWD 39,200 1370 6.0 72,860 33,660 12,450 653
SITG 45,983 8153 6.2 73,450 27,467 6257 742
DSR 39,130 1300 5.1 66,400 27,270 6060 767
SSD 43,765 5935 5.3 65,190 21,425 215 826
MT 39,745 1915 5.4 66,420 26,675 5465 736
IMP 39,560 1730 5.9 72,570 33,010 11,800 671
IPM 41,240 3410 6.1 75,030 33,790 12,580 676
MH 39,870 2040 5.1 62,730 22,860 1650 782
80 8.0
Rice Yield t/ha

6.6 6.2
60 6.0 5.9 6.1 6.0
over CM
5.3 5.4
585.1 54 54 575.1
40 4.0
20 2.0
53 1 59 5 30 2 29 11 26 3 56 59 8
0 0.0
MSR I AWD SITG DSR SSD MT IMP IPM MH
Fig. 10.12 Technology-wise % adoption, % addl.net return over CM and rice yield in Andhra
Pradesh
10.1.5.2 Costs and Benefits of Technology Adoption—Tamil Nadu
In case of Tamil Nadu, the average cost of cultivation varies from Rs. 29,740/ha under
CM to Rs. 37,680/ha with SITG. One interesting aspect is that yield increase is signif-
icantly higher with technology adoption where the average yield under technologies
as a whole is about 6.3 t/ha compared to 5.1/ha with conventional method. Increase
in net income is higher with IMP (Rs. 12,015/ha) which is about 38% higher than
the CM, whereas the cost of production under AWD is lower at Rs. 526/qtl followed
by IMP and MH.
% adopƟon & addl. net return 120 6.6

6.5 6.5
100
6.5
Rice Yield t/ha

80
6.4 6.4
over CM
60 6.4
100 100 100 6.3
40 6.3
6.3
20 45
6.2 6.2 6.2
24 31 23 15 38 31
0 6.2
SRI AWD SITG IMP MH
Fig. 10.13 Technology-wise % adoption, % addl.net return over CM and rice yield in Tamil Nadu
Table 10.9 Economics of rice technologies—Tamil Nadu

(Rs./ha) (Rs./qtl)
CM 29,740 5.1 61,245 31,505 583
SRI 36,580 6840 6.4 75,520 38,940 7435 572
AWD 33,120 3380 6.3 74,320 41,200 9695 526
SITG 37,680 7940 6.2 73,755 36,075 4570 608
IMP 35,130 5390 6.5 78,650 43,520 12,015 540
MH 33,765 4025 6.2 75,020 41,255 9750 545
The SITG, IMP and MH are fully adopted (100%) by the sample farmers
(Fig. 10.13). Yields also have comparative advantage in technology adoption
(1.1–1.3 t/ha). Cost of production is higher for SITG when compared to CM. All
other four practices are generally showing comparatively lower cost than the CM
(Table 10.9).
10.1.5.3 Costs and Benefits of Technology Adoption—Karnataka
In the case of Karnataka, the average cost of cultivation varies from Rs. 34,830/ha
under conventional method to Rs. 45,983/ha with SITG (Table 10.10). Regarding
the technology cost, the supplemental irrigation costs are more due to possible water
scarcity, but the yield is higher (6.2 t/ha). In the case of increased net income over
conventional method, only SRI resulted in higher income (Rs. 10,396/ha) and the
Table 10.10 Economics of rice technologies—Karnataka

(Rs./ha) (Rs./qtl)
CM 34,830 5.3 76,850 42,020 657
DSR 35,130 300 5.4 78,300 43,170 1150 651
SRI 43,284 8454 6.6 95,700 52,416 10,396 656
IMP 36,500 7950 5.4 78,600 42,100 80 676
SITG 45,983 11,153 6.2 89,900 43,917 1897 742
AWD 37,130 4300 5.6 81,200 44,070 2050 663
IPM 41,050 6220 5.8 84,100 43,050 1030 708
MH 35,767 8935 5.5 79,750 43,983 1963 650
MT 36,765 1935 5.6 81,200 44,435 2415 657
70 7.0
6.6
60 6.2 6.0
5.6 5.8
5.4 5.4 5.5
Rice Yield t/ha

50 5.0
40 4.0
30 63 3.0
56
47
20 38 2.0
35
10 1.0
3 2 25 7 0 5 5 2 5
0 0.0
DSR SRI IMP SITG AWD IPM MH
Fig. 10.14 Technology-wise % adoption, % addl.net return over CM and rice yield in Karnataka
least was observed under IMP and this might be due to lower rice yield. Higher cost
of cultivation and marginal yield increase with technologies resulted in only marginal
increase in net income over the conventional method, thus warranting strategies to
increase the yield (Fig. 10.14).
Adoption of SITG, MH and IMP practices are high in Karnataka state. The per-
centage of additional net returns over CM is more for SRI followed by AWD, SITG
and MH accounting for 5% each.
224
Table 10.11 Increase in cost of cultivation with technologies over the conventional method (%)
Technology Northern Western Eastern Central Southern Average
PU HA UP GU OD WB MP AP TN KA
AMOM 16.98 15.86 12.24 16.01 24 17.02
AWD 5.67 7.76 19.05 3.62 6.6 11.37 9.01
DI 37.43 37.43
DSR −6.19 7.08 11.75 −8.5 3.44 0.86 1.41
IMD 4.1 4.10
IMP 9.06 28.76 26.23 8.68 15.4 3.07 4.57 4.79 18.12 13.19
IPM 16.81 9.01 17.86 14.56
MH 13.19 7.99 5.39 2.69 13.53 8.56
MSRI 14.43 20.48 14.8 21.46 6 19.19 14.42 15.83
MT 2.64 9.43 5.06 5.56 5.67
SITG 20.48 25.79 9.31 24.52 22.67 12.1 21.55 32.02 26.7 21.68
SRI 25.29 22.61 19.14 24.27 23 22.86
SSD 15.68 15.68
Note DSR direct sown rice; MSRI modified system of rice intensification; MT machine transplanting; IMP improved management practices; AMOM application
of more organic manure; SITG supplemental irrigation through groundwater; DI drip irrigation; AWD alternate wetting and drying; IMD irrigation at minimum
depth; IPM integrated pest management; MH machine harvesting SSD subsurface drainage
10 Economics of Adoption of Rice Production and Management Technologies
10.2 Comparison of Cost and Yield ... 225
10.2 Comparison of Cost and Yield Attributes of Rice Due

to Technology Adoption
It is interesting to note that in most cases discussed earlier, the cost of production per
quintal (qtl) was almost same with and without technology adoption. This is possibly
due to higher cost of cultivation without corresponding increase in yield. Otherwise,
both cost and yield did not change much due to technology adoption. In order to get
an idea on this, a comparison of the cost of cultivation under different technologies
across regions has been made (Table 10.11). It is seen that invariably in all the
regions, the cost of cultivation of rice with the technology adoption has increased
from negligible percentage with DSR to as high as 37% with DI. Average increase
in cost shows 37% for DI followed by 23% with SRI, 22% with SITG, 17% with
AMOM, etc. The high cost under DI might be due to its initial high cost of drip
irrigation system. However, this technology has been in its initial stage of adoption
under government/research support programs. The cost of cultivation for DSR in
Punjab and MP states is lesser than CM due to saving in labour cost.
In the case of yield increase, except AMOM, DI, SSD, other technologies exhibited
increased yields than the cost increment (Table 10.12). In the case of SITG, there
is no significant difference. Since the difference in the cost and yield increase is
varying across regions and technologies as seen in Fig. 10.15, technology-focused
efforts are needed to increase the yield and reduce the cost.
40.00
35.00
30.00
25.00
% 20.00
15.00
10.00
5.00
0.00
Technologies
Cost increase (%) Yield increase (%)
Fig. 10.15 Technology-wise cost and yield increase

226
Table 10.12 Increase rice yield with technologies over the conventional method (%)
Technology Northern Western Eastern Central Southern Average
PU HA UP GU OD WB MP AP TN KA
AMOM 24.49 12.5 7.14 10.45 10.2 12.96
AWD 32.65 27.08 22.45 25 23.53 5.66 22.73
DI 22.92 22.92
DSR −2.04 5.04 20.83 2.08 6.25 1.89 5.68
IMD 2.08 2.08
IMP 26.53 14.58 13.45 18.75 16.16 12.5 22.92 27.45 1.89 17.14
IPM 20.41 27.08 9.43 18.97
MH 2.94 16.67 6.25 21.57 3.77 10.24
MSRI 34.69 25 23.95 22.29 22.92 24.9 37.5 27.32
MT 12.43 14.58 12.5 5.7 11.30
SITG 30.61 18.75 11.34 20.32 17.37 22.92 29.17 21.57 16.98 21.00
SRI 26.23 26.53 29.17 25.49 24.53 26.39
SSD 10.42 10.42
Note DSR direct sown rice; MSRI modified system of rice intensification; MT machine transplanting; IMP improved management practices; AMOM application
of more organic manure; SITG supplemental irrigation through groundwater; DI drip irrigation; AWD alternate wetting and drying; IMD irrigation at minimum
depth; IPM integrated pest management; MH machine harvesting SSD subsurface drainage
10 Economics of Adoption of Rice Production and Management Technologies
10.3 Constraints in Technology Adoption 227
10.3 Constraints in Technology Adoption
The major constraints faced by the farmers are lack of skills in adopting the technolo-
gies and the cost involved in actually getting these technologies in the field. A recent
study indicated that transaction cost associated with the technology adoption is higher
and this discouraged the farmers in their timely adoption (Palanisami et al. 2015a).
Another study on technology adoption had indicated that besides timely water sup-
plies, technologies not matching with the local situation, high capital or initial cost
of some of the technologies such as drip and sprinkler, maintenance problems, etc.
are the major constraints (Table 10.13) (Palanisami et al. 2015b; GOI 2008).
Table 10.13 Major constraints in technology adoption

State Total no. High Too Lack of MaintenanceNeed Not
of capi- technical skills problem skilled matching
farmers tal/initial to adopt labour with
cost local
situation
Northern region
Haryana 5 (100) – – – 2 (40) – 3 (60)
Punjab 94 (100) 28 (30) 7 (8) 1 (1) 4 (4) 16 (17) 38 (39)
Uttar 144 40 (28) 27 (19) 21 (15) 34 (24) 14 (10) 8 (6)
Pradesh (100)
Western region
Gujarat 212 32 (15) 23 (11) 46 (22) 51 (24) 12 (6) 48 (22)
(100)
Eastern region
Odisha 8 (100) – – 8 (100) − – –
West 464 48 (10) 35 (8) 11 (2) 100 (22) 21 (5) 249 (53)
Bengal (100)
Central region
Madhya 18 (100) – – 8 (44) – – 10 (56)
Pradesh
Southern region
Tamil 90 (100) 29 (32) 16 (18) 19 (21) 18 (20) 8 (9) 0
Nadu
Karnataka 480 197 (41) 21 (5) 75 (16) 14 (3) 115 (24) 58 (12)
(100)
Andhra 192 70 (36) 9 (5) 24 (13) 22 (11) 50 (26) 17 (9)
Pradesh (100)
Source Palanisami et al. 2015; Field surveys of FPARP (GOI 2008)
Figures in parentheses indicate percentage to total number of farmers
10.4 Summary
In the northern region, IMP, AMOM, SITG and MH are the major technologies
adopted. In the central region SRI, SITG, DSR and IMP are the major technologies
adopted with higher net returns from DSR, IMP and SITG. The eastern region is
familiar with IMP, DSR, MSRI, MT and MH technologies and among them; MSRI,
IMP and DSR realized higher net income. In the western region, SRI, MSRI, SITG,
MT and AMOM are the technologies noted for their higher adoption level. In the
southern region, MSRI, AWD, SITG, DSR, SSD and MT are observed for making
high net income compared to traditional practices. Interestingly, the cost of produc-
tion of rice with technology adoption has not resulted in cost reduction compared to
conventional method mainly due to more increase in cost of cultivation of rice than
yield increase thus highlighting need for introducing measures to increase the rice
productivity with reduced cultivation costs. This may be possible when the adoption
level as a whole is increased in all the regions, as the present adoption level of the
technologies is only from 5 to 30%. The major constraints faced by the farmers in
technology adoption include lack of technical skills, high initial cost of the technol-
ogy, untimely canal water supplies and high transaction cost in actually getting these
technologies to the field. Therefore, region-specific technology promotion programs
such as awareness and capacity building could help increase the adoption of these
technologies.
Chapter 11
Climate Change and Rice Production
in India: A Way Forward

11.1 Introduction
It is expected that climate change and its variability could have serious impact on
agriculture and water sectors and eventually on the food security and livelihoods
of a large section of the rural population in developing countries (IPCC AR4 2007;
IPCC 2018). The climate change, in the long run, may lead to new combinations
of changes with regard to atmospheric constituents, temperature, solar radiation and
pests, diseases and weeds. Consequently, in the Indian subcontinent, such expected
changes in climate would alter regional agricultural systems thereby a reduction in
rice production, thus leading to food insecurity.
The results based on region-wise analysis evinced that climate change has an
impact on rice productivity and production. Thus, there is an increasing concern
among farmers, researchers and policymakers due to potential impacts, the future
challenges and ways and means to deal with such challenges. The rice, being the
major staple food crop in India, the expected yield reductions due to climate change
are seemingly critical. By some estimates, over a quarter of all rice production in
a majority of the river basins in India may be lost due to climate change and its
variability leading to extreme weather consequences like frequent floods and droughts
(Palanisami et al. 2014). Climate modellers are using innovative approaches to predict
possible climate change and providing inputs to plan future strategies. This has to be
seriously considered by the governments to invest in preparing the farming sector to
cope with the impacts and sustain rice production and income of rice farmers.

230 11 Climate Change and Rice Production in India: A Way Forward
11.2 Overview of Climate Change Impact on Rice

Production in India
Currently, the area under rice in the country is about 44 million hectares. The total
area under rice in 13 rice growing states selected in this study is approximately
38.44 million hectares constituting 87% of the total area and 89% of the total rice
production in the country. As discussed in Chap. 2, the overall growth of rice area
during the four decades was 0.4%, the growth in yield was 1.95% and the growth in
rice production was 2.36%. Except for a decline in growth rates in states like Bihar,
Chhattisgarh, Madhya Pradesh and Tamil Nadu, the overall trends in the area, pro-
ductivity and production showed positive growth rates over the years. However, when
climate change scenarios were considered, future rice production in different regions
of India has shown varying levels of rice production as documented in Table 11.1
and Fig. 11.1.
In the northern region, as per RCP4.5 (medium emission) and RCP8.5 (high emis-
sion) scenarios, there would be a substantial reduction in rice yield during 2021–2050
(mid-century) and 2071–2100 (end-century) periods. Under both the scenarios, the
percentage reduction in the later period will be more. The changes in mean yield
show a variation from 6.9 to 12.7% as per RCP4.5 scenario and 8.5–21.4% as per
RCP8.5 scenario. Yield variability as estimated by RCP8.5 scenario is more than that
predicted by RCP4.5 scenario for both periods. The first scenario predicts an increase
of 28.7% yield variability during 2021–2050, while the second scenario predicts an
increase of 47% for the same period. For the second period, RCP4.5 predicts a small
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
% Change
-1.00 Southern Central Eastern Western Northern India

-2.00
-3.00
-4.00
-5.00 Regions
-6.00
-7.00
-8.00
-9.00
-10.00
-11.00 RCP4.5 Mid RCP8.5 Mid RCP4.5 End RCP8.5 End
-12.00
Fig. 11.1 Region-wise impact of climate change on rice yield under different scenarios
Table 11.1 Impact of climate change on rice yield and production in the future, India
Regions Current Current Current Projected Projected mean rice Projected rice production (Million tonnes)
area yield produc- area yield (Kg/ha)a
(Mil- (kg/ha) tion (Mil- Medium High Medium emission High emission (RCP8.5)
lion ha) (Mil- lion ha) emission emission (RCP4.5)
lion (RCP4.5) (RCP8.5)
tonnes)
MC EC MC EC MC EC MC EC
Northern 10.06 3218 32.38 10.28 2860 2697 2767 2199 29.39 27.72 28.43 22.59
(−9.23) (−14.40) (−12.19) (−30.22)
Central 3.54 1704 6.04 3.49 1451 1211 1433 840 5.06 4.23 5.00 2.93
(−16.17) (−30.02) (−17.24) (−51.47)
Eastern 16.63 2034 33.81 16.59 2062 2053 2067 2082 34.20 34.05 34.28 34.53
(1.15) (0.72) (1.38) (2.11)
Western 0.79 2076 1.64 0.80 2048 2019 2046 1981 1.64 1.62 1.64 1.59
(0.27) (−1.12) (0.20) (−2.99)
Southern 7.42 2896 21.49 7.44 3220 3491 3250 3892 23.94 25.96 24.16 28.94
(11.41) (20.79) (12.44) (34.65)
All India 44.14 2416 106.65 44.32 2345 2289 2332 2174 103.93 101.45 103.33 96.33
11.2 Overview of Climate Change Impact on Rice Production in India
(−2.55) (−4.87) (−3.11) (−9.67)

Figures in parentheses indicate % change in rice production over the current production
MC denotes mid-century (2021–2050) and EC denotes end-century (2071–2100)
Current area, yield and production refer to the year 2013–14
a The projected mean yield was calculated from the current yield by using the percentage change in mean yield in different regions of India. The future production
was worked out by multiplying the projected rice area with a projected mean yield
231
reduction in yield variability of 2.1%, whereas RCP8.5 scenario predicts an increase

in yield variability of 25.9%.
In the western region, as per RCP4.5 and RCP8.5 scenarios, there would be
only marginal reduction of rice yield during the periods 2021–2050 and 2071–2100.
The changes in mean yield show a variation of 1.9–5.6% during the first period,
whereas during the second period, it shows a reduction of 5.8% as per RCP4.5 and
an increase of 14.7% as per RCP8.5. The model projects drastic yield variability
under both scenarios. The predicted percentage of yield variability as estimated by
RCP8.5 scenario is more than that predicted by RCP4.5 scenario for both periods.
In the central region, as per RCP4.5 and RCP8.5 scenarios, there would be a
substantial reduction in percentages of rice yield during the periods 2021–2050 and
2071–2100. The reduction in mean yield shows a variation from 14.85 to 15.94%
during the first period as per the two scenarios and as per RCP8.5 scenario, it varies
from 28.92 to 50.71% during the second period. As expected, yield variability as
estimated by RCP8.5 scenario was more than that predicted by RCP4.5 scenario for
both periods. The first scenario predicts an increase of 184% yield variability dur-
ing 2021–2050 while the second scenario predicts an increase of 240% for the same
period. For the second period, RCP4.5 predicts an increase of 344% in yield variabil-
ity, whereas RCP8.5 scenario predicts an increase in yield variability of 667%. The
variations are quite large and hence need more careful considerations for planning
any future measures.
In the eastern region, as per RCP4.5 and RCP8.5 scenarios, there will be a marginal
increase in rice yield during the periods 2021–2050 and 2071–2100. The increase in
mean yield varies from 1.40 to 1.63% during the first period as per the two scenarios
and it varies from 0.96 to 2.36% during the second period as per the two scenarios.
yield variability during 2021–2050 while the second scenario predicts an increase of
41% for the same period. For the second period, RCP4.5 predicts a higher increase
variability of 139%. Thus, we could conclude that climate change may not have
an adverse effect on the mean rice yield in eastern region but yield variability may
be substantial. Hence, measures are to be taken to decrease the yield variability or
initiate policy and trade measures to compensate for the yield variations.
In the southern region, as per RCP4.5 and RCP8.5 scenarios, climate change
has a positive effect on rice yield. Overall, there will be an increase in rice yield
during the periods between 2021–2050 and 2071–2100 under both the scenarios.
The percentage increase in mean yield shows a variation from 11.2 to 12.23% during
the first period as per the two scenarios, and it varies from 20.57 to 34.40% during
the second period as per the two scenarios. Yield variability is expected to reduce
in both periods under the two scenarios. The predicted reduction in variability is
more with respect to the RCP8.5 scenario. The first scenario predicts a decrease in
variability of 9.1% during 2021–2050 while the second scenario predicts a decrease
of 20.3% for the same period. For the second period, RCP4.5 predicts a reduction
of 18% in yield variability whereas RCP8.5 scenario predicts 101% reduction in
11.2 Overview of Climate Change Impact on Rice Production in India 233
variability. Thus, we find that climate change would increase yield accompanied by
a substantial reduction in yield variability in the region.
In the southern region, rice production will increase more in the end-century under
both medium and high emission scenarios showing the stability of rice production in
the region. Being the traditional rice belt of the country, it is crucial to know how best
the stability in rice production could be achieved through appropriate measures such
as the adoption of different climate resilient practices in rice. Similarly, in the eastern
region, rice production is expected to increase marginally and hence efforts should
be needed to stabilize the rice production in the region as well. As both southern and
eastern regions combined represent 52% of the rice production in the country, it is
a good signal that climate change will not have any significant negative impact on
rice production. Other regions, accounting for 38% of rice production in the country,
will be facing a substantial climate change impact on rice production. It is possible
that this impact can be addressed through several adaptation and policy measures so
that rice production or supply in the market can be maintained to meet the demand.
Assuming medium emission scenarios will be prevalent in the future, rice produc-
tion will be approximately 104 million tonnes during the mid-century (2021–2050)
and 101 million tonnes during the end-century (2071–2100) indicating that there will
be an overall reduction in rice production by 2.5% during the mid-century period
and about 5% during end-century period (Table 11.1). The mid-century scenario is
not alarming in terms of production. However, the future focus of rice production in
the country should be on sustainable practices that will reduce the impact of climate
change on the environment. Moreover, efficiency in input usage has to be improved,
thus emphasizing not only on climate adaptation but also climate mitigation mea-
sures in rice production. This is possible given that there is a potential to improve
the efficiency of current technologies used in rice farming systems in the country
that can contribute to reduce the greenhouse gas emissions. This requires policy and
investment support to reduce the emissions in the mid- and end-century scenarios.
Hence, a 5% reduction in rice production in the future can be avoided if proper
climate change-related policies are implemented.
Overall, the results indicated a nonlinear effect of both maximum temperature and
rainfall in reducing the yield of rice crop in most of the rice growing regions. Further,
the variations in yield could increase with an increase in the average maximum
temperature.
To summarize, in southern region, climate change will induce an increase in yield
accompanied by a substantial reduction in yield variability, whereas in the eastern
region, there will be a marginal increase in rice yield accompanied by high increase
in yield variability. Though these two regions offer more scope for increasing rice
production in the future, eastern region’s rice production is subject to the risk of high
yield variability which needs to be addressed through technological interventions. In
the case of northern, western and central regions, there will be a marginal reduction
in rice yield accompanied by medium to high variability in rice yield during the mid-
and end-century periods. Among these three regions, central region is more prone to
climate change and variability, and hence, special attention is needed for stabilizing
rice production as rice area and productivity are already showing declining growth
rate in the region. Suitable adaptation strategies have been examined in detail in
different regions and those showing promising responses at the farm level have to be
promoted. In what follows, possible interventions and approaches are suggested.
11.3 A Way Forward
11.3.1 Future Rice Production Policy
The Population Foundation of India projected that the country’s population will be
1546 million, 1695 million and 1824 million by the end of the year 2030, 2040
and 2050, respectively (CRRI 2013). Accordingly, it is estimated that the domestic
demand for rice will be 121.2, 129.6 and 137.3 million tonnes during the end of
the above periods. In addition to this, India exports about 3.5 million tonnes of
basmati and 6.9 million tonnes non-basmati rice every year (CRRI 2013). To meet
out the above challenges, the productivity of rice has to be increased, i.e. from
current level of 2.4 tonnes per ha to 3.3 tonnes per ha. This should not be difficult to
achieve with improvement in the technologies and introduction of new rice varieties
although the growth rate in production during the time periods between 2007–08 and
2011–12 was 0.36%, far below the population growth rate of 1.63%. No doubt, the
current decelerating trend in production and yield is of concern and hence has to be
necessarily reversed to meet the growing demand and also to meet the needs of export
markets. Moreover, the profit margin in rice cultivation has been declining and thus
making cultivation of this crop unattractive to farmers, especially to the smallholders,
who constitute the majority of rice growing farmers in the country. Therefore, a two-
pronged strategy of developing new technologies, by expanding research investments
to bring efficiency in production, and implementation of favourable government
policies will help in increasing rice production and productivity in the country to meet
the demand arising in the future. As there will be an only a marginal increase (0.4%)
in rice area in the future, any increase in rice production should necessarily come
from productivity increases only. Further, adaptation strategies will help improve the
rice productivity to a larger extent as indicated in the region-wise chapters of this
book.
11.3.2 Upscaling and Mainstreaming the Adaptation

Strategies
The evidence of increased climate variability are though obvious, the severity, how-
ever, demands more attentions from water and agricultural managers. For new
weather patterns repeating for several subsequent seasons, farmers have to adapt
strategies to maintain their crop productivity. The reassuring aspect of this argument
11.3 A Way Forward 235
is that these adaptation options to cope with climate variability would also help to
minimize the impacts from other factors like population, urbanization, change in
consumer and market demand, poverty, and climate change in the future. Adaptation
will benefit from improved forecasting and climate modelling which in turn imply
the need for strengthening of data collection initiatives. Agriculture sector needs
better tools for prediction and monitoring and introducing sustainable and efficient
measures.
Since climate change is a continuous phenomenon and varies between regions,
any adaptation and mitigation measures have to be specific in order to minimize
the reduction and as well as variability (risk) in rice yield. Adaptations are nothing
but adjustments or interventions which are followed to manage the losses or to take
advantage of the opportunities offered by climate change (IPCC 2001, 2007). Most of
the adaptations that occur at farm-level focus mainly on farming-related interventions
or adjustments. Further, they are related to short-term periods and influenced by
seasonal climate variations. The earlier chapters have indicated a wide range of
adaptation strategies specific to regions. Some of the strategies are direct seeding of
rice, modified system of rice intensification (SRI), supplemental irrigation, alternate
wetting and drying and improved management and extension practices.
However, access to and adoption of these strategies by farmers varied largely
between regions and between states (Palanisami et al. 2014). Hence, region-specific
appropriate and affordable adaptation strategies need to be identified for upscaling
by the respective state governments since agriculture is a state subject in India. Trade
off agricultural produce between the states should be made easier. Transaction cost of
the adaptation strategies is a major component in technology adoption. Transaction
costs refer to the costs incurred while taking the technology to the fields. Normally,
it accounts for 2–5% of the total cost of production of crops and minimizing the
transaction cost at farm level will help to speed up the adoption of technologies and
this should be well addressed by the extension agencies while promoting various
adaptation strategies in different locations. Technology upscaling is important and
should be supported by the convergence of several ongoing government programs.
Developing public–private partnership in implementation, awareness creation and
designing need specific capacity-building training programmes for the extension
workers and farmers would improve adoption levels. Climate-smart technologies
and a package of practices should be included in all the crop production manuals
of the state agriculture departments and universities and made available in local
languages.
Further, risk in rice production (in terms of high variability in rice yield) is expected
to increase more due to climate change and will influence the rice production in all the
regions. Hence, necessary weather-based crop insurance products mostly available
for rainfed crops can also be introduced to rice to address the risk in rice production
(David et al. 2012).
It is estimated that the adaptation practices namely system of rice intensifica-
tion, machine transplantation, alternate wetting and drying and direct seeding could
reduce the water and labour use by 10–15% and stabilize rice production in the long
term. Hence, simply by promotion of these adaptation practices will minimize the
income losses to the farmers. Moreover, there is a need for technology upscaling
which should be backed up with well-planned capacity-building programmes for the
farmers. Further, an assessment of farm-level adaptation strategies over time and
space is important in providing information necessary to the implementing agencies
including government for formulating the needed policies. The adaptation strategies
related to rice production include physical investment to create farm structure as
well as management technologies such as improved crop and water management
practices, several of which have been already studied by the researchers (Palanisami
et al. 2011; Abraha and Savage 2006; Till et al. 2010). Besides, subjecting the farm-
ers to the actual cost of adaptation vis-à-vis to the expected income associated with
adaptation measures may motivate them towards better adaptation measures.
11.3.3 Government’s Commitment to Address Climate

Change
The release of the first National Action Plan on Climate Change (NAPCC) on cli-
mate mitigation and adaptation through sustainable development in 2008 evinced
India’s commitment to adapt to climate change. The plan has identified eight core
‘National Missions’ to be followed and directed related Ministries of Government
to share detailed implementation plans to the Prime Minister’s Council on Climate
Change. India is thus one of the first countries to come out with NAPCC. However,
considering the vast area, the diverse nature of the different states in the country and
also the varied impacts of climate across ecosystems and sectors, it is important to
device specific strategies within each state, build institutional capacity and develop
measures to implement the adaptation plans. Towards this end, the state governments
are being sensitized at different levels and expected to device the state climate change
adaptation frameworks under the overall supervision of the Ministry of Environment
and Forests (MoEF), Government of India. Within the states, the responsibility has
been given to the Department of Environment and Forests. One of the primary gaps
connected to the planning exercise is the lack of adequate scientific/or technological-
based inputs to the state climate and sector adaptation plans. The results on impact
due to climate change on the rice production outlined in the book will help to address
some of these gaps.
The National Mission for Sustainable Agriculture and the National Water Mission
have been identified as the most critical missions in the country. These missions
are critical because they hold millions of small-scale farmers in India and their
livelihoods, who are highly vulnerable to climate change. The National Mission on
Sustainable Agriculture aims to devise strategies to make agriculture more resilient
to climate change. The National Water Mission aims to improve water use efficiency
by 20%, explore the possibilities of basin-level management strategies in Indian
rivers and capacity-building options. Further, these missions have to be supported
with inputs from information and biotechnology, as well as credit and insurance
11.3 A Way Forward 237
support mechanisms. In the process, the respective state governments need support
to develop and implement comprehensive adaptation and policy frameworks. Some
of them have initiated measures, but need expertise and investments, especially in
terms of integrating adaptation technologies with larger development programs and
further upscaling.
11.4 Specific Measure Relevant to Rice
11.4.1 Augmenting and Managing Water Resources
The scarcity of water is expected to become an ever-increasing problem in India

as a result of climate change impacts. The situation is quite alarming and cautions
that India needs to take drastic measures to improve water management, especially
groundwater usage (World Bank 2018). The water balance of the regions will change
in the future due to the accelerated rate of evaporation from soil and water bodies,
change in run-off and soil moisture storage and increased rate of transpiration from
plants. Furthermore, it is expected that groundwater depletion will take place due to
the combination of shorter duration but more intense rainfall (meaning more run-off
and less infiltration) combined with increased evapo-transpiration and increased irri-
gation (Konikow and Kendy 2005). Lowering of groundwater table and consequently
an increase in the energy needed to pump water will make irrigation more expensive
particularly for rice in selected regions. Therefore, the strategies and tools must be
strengthened to be able to adapt to climate change and to cope with between and
within crop season weather variability. In future, a larger amount of water (ranging
from 1200 to 2500 mm per hectare) will be needed since India has to produce more
rice to meet food grains need for consumption and export. Currently, to produce
1 kg of rice, about 2500–3500 l of water (water productivity) is used in India. There-
fore, to meet current challenge of declining water resource base and competition
from other sectors, the water productivity of rice needs to be improved by bringing
down the water productivity to the level of 2000 L/kg of rice (CRRI 2013; Deelstra
et al. 2016; Kakumanu et al. 2016). Our recent research findings show that by using
improved crop and water management practices, it is possible to increase the rice
productivity and bring down the water use significantly thus improving the water use
efficiency (WUE). For e.g., in Krishna, Godavari and Cauvery basins, the average
WUE under SRI is 9.1 kg/ha/mm compared to 4.6 kg/ha/mm under conventional
practices. Similarly, with alternative wetting and drying (AWD) method, the aver-
age WUE is 4.2 kg/ha/mm compared to 2.5 kg/ha/mm under conventional practices
(Palanisami et al. 2014; Kakumanu et al. 2016).
In order to achieve the goal of improving the water availability and to meet the
demand by 2050, the Ministry of Water Resources, Government of India, is tak-
ing several steps. This very optimistic goal is expected to be achieved by several
measures, including the implementation of integrated water resource management
(IWRM) plans, enhancing efficiency in water use and demand management (National
Water Mission 2008). However, to reach the goal and ensure implementation, sub-
stantial investments and cooperation by the governments are necessary.
11.4.2 Managing Soil Health
At the farm level, small-scale farmers are interested in improved soil and land man-
agement technologies that can also benefit them with carbon credits. They need
technological support to address the challenges in soil management.
The different land-use and soil management options will concomitantly contribute
to restoring soil organic matter (SOM) stocks (Lal 2004). Such soil organic matter
is also intimately linked to climate effects on the water cycle. Sufficient SOM stocks
reduce erosion risks linked to intense precipitation events, help conserve water within
the soil profile and help retain nutrients in soils, especially in sandy soils. Thus, build-
ing SOM stocks in soils is one of the best possible measures to buffer the effects
of climate extremes on crop productivity. In addition, by building SOM stocks in
soils under their custody, smallholder farmers contribute to carbon sequestration and
storage. In future scenarios, smallholder farmers could get credits (i.e. remuneration)
for this important environmental service of CO2 sequestration and carbon storage in
their lands. Increased crop productivity, reduced inter-annual variability and remu-
neration for carbon storage would go a long way in helping smallholder farmers in
rural India.
Therefore, the management of land and soil is crucial for optimizing water use
efficiency. In other words, it is at the farm level, that changes have to take place.
This might be possible by initiating various measures at the village level, includ-
ing village knowledge centres, and SHGs to provide necessary help to farmers and
women with information and resources to adapt to climate change and training and
capacity-building initiatives (Kakumanu et al. 2018). The ongoing information access
revolution related to the spread of mobile networks and the rapidly developing sit-
uation of mobile Internet access devices could prove to be a new means for state
institutions and local agencies for public interaction.
11.4.3 Developing Multiple Abiotic Stress-Tolerant Rice

with Climate Resilience
The yield of a crop is influenced by biotic stresses, abiotic stresses and nutritional
factors either separately or in combinations. The realization of full yield potential
is prevented by abiotic stress primarily due to drought, heat, cold, salinity and sub-
mergence. Thus, abiotic stresses always occur in combination. For instance, heat
and drought stress or submergence and salinity occur simultaneously. The situation
11.4 Specific Measure Relevant to Rice 239
will become further complicated as the climate keep changing in future with rapid
fluctuations in the occurrence of such stresses. Therefore, tolerance towards stress is
a complex phenomenon and no single approach could provide an immediate solution
to the multiple environmental stresses that a plant experiences during its lifecycle.
Further, a systematic evaluation of novel germplasm and putative stress-related genes
under field conditions has to be undertaken to develop a robust set of new germplasm
and better adapted to withstand stresses. Further, identification and pyramiding of
multiple genes for stress tolerance in high-yielding genetic backgrounds are feasible
and the efforts would be practicable to tailor novel rice genotypes in future (CRRI
2013).
11.4.4 Climate Resilient Rice Production Technologies
The impacts due to climate change can be addressed with the following strategies (i)
making adjustments in rice production methods and introduction of new rice strains to
withstand high temperature, (ii) developing and popularizing multiple stress-tolerant
varieties to withstand and perform better, (iii) adopting environment-friendly and
least environment footprint rice-based cropping system, (iv) achieving a sustainable
rice production, development of cultivation practices to maintain soil health and
conserve natural resource base has to be promoted, (v) accelerating adaptation to cli-
mate change and better management of agricultural risks towards resilient agriculture
and (vi) promoting partnership with rural communities and other stakeholders in a
climate-smart model for agricultural development along with innovative agricultural
risk management measures (CRRI 2013).
The National Rice Research Institute (NRRI) in its Vision 2050 clearly states
that entrepreneurship development is important. By providing specialized training
to educated rural youth, one can encourage and make them successful entrepreneurs.
The trained youths, in the long run, will act as change agents to disseminate and
promote the adoption of rice-related interventions including spread of high-yielding
varieties (HYV) and seed production, production and marketing of rice, farm mecha-
nization, value-added rice products, etc. (CRRI 2013). Further, strengthening gender
participation needs to be ensured. With much emphasis on women empowerment and
sensitization on gender equality and equity, rice researchers have to develop projects
for generation of technology that would suit the needs, aspiration and problems of
women in agriculture. Special attention on the drudgery of women in rice farming
needs to be made such that the occupational hazards while transplanting, weeding
and post-harvest can be mitigated and ultimately the efficiency of women in rice
production can be increased.
11.5 Summary
The preceding discussions on climate change clearly indicate that such changes
would lead to a negative impact on crop yields in most of the regions in India. One
way of addressing these impacts will be through the adoption of feasible strategies
across the rice value chain that will help in sustaining rice production. Since the
current technology adoption level is comparatively low (ranging from 8 to 12%) more
efforts are needed to promote the strategies to be adopted by the farmers (Palanisami
et al. 2013, 2015). Therefore, future research and policy interventions have to focus
on studying the economics of different adaptation strategies already being followed
by the farmers as well as the new strategies that are being field tested in different
locations by different agencies. Moreover, studies on farmers’ responses to these
strategies are also warranted as the most efficient strategies in terms of profit are not
followed by farmers owing to operating constraints in adopting them. Provisions of
assured and timely credit, marketing facilities with remunerative prices and weather-
based crop insurance products will help improve the technology adoption. Public and
private sector partnership in promoting the interventions need to be strengthened. In
this context, the following recommendations are important:
• Capacity building to farmers, through awareness programs, to educate them on pos-
sible impacts on climate change, improving adaptation strategies and the effective
use of support systems. The development and implementation of the adaptation
measures need to be done through ‘learning-by-doing’.
• Build creative partnerships to facilitate adaptation through stakeholder participa-
tion; work closely with local institutions; demonstrate multiple adaptation strate-
gies; and carry out a cost-benefit analysis of adaptation strategies. A careful eco-
nomic analysis of the instruments will facilitate adaptation, strengthening farmer
producer groups/organizations to support farm mechanization, trade and better
price to rice farmers.
• Crop diversification for effective use of land and water, which in turn will ensure
better income to rice farmers.
• Existing weather-based insurance products to be fine-tuned according to the
regions and crops.
• Use of ICT-based tools, mobile phones and rice apps to improve farmer connec-
tivity and rice productivity. Several states have already developed user-friendly
software/portals to facilitate the use of online extension services to reach larger
farmer numbers.
• The availability of relevant data/information must be ensured and such information
is very crucial for making climate projections and identifying vulnerable groups
and hot spots.
• Information sharing, including the early warning system, between sectors is essen-
tial for developing effective and efficient climate change adaptation strategies, both
in the short and long term.
• State level initiatives like ‘Climate Water Forum’ as was done in Andhra Pradesh
state, India will help to address key emerging issues and make them more sustained.
11.5 Summary 241
• Policy and trade measures to make it easier for farmers to sell their produce. This
includes assuring acceptable minimum support price and facilities to store and
market the harvested products to minimize post-harvest losses.
References
Abraha MG, Savage MJ (2006) Potential impacts of climate change on the grain yield of maize for
the midlands of KwaZulu-Natal, South Africa. Agric Ecosyst Environ 115:150–160
African Devlopment Bank et al. (2009). Poverty and climate change: reducing the vulnerability of
the poor through adaptation. Available at http://documents.worldbank.org/curated/en/
534871468155709473/pdf/521760WP0pover1e0Box35554B01PUBLIC1.pdf
Adhikari BM, Bag K, Bhowmick MK, Kundu C Status paper on rice in West Bengal. Retrieved
from http://www.rkmp.co.in
Aggarwal PK (2008) Global climate change and Indian agriculture: Impacts, adaptation and
mitigation. Indian J Agric Sci 78(10):911–919
Aggarwal PK, Hebbar KB, Venugopalan MV, Rani S, Bala A, Biswal A, Wani SP (2008)
Quantification of yield gaps in rain-fed rice, wheat, cotton and mustard in India. Global theme
on agro-ecosystems report no. 43. International Crops Research Institute for the Semi-Arid
Tropics. Patancheru, Andhra Pradesh, India
Aggarwal PK, Singh AK, Samra JS, Singh G, Gogoi AK, Rao GGSN, Ramakrishna YS (2009)
Introduction. In: Aggarwal PK (ed) Global climate change and Indian agriculture. ICAR,
New Delhi, pp 1–5
Aggarwal PK, Kropff MJ, Cassman KG, Berge HFM (1997) Simulating genotypic strategies for
increasing rice yield potential in irrigated, tropical environments. Field Crops Res 51:5–17
Aggarwal PK, Mall RK (2002) Climate change and rice yields in diverse agro-environments of
India: effect of uncertainties in scenarios and crop models on impact assessment. Clim Change
52(3):331–343
Aggarwal PK, Pathak H, Kumar N (2009) Global climate change and indian agriculture: a review
of adaptation strategies. Trust for Advancement of Agricultural Sciences, New Delhi
Agricultural Statistics (2014) Agricultural statistics at a glance. Directorate of economics and
statistics. Government of India
Agricultural Statistics (2016) Agricultural statistics at a glance. Directorate of economics and
statistics. Government of India
Akaike H (1973) Information theory and an extension of the maximum likelihood principle. In:
Petrov BN, Csaki F (eds) Proceedings of the 2nd international symposium on information
theory. Akademiai Kiado, Budapest, pp 267–281
Anderson J, Hazell P (1989) Variability in grain yields: implications for agricultural research and
policy in developing countris. Intrnational Food Policy Research Institute. The Johns Hopkins
University Press. Baltimore and London
Annamalai H, Hafner J, Sooraj KP, Pillai P (2013) Global warming shifts the monsoon circulation,
drying South Asia. J Clim 26(9):2701–2718

https://doi.org/10.1007/978-981-13-8363-2
244 References
Asche F, Tveterås R (1999) Modeling production risk with a two-step procedure. J Agric Resour
Econ 24:424–439
Aye GC, Ater PI (2012) Impact of climate change on grain yield and variability in Nigeria: a
stochastic production model approach. Mediterr J Soc Sci 3(16):142–150
Bal PK, Ramachandran A, Palanivelu K, Thirumurugan P, Geetha R, Bhaskaran B (2016) Climate
change projections over India by a downscaling approach using PRECIS. Asia Pacific J Atmos
Sci 52(4):353–369
Baltagi BH (2005) Econometric analysis of panel data, 3rd edn. Wiley, New York
Barnwal P, Kotani K (2010) Impact of variation in climate factors on crop yield: a case of rice crop
in Andhra Pradesh, India. Working paper. IUJ Research Institute, International University of
Japan, Minami-Uonuma
Barton DN, Kakumanu KR, Palanisami K, Tirupataiah K (2012) Analysis of economic incentives
to manage risk at farm level in the context of climate change. In: Nagothu US, Gosain AK,
Palanisami K (eds) Water and climate change: an integrated approach to address adaptation
challenges. Macmillan Publishers India Ltd., New Delhi, pp 143–168. ISBN
978-935-059-059-1
Basavaraja H et al (2008) Technological change in paddy production: a comparative analysis of
traditional and sri methods of cultivation. Indian J Agric Econ 63(4)
Bindi M, Fibbi L, Bernardo Gozzini, Serena Orlandini, Miglietta F (1996) Modelling the impact of
future climate scenarios on yield and yield variability of grapevine. Clim Res 7:213–224.
https://doi.org/10.3354/cr007213
Boubacar I (2010) The effects of drought on crop yields and yield variability in Sahel. Int J Econ
Finance 4(12)
Buchinsky M (1998) Recent advances in quantile regression models: a practical guideline for
empirical research. J Human Resour 33(1):88–126. Retrieved from https://EconPapers.repec.
org/RePEc:uwp:jhriss:v:33:y:1998:i:1:p:88-126
Bouman B, Barker R, Humphreys E, Tuong TP, Atlin G, Bennett J, Dawe D, Dittert K,
Dobermann A, Facon T, Fujimoto N, Gupta R, Haefele S, Hosen Y, Ismail A, Johnson D,
Johnson S, Khan S, Shan L, Masih I, Matsuno Y, Pandey S, Peng T, Muthukumarisami T,
Wassman R (2007) Rice: feeding the billions. In: Molden D (ed) Water for food, water for life:
a comprehensive assessment of water management in agriculture. Earthscan/IWMI, London,
UK/Colombo, Sri Lanka, pp 515–549
Cabas-Monje JW, Weersink AO, Edward O (2010) Crop yield response to economic sit and
climatic variables. Clim Change 101:599–616. https://doi.org/10.1007/s10584-009-9754-4
Cambell B, Mannm W, Melendez-Ortiz R, Streck C, Tennigkeit T (2011) Addressing agriculture
in climate change negotiations: a scoping report. Retrieved from https://ccafs.cgiar.org/publi
cations/addressing-agriculture-climate-change-negotiations-scoping-report#.XATlSPZuLIU.
Accessed on 3 Dec 2018
Central Rice Research Institute (CRRI) (2013) Vision 2050. Indian Council of Agricultural
Research, Cuttack, India
Chapagain T, Riseman A, Yamaji E (2011) Achieving more with less water: Alternate wet and dry
irrigation (AWDI) as an alternative to the conventional water management practices in rice
farming. J Agric Sci 3(3):3–13
Chaturvedi RK, Joshi J, Jayaraman M, Govindasamy B (2012) Multi-model climate change
projections for India under representative concentration pathways. Curr Sci 103(7):791–802
Chen C-C, Chang CC (2005) The impact of weather on crop yield distribution in taiwan: some
new evidence from panel data models and implications for crop insurance. Agric Econ 33
(s3):503–511
Chen C-C, McCarl BA, Schimmelpfennig D (2004) Yield variability as influenced by climate: a
statistical investigation. Clim Change 66(2):239–261
Chen S, Cai SG, Chen X, Zhang GP (2009) Genotypic differences in growth and physiological
responses of transplanting and direct seeding cultivation in rice. Rice Sci 16(2):143–150
References 245
Cheralu C (2011) Status paper on rice in Andhra Pradesh. Rice Knowl Manag Portal 1–36.
Retrieved from http://www.rkmp.co.in/sites/default/files/ris/rice-state-wise/RiceState
Wise Andhra Pradesh_0.pdf
Cherchi A, Annamalai H, Masina S, Navarra A (2014) South Asian summer monsoon and the
eastern Mediterranean climate: the monsoon-desert mechanism in CMIP5 simulations. J Clim
27(18):6877–6903
Choi I (2001) Unit root tests for panel data. J Int Money Finance 20(2):249–272
Cline WR (2007) Global warming and agriculture: impact estimates by country. Peter Institute for
international economics, Columbia University Press, 250 pp
Das SR (undated) Status paper on rice in Orissa. Rice Knowledge Management Portal, Directorate
of Rice Research, Hyderabad. Retrieved from http://www.rkmp.co.in/sites/default/files/ris/rice-
statewise/Status%20Paper%20on%20Rice%20in%20Orissa.pdf
De-Graft AH, Kweku KC (2012) The effects of climatic variables and crop area on maize yield and
variability in Ghana. Russ J Agric Soc-Econ Sci 10:10–13. https://doi.org/10.18551/rjoas.
2012-10.02
Deelstra J, Kakumanu KR, Kallam SR, Nagothu US, Geethalakshmi V, Lakhsman A, Arasu MS
(2016). Water productivity under different rice growing practices: results from farmer led field
demonstrations in India. In: Nagothu US (eds) Climate change and agricultural development:
improving resilience through climate smart agriculture, agroecology and conservation.
Earthscan Food and Agriculture, Routledge, UK. ISBN 978-1-138-92227-3
Deressa T, Hassan R (2009) Economic impact of climate change on crop production in ethiopia:
evidence from cross-section measures. J Afr Econ 18:529–554. https://doi.org/10.1093/jae/
ejp002
Dey MM, Upadhyaya HK (1996) Yield loss due to drought, cold, and submergence in Asia. In:
Evenson RE, Herdt RW, Hossain M (eds) Rice research in Asia: progress and priorities. CAB
International/IRRI, Wallingford, UK, pp 35–58
Dwivedi JL (2011). Status paper on rice in Uttar Pradesh. Rice Knowledge Management Portal.
Retrieved from http://www.rkmp.co.in
El-Ashry M (2009) Adaptation to climate change: building resilience and reducing vulnerability.
Recommendations from the 2009 Brooking Blum Roundtable, United Nations Foundations
Farooq M, Kadambot HMS, Rehman H, Aziz T, Lee D, Wahid A (2011) Rice direct seeding:
experiences, challenges and opportunities. Soil Tillage Res 111(2011):87–98
Food and Agriculture Organization of the United Nations (FAO) (1992) Agrostat. FAO
Publication, Rome, Italy
Food and Agriculture Organization of the United Nations (FAO) (1996) Agro−ecological zoning:
guidelines FAO Soils Bulletin 73, Rome, Italy
Food and Agricultural Organization of the United Nations (FAO) (2000) Two essays on climate
change and agriculture: a developing country perspective. FAO Economic and Social
Development Paper, p 145
Food and Agriculture Organization of the United Nations (FAO) (2009) OECD-FAO Agricultural
outlook 2009–2018
Garrett HE, Woodworth RS (1971) Statistics in psychology and education. Vakils, Feffer and
Simons, Bombay
Geethalakshmi V, Lakshmanan A, Rajalakshmi D, Jagannathan R, Sridhar G, Ramaraj AP,
Bhuvaneswari K, Gurusamy L, Anbhazhagan R (2011) Climate change impact assessment and
adaptation strategies to sustain rice production in Cauvery Basin of Tamil Nadu. Curr Sci
101:342–347
Government of India (2004) National action plan on climate change in India. Retrieved from
www.moef.nic.in/downloads/home/pg01-52.pdf
Government of India (2012) Twelfth five year (2012–2017) economic sectors, vol II. Government
of India, p 3
Granger CWJ, Newbold P (1974) Spurious regressions in econometrics. J Econ 2(2):205–210
246 References
Gujja B, Thiyagarajan TM (2009) New hope for indian food security? The system of rice
intensification. Gate Keeper 143, International Institute for Environment and Development,
London
Gujarati DN (2004) Basic econometrics. McGraw Hill, New York
Gupta S, Sen P, Srinivasan S (2012) Impact of climate change on Indian economy: evidence from
food grain yields. Centre for Development Economics Working Paper 218, Delhi
Haris AA, Biswas S, Chhabra V (2010) Climate change impacts on productivity ofrice
(OryzaSativa) in Bihar. Indian J Agron 55:295–298
Harvey AC (1976) Estimating regression models with multiplicative heteroscedasticity.
Econometrica 44:461–465
Hasanthika WKAMA, Edirisinghe JC, Rajapakshe RDDP (2013) Climate variability, risk and
paddy production. J Environ Prof Sri Lanka 2(2):57–65
Hoff H (2004) Climate change adaptation in India. Research report. Postdam Institute for Climate
Impact Research, Germany
Hou AH, Chen GX, Wang ZP, Cleemput OV, Patrick WH (2000) Methane and nitrous oxide
emissions from a rice field in relation to soil redox and microbiological processes. Soil Sci Soc
Am J 64(6):2180–2186
Holst R, Yu X, Grun C (2010) Climate change, risk and grain production in China. In: 2010
AAEA, CAES, WAEA joint annual meeting. Agriculture and Applied Economics Association,
Colorado, p 29
Hundal SS, Kaur P (2007) Climatic variability and its impact on cereal productivity in Indian
Punjab. Curr Sci 92:506–512
Hurlin C, Mignon V (2006) Second generation panel unit root tests. Retrieved from http://halshs.
archives-ouvertes.fr/docs/00/15/98/42/PDF/UnitRoot_Ev5.pdf
Im KS, Pesaran MH, Shin Y (2003) Testing for unit roots in heterogeneous panels. J Econ
115:53–74
Indiastat (2013) Stat-wise production of rice in India. Retrieved from www.Indiastat.com
IPCC (2001) Climate change 2001: impacts, vulnerability and adaptation. In: Contribution of
working group III to the third assessment report on the Intergovernmental Panel on Climate
Change. Cambridge University Press, Cambridge
IPCC (2007) Climate Change 2007: the physical science basis: summary for policymakers
working group I to the Intergovernmental Panel on Climate Change fourth assessment report.
Intergovernmental Panel on Climate Change, Geneva
IPCC (2007) Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II and
III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Core
Writing Team, Pachauri, R.K and Reisinger, A. (eds.)]. IPCC, Geneva, Switzerland, 104 pp
IPCC (2014) Climate change: impacts, adaptation and vulnerability. Cambridge University Press,
Cambridge
IPCC (2018) Summary for policymakers. In: Masson-Delmotte V, Zhai P, Pörtner HO, Roberts D,
Skea J, Shukla PR, Pirani A, Moufouma-Okia W, Péan C, Pidcock R, Connors S,
Matthews JBR, Chen Y, Zhou X, Gomis MI, Lonnoy E, Maycock T, Tignor M, Waterfield T
(eds) Global warming of 1.5 °C. An IPCC special report on the impacts of global warming of
1.5 °C above pre-industrial levels and related global greenhouse gas emission pathways, in the
context of strengthening the global response to the threat of climate change, sustainable
development, and efforts to eradicate poverty. World Meteorological Organization, Geneva,
Switzerland, 32 pp
Isik M, Devadoss S (2006) An analysis of the impact of climate change on crop yields and yield
variability. Appl Econ 38:835–844
Isik M, Khanna M (2003) Stochastic technology, risk preferences and adoption of site−specific
technologies. Am J Agric Econ 85:305–317
Joshi E, Dinesh K, Lal B, Nepalia V, Gautam P, Vyas AK (2013) Management of direct seeded
rice for enhanced resource—use efficiency. Plant Knowl J 2(3):119–134
References 247
Jung H-S, Choi Y, Oh J-H, Lim G-H (2002) Recent trends in temperature and precipitation over
South Korea. Int J Climatol 22:1327–1337
Judge GG, Hill RC, Griffiths WE, Lutkepohl H, Lee TC (1988) Introduction to the theory and
practise of econometrics, 2nd edn. Wiley, New York
Just RE, Pope RD (1978) Stochastic specification of production functions and economic
implications. J Econometrics 7:67–86
Kakumanu KR, Kotapati GR, Nagothu US, Palanisami K, Kallam SR (2018) Adaptation to
climate change and variability: a case of direct seeded rice in Andhra Pradesh. J Water Clim
Change, India. https://doi.org/10.2166/wcc.2018.141
Kakumanu KR, Kaluvai YR, Nagothu US, Lati NR, Kotapati GR, Karanam S (2018) Building
farm-level capacities in irrigation water management to adapt to climate change. Irrig Drainage
67:43–54. https://doi.org/10.1002/ird.2143
Kakumanu KR, Tesfai M, Borrell A, Nagothu US, Suresh Reddy K, Gurava Reddy K (2016)
Climate smart rice production systems: studying the potential of alternate wetting and drying
irrigation. In: Nagothu US (eds) Climate change and agricultural development: Improving
resilience through climate smart agriculture, agroecology and conservation. Earthscan Food
and Agriculture, Routledge, UK. ISBN 978-1-138-92227-3
Kelkar U, Bhadwal S (2007) South Asian regional study on climate change impacts and
adaptation: implications for human development. In: Occasional paper. Human Development
Report Office, United Nations Development Programme, New Delhi, India
Kendall MG, Babington Smith B (1939) The problem of m rankings. Ann Math Stat 10(3):
275–287
Kim M-K, Pang A (2009) Climate change impact on rice yield and production risk. J Rural Dev
Nongchon-Gyeongje 2:1–13. Korea Rural Economic Institute
Konikow LF, Kendy E (2005) Groundwater depletion: a global problem. Hydrogeol J 13:317–320
Komiya S, Shoji Y, Noborio K, Yazaki T, Toojinda T, Siangliw M (2010) Climatic impacts on
greenhouse gas emissions in rice paddy fields. Paper presented at 19th world congress of soil
science, soil solutions for a changing world, Brisbane, Australia, 1–6 Aug
Koenker R, Bassett G (1978) Regression quantiles. Econometrica 46(1):33–50
Kulkarni RS, Raghuprasad KP, Mallika M (2013) Achievements, challenges, technology
dissemination of rice in India. Report Innovation in rice Production, p 215
Kumar K, Parikh J (1998a) Climate change impacts on Indian agriculture: results from a crop
modelling approach. In: Dinar A, Mendelsohn R, Evenson R, Parikh J, Sanghi A, Kumar K,
McKinsey J, Lonergan S (eds) Measuring the impact of climate change on Indian agriculture.
World Bank Technical Paper No. 402, Washington, DC
Kumar K, Parikh J (1998b) Climate change impacts on Indian agriculture: The Ricardian
approach. In: Dinar A, Mendelsohn R, Evenson R, Parikh J, Sanghi A, Kumar K, McKinsey J,
Lonergan S (eds) Measuring the impact of climate change on Indian agriculture. World Bank
Technical Paper No. 402, Washington, DC
Kumar P, Joshi PK, Birthal PS (2009) Demand projections for food grains in India. Agric Econ
Res Rev 22(2):237–243
Kumar P, Shinoj P, Raju SS, Anjani K, Msangi S (2010) Agric Econ Res Rev 23:9
Kumbhakar SC (1997) Efficiency estimation with heteroscedasticity in a panel data model. Appl
Econ 29:379–386
Lakshmanan A, Geethalakshmi V, Nagothu US (2009) Azolla and cyanobacterial systems in
supplementing nitrogen to rice besides minimizing methane flux. Technical brief 2,
CLIMARICE: climate change and persistent droughts: impact, vulnerability and adaptation
in rice growing sub-divisions in India. Tamilnadu Agricultural University, Coimbatore
Lal R (2004) Soil carbon sequestration impacts on global climate change and food security.
Science 304:1623–1627
Li Y, Simunek J, Jing L, Zhang Z, Ni L (2014) Evaluation of water movement and water losses in
a direct-seeded-rice field experiment using hydrus-1D. Agric Water Manage 142:38–46
248 References
Levin A, Lin CF, Chu CS (2002) Unit root tests in panel data: asymptotic and finite-sample
properties. J Econometrics 108:1–24
Maddala G, Wu S (1999) A Comparative study of unit root tests and a new simple test. Oxford
Bull Econ Stat 61:631–652
Mahajan G, Chauhan BS, Gill MS (2013) Dry-seeded rice culture in Punjab state of India: lessons
learned from farmers. Field Crops Res 144:89–99
Mall RK, Singh R, Gupta A, Srinivasan G, Rathore LS (2006) Impact of climate change on Indian
agriculture: a review. Clim Change 78(2–4):445–478
McCarl BA, Villavicencio X, Wu M (2008) Climate change and future analysis: is stationarity
dying? Am J Agric Econ 90(5):1241–1247
Mearns LO, Rosenzweig C, Goldberg R (1997) Mean and variance change in climate scenarios:
methods, agricultural applications, and measures of uncertainty. Clim Change 35:367–396
Mehala V et al (2016) Impact of direct seeded rice on economics of paddy crop in Haryana. Int J
Agric Sci 8(62):3525–3528. ISSN 0975-3710, E-ISSN 0975-9107
Mendelsohn R, Dinar A, Dalfelt A (2000) Climate change impacts on African agriculture. School
of Forestry and Environmental Studies, Yale University, Mimeo
Mendelsohn R, Nordhaus WD, Shaw D (1996) The impact of global warming on agriculture:
reply. Am Econ Rev 86:1312–1315
Mendelsohn R, Nordhaus WD, Shaw D (1994) The impact of global warming on agriculture:
A Ricardian analysis. Am Econ Rev 84:753–771
Moss RH, Edmonds JA, Hibbard KA, Manning MR, Rose SK, Van Vuuren DP, Carter TR,
Emori S, Kainuma M, Kram T, Meehl GA, Mitchell JFB, Nakicenovic N, Riahi K, Smith SJ,
Stouffer RJ, Thomson AM, Weyant JP, Wilbanks TJ (2010) The next generation of scenarios
for climate change research and assessment. Nature 463(7282):747–756
Nandhini US, Alagumani T, Shibi S (2006) Economic analysis of agriculture in southern parts of
coastal India. Agricultura Tropica et Subtropica 39(4):279–284
Oliver MMH, Talukder MSU, Ahmed M (2008) Alternate wetting and drying irrigation for rice
cultivation. J Bangladesh Agric Univ 6(2):409–414. ISSN 1810-3030
Paddy cultivation guide: different methods of rice cultivation in India. Retrieved from https://
www.farmingindia.in/paddy-cultivation/
Palanisami K, Paramasivam P, Ranganathan CR, Aggarwal PK, Senthilnathan S (2009)
Quantifying vulnerability and impact of climate change on production of major crops in
Tamil Nadu, India. In: Taniguchi M, Burnett WC, Fukushima Y, Haigh M, Umezawa Y
(eds) Headwaters to the Ocean—hydrological change and watershed management. Taylor and
Francis Group, London, pp 509–551
Palanisami, K, Suresh Kumar D, Malik, RPS, Raman S, Kar G, Kadiri M (2015) Managing water
management research: analysis of four decades of research and outreach programmes in India.
Econ Polit Wkly l(26&27)
Palanisami K, Ranganathan CR, Kakumanu KR, Nagothu US (2011) A hybrid model to quantify
the impact of climate change on agriculture in Godavari river basin. Energy Environ Res 1
(1):32–52
Palanisami K, Ranganathan CR, Nagothu US, Kakumanu Krishna Reddy (2014) Climate change
and agriculture in India. Routledge Taylor and Francis Group, New Delhi. ISBN
978-0-415-73599-5
Palanisami K, Krishna Reddy K, Ranganathan CR, Nagothu US (2015) Farm-level cost of
adaptation and expected cost of uncertainty associated with climate change impacts in Major
River basins in India. Int J Clim Change Strat Manage 7(1):76–96
Pandey MP, Verulkar SB, Sarawgi AK. Status paper on Rice in Chattisgarghl. Retrieved from
http://www.rkmp.co.in
Pathak AR, Dr Mehta AM, Dr. Vashi RD. Retrieved from http://www.rkmp.co.in
Pathak H, Ladha JK, Aggarwal PK, Peng S, Das S, Singh Y, Singh B, Kamra SK, Mishra B,
Sastri ASRAS, Aggarwal HP, Das DK, Gupta RK (2003) Climatic potential and on-farm yield
trends of rice and wheat in the Indo-Gangetic Plains. Field Crops Res 80(3):223–234
References 249
Pathak H, Sankhyan S, Dubey DS, Bhatia A, Jain N (2013) Dry direct-seeding of rice for
mitigating greenhouse gas emission: field experimentation and simulation. Paddy Water
Environ 11:593–601
Pattanayak A, Kumar KSK (2013) Weather sensitivity of rice yield: evidence from India. Working
papers 2013-081. Madras School of Economics, Chennai, India
Pepsico International (2009). Direct seeding of paddy—the work of Pepsico reported in India
water portal. Retrieved from http://www.indiawaterportal.org/post/6754. Accessed on 11 Aug
2013
Prakash SS, Patil CR, Amrutha TJ, Siddayya S (2013) Input use and production pattern of paddy
cultivation under leased-in land in Tungabhadra project area. Karnataka J Agric Sci 26(2):
224–228
Pattanayak A, Kumar KSK (2013) Weather sensitivity of rice yield: evidence from India. Working
papers 2013-081. Madras School of Economics, Chennai. India. Retrieved from https://ideas.
repec.org/p/mad/wpaper/2013-081.html
Rajanna MP (2010) Status paper on rice in Karnataka. Directorate of rice research. Rice
Knowledge Management Portal (RKMP), Hyderabad, India. Retrieved from http://rkmp.co.in/
sites/default/files/ris/rice-statewise/Status%20Paper%20on%20Rice%20in%20Karnataka.pdf
Ranganathan C (2009) Quantifying the impact of climatic change on yields and yield variability of
major crops and optimal land allocation for maximizing food production in different
agro-climatic zones of Tamil Nadu, India: An econometric approach. Working paper, Research
Institute for Humanity and Nature, Kyoto, Japan
Rang A, Mangat GS, Kaur R. Status paper on rice in Punjab. Retrieved from http://www.rkmp.co.in
Rao SK. Status paper on rice in Madhya Pradesh. Retrieved from http://www.rkmp.co.in
Reddy TY, Reddi GH (1995) Principles of agronomy. Kalyani publishers, India
Reiner W, Aulakh MS (2000) The role of rice plants in regulating mechanisms of methane
emissions. Biol Fertil Soils 31:20–29
Saha A, Havenner A, Talpaz H (1997) Stochastic production function estimation: small sample
properties of ML versus FGLS. Appl Econ 29:459–469
Sahana S et al (2017) Adoption and attitude of the farmers on mechanization of paddy in command
areas of Karnataka. Int J Agric Sci 9(29):4382–4384. ISSN 0975-3710, E-ISSN 0975-9107
Sarker MAR, Alam K, Gow J (2012) Exploring the relationship between climate change and rice
yield in Bangladesh: an analysis of time series data. Agric Syst Elsevier 112:11–16
Saseendran SA, Singh KK, Rathore LS, Singh SV, Sinha SK (2000) Effects of climate change on
rice production in the tropical humid climate of Kerala, India. Clim Change 44:495–514
Schlenker W, Roberts MJ (2008) Estimating the impact of climate change on crop yields: the
importance of nonlinear temperature effects. In: NBER working paper 13799
Seo, SN, Mendelsohn R (2008) A structural Ricardian analysis of climate change impacts and
adaptations in African agriculture. In: Policy research working paper No. 4603. World Bank,
Washington, DC. © World Bank. https://openknowledge.worldbank.org/handle/10986/6770.
License: CC BY 3.0 IGO
Shanmugam TR, Mahendran K, UshaNandhini S, Shibi S (2007) Impact assessment of agricultural
research and technology in addressing rice productivity constraints in Tamil Nadu. Project
report submitted to the Department of Agricultural Economics, Tamil Nadu Agricultural
University, Coimbatore 3, India
Shetty PK, Hegde MR, Mahadevappa M (eds) Innovations in rice production. National Institute of
Advanced Studies, Bangalore, pp 1–243
Sperber KR, Annamalai H, Kang I-S, Kitoh A, Moise A, Turner A, Wang B, Zhou T (2013) The
Asian summer monsoon: an intercomparison of CMIP5 versus CMIP3 simulations of the late
20th century. Clim Dyn 41(9–10):2711–2744
Suresh Kumar D, Palanisami K (2010) Impact of drip irrigation on farming system: evidence from
southern India. Agric Econ Res Rev 23:265–272
250 References
Thakur J (1994) Rice production constraints in Bihar (India). Paper presented at the workshop. In:
Thaware BL, Kunkerkar RL, Shivade HA (eds) Status paper on rice in Maharashtra. Retrieved
from http://www.rkmp.co.in
Thiyagarajan K, Kalaiyarasi R. Status paper on rice in Tamil Nadu. Retrieved from http://www.
rkmp.co.in
Technical Program (2011) Technical report of agricultural economics. Regional Agricultural
Research Station, Lam, Guntur, Andhra Pradesh. Unpublished manuscript
Till B, Artner A, Siebert R, Sieber S (2010). Micro-level practices to adapt to climate change for
African small-scale farmers. IFPRI Discussion Paper 00953
UNESCO-WWAP (2009) Climate change and water—an overview from the world water
development report 3: Water in the changing world. A world water assessment programme
special report, Perguia, Italy
Wang JX, Mendelsohn R, Dinar A, Huang JK, Rozelle S, Zhang LJ (2009) The impact of climate
change on China’s agriculture. Agric Econ 40(3):323–337
World Bank (2018) India—Atal Bhujal Yojana (ABHY)—National groundwater management
improvement program (English). Washington, DC
Zainal Z, Shamsudin MN, Mohamed ZA, Adam SU (2012) Economic impact of climate change on
the Malaysian palm oil production. Trends Appl Sci Res 7(10):872–880. https://doi.org/10.
3923/366tasr.2012.872.880

10.1007@978 981 13 8363 2 PDF

Uploaded by

Copyright:

Available Formats

10.1007@978 981 13 8363 2 PDF

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

10.1007@978 981 13 8363 2 PDF

Uploaded by

Copyright:

Available Formats

India Studies in Business and Economics

More information about this series at http://www.springer.com/series/11234

Udaya Sekhar Nagothu C. R. Ranganathan

Climate Change and Future

ISSN 2198-0012 ISSN 2198-0020 (electronic)

Warming of the climate system in recent decades is unequivocal. A warmer climate

To address these issues and the emerging challenges, a collaborative research

New Delhi, India K. Palanisami

1 Climate Change and Agriculture in India . . . . . . . . . . . . . . . . . ... 1

3.4.1 Seasonality of Rainfall . . . . . . . . . . . . . . . . . . . . . . . . 40

5.5 Analysis of the Effect of Climate Change on Rice Yield . . . . . . 96

7.5 Analysis of the Effect of Climate Change on Rice Yield . . . . . . 153

9.5.3 Panel Root Test: Fisher’s Tests . . . . . . . . . . . . . . . . . . 196

K. Palanisami is an agricultural economist and international water resources

Krishna Reddy Kakumanu is an Associate Professor at the Centre for Natural

Udaya Sekhar Nagothu is a Research Professor and Director (Centre for

C. R. Ranganathan is an Adjunct Professor at the Department of Computing,

K. Palanisami, Krishna Reddy Kakumanu, Udaya Sekhar Nagothu

Climate change is a continuous phenomenon, and its variability is a complex phe-

Table 1.1 Projected temperature and precipitation changes in India: 2070–2099

1.2 Climate Change Impacts on Water and Agriculture

Agriculture system in India is subject to severe climatic inter-annual variability,

texture, structure and profile conditions, demands about 1200–2500 mm of water

1.3 Relevance to Policy

1.4 Outline of the Book

D. Suresh Kumar (B)

© Springer Nature Singapore Pte Ltd. 2019 7

2.2 Trend in Area, Production and Productivity

2.2.1 Trend in Area Under Rice

With wide variations in agro-economic conditions and socio-economic diversity, the

Tamil Nadu Karnataka Andhra Pradesh

Fig. 2.1 Trend in area of rice in Southern India

Fig. 2.2 Trend in area of rice in central India

Bihar Orissa West Bengal

Fig. 2.3 Trend in area of rice in eastern India

Fig. 2.4 Trend in area of rice in western India

Punjab UƩar Pradesh Haryana

Fig. 2.5 Trend in area of rice in northern India

2.2.2 Trend in Rice Production

Tamil Nadu Karnataka Andhra Pradesh

Fig. 2.6 Trend in production in southern India

2.2.3 Trend in Rice Productivity

Madhya Pradesh Maharashtra

Fig. 2.7 Trend in production of rice in central India

Bihar Orissa West Bengal

Fig. 2.8 Trend in production of rice in eastern India

Fig. 2.9 Trend in production of rice in western India

Fig. 2.10 Trend in production of rice in northern India

2.3 Constraints in Rice Production

1982-83 1982-83 1982-83

Fig. 2.12 Trend in yield of rice in central India

Fig. 2.13 Trend in yield of rice in eastern India

1992-93 1992-93 1992-93

2004-05 2004-05 2004-05

Fig. 2.14 Trend in yield of rice in western India

Fig. 2.15 Trend in yield of rice in northern India

Table 2.4 Major rice production constraints across states in India

BPH is a dominant pest of rice specifically in coastal Andhra Pradesh. Recently,

2.4 Climate Change and Rice Production