El Niño and positive Indian Ocean Dipole conditions simultaneously reduce the production of multiple cereals across India

Abstract

Natural climate phenomena like El Niño Southern Oscillation (ENSO) and the Indian Ocean Dipole (IOD) influence the Indian monsoon and thereby the region’s agricultural systems. Understanding their influence can provide seasonal predictability of agricultural production metrics to inform decision-making and mitigate potential food security challenges. Here, we analyze the effects of ENSO and IOD on four agricultural production metrics (production, harvested area, irrigated area, and yields) for rice, maize, sorghum, pearl millet, and finger millet across India from 1968 to 2015. El Niños and positive-IODs are associated with simultaneous reductions in the production and yields of multiple crops. Impacts vary considerably by crop and geography. Maize and pearl millet experience large declines in both production and yields when compared to other grains in districts located in the northwest and southern peninsular regions. Associated with warmer and drier conditions during El Niño, >70% of all crop districts experience lower production and yields. Impacts of positive-IODs exhibit relatively more spatial variability. La Niña and negative-IODs are associated with simultaneous increases in all production metrics across the crops, particularly benefiting traditional grains. Variations in impacts of ENSO and IOD on different cereals depend on where they are grown and differences in their sensitivity to climate conditions. We compare production metrics for each crop relative to rice in overlapping rainfed districts to isolate the influence of climate conditions. Maize production and yields experience larger reductions relative to rice, while pearl millet production and yields also experience reductions relative to rice during El Niños and positive-IODs. However, sorghum experiences enhanced production and harvested areas, and finger millet experiences enhanced production and yields. These findings suggest that transitioning from maize and rice to these traditional cereals could lower interannual production variability associated with natural climate variations.

Full text

Environ. Res. Lett. 19 (2024) 104059 https://doi.org/10.1088/1748-9326/ad6a6f
OPEN ACCESS
RECEIVED
1 February 2024
REVISED
12 July 2024
ACC EPT ED FOR PUB LICATI ON
2 August 2024
PUBLISHED
9 September 2024
Original content from
this work may be used
under the terms of the
Creative Commons
Attribution 4.0 licence.
Any further distribution
of this work must
maintain attribution to
the author(s) and the title
of the work, journal
citation and DOI.
LETTER
El Ni˜
no and positive Indian Ocean Dipole conditions
simultaneously reduce the production of multiple cereals across
India
Madhulika Gurazada1,∗, Sonali McDermid2,3, Ruth DeFries4, Kyle F Davis5,6, Jitendra Singh7
and Deepti Singh1
1School of the Environment, Washington State University, Vancouver, WA, United States of America
2NASA Goddard Institute for Space Studies, New York, NY, United States of America
3Department of Environmental Studies, New York University, New York, NY, United States of America
4Department of Ecology, Evolution, and Environmental Biology, Columbia University, New York, NY, United States of America
5Department of Geography and Spatial Sciences, University of Delaware, Newark, DE, United States of America
6Department of Plant and Soil Sciences, University of Delaware, Newark, DE, United States of America
7Institute for Atmospheric and Climate Science, ETH Zurich, Zurich, Switzerland
∗Author to whom any correspondence should be addressed.
E-mail: madhulika.gurazada@wsu.edu
Keywords: natural climate variability, Indian monsoon, climate impacts, agricultural impacts, food security
Supplementary material for this article is available online
Abstract
Natural climate phenomena like El Ni˜
no Southern Oscillation (ENSO) and the Indian Ocean
Dipole (IOD) influence the Indian monsoon and thereby the region’s agricultural systems.
Understanding their influence can provide seasonal predictability of agricultural production
metrics to inform decision-making and mitigate potential food security challenges. Here, we
analyze the effects of ENSO and IOD on four agricultural production metrics (production,
harvested area, irrigated area, and yields) for rice, maize, sorghum, pearl millet, and finger millet
across India from 1968 to 2015. El Ni˜
nos and positive-IODs are associated with simultaneous
reductions in the production and yields of multiple crops. Impacts vary considerably by crop and
geography. Maize and pearl millet experience large declines in both production and yields when
compared to other grains in districts located in the northwest and southern peninsular regions.
Associated with warmer and drier conditions during El Ni˜
no, >70% of all crop districts experience
lower production and yields. Impacts of positive-IODs exhibit relatively more spatial variability. La
Ni˜
na and negative-IODs are associated with simultaneous increases in all production metrics
across the crops, particularly benefiting traditional grains. Variations in impacts of ENSO and IOD
on different cereals depend on where they are grown and differences in their sensitivity to climate
conditions. We compare production metrics for each crop relative to rice in overlapping rainfed
districts to isolate the influence of climate conditions. Maize production and yields experience
larger reductions relative to rice, while pearl millet production and yields also experience
reductions relative to rice during El Ni˜
nos and positive-IODs. However, sorghum experiences
enhanced production and harvested areas, and finger millet experiences enhanced production and
yields. These findings suggest that transitioning from maize and rice to these traditional cereals
could lower interannual production variability associated with natural climate variations.
1. Introduction
Oscillations in tropical sea-surface temperatures drive
variations in global rainfall and temperature pat-
terns that affect food production and livelihoods
worldwide [1,2]. India is particularly vulnerable
to such natural climate variations, as the coun-
try’s agricultural activities and water availability
closely depend on the Indian summer monsoon
rainfall (ISMR). Agriculture is crucial to the Indian
© 2024 The Author(s). Published by IOP Publishing Ltd

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
economy, contributing substantially to employment,
food security, and economic growth. According to
the Indian Economic Association [3], agriculture
employed about 46.5% of the workforce in 2020–21
and constituted one-fifth of India’s economy (Gross
Value Added). Fluctuations in temperature and mon-
soonal rainfall can impact food production, food
prices, agricultural activities, and socioeconomic sta-
bility of the agricultural workforce [4], thereby affect-
ing food security within India and beyond [5–8].
El Ni˜
no Southern Oscillation (ENSO) is a primary
mode of natural variability in Earth’s climate sys-
tem and a major driver of ISMR variability [9]. The
Indian Ocean Dipole (IOD) also influences the ISMR
and modulates monsoon-ENSO teleconnections [10–
13]. Typically, the Indian monsoon experiences below
average rainfall during El Ni˜
no [9] and excess rain-
fall during La Ni˜
na [14]. Positive IOD events typ-
ically enhance the monsoon but are associated with
delayed monsoon onset over southwestern India [15].
Strong El Ni˜
no events often co-occur with positive
IOD (IOD+) events, and recent research suggests
these co-occurrences have become more frequent in
recent decades [16]. However, the impacts of their
co-occurrences are not well understood due to the
relatively short instrumental record. Understanding
how ENSO and IOD influence the summer monsoon
and grains typically grown during the monsoon sea-
son (referred to locally as kharif grains) can inform
the seasonal predictability of agricultural production
and development of early warning systems to manage
food security impacts.
Previous studies have quantified the impact of
ENSO and IOD on the yields and production of cer-
eal grains such as rice, wheat, maize, and soybean
in several regions including Africa, Philippines, and
globally [17–21]. Climate variability explains over
60% of yield variability in major global breadbaskets
[21]. However, a comprehensive evaluation of ENSO
and IOD impacts on grain yields and production at
finer spatial scales in India has not yet been con-
ducted. Previous studies have examined their influ-
ence on national-level yields of rice, maize, wheat,
soybean and sorghum [17,22–24], but studies over
India are limited to certain sub-regions. For instance,
Bhatla et al [25] show that El Ni˜
no negatively affects
rice, maize, pulse, and sugarcane production over the
Indo-Gangetic basin. Nageswararao et al [26], found
that ENSO had a positive influence on wheat in the
Himalayan region, gram in Uttarakhand, rapeseed–
mustard and oilseeds in Uttar Pradesh, Chhattisgarh,
and Rajasthan during October–April. Moreover, the
impacts of climate variability modes on traditional
grains such as sorghum (∼1.3% of total Indian pro-
duction of five major cereals in 2015) and pearl
and finger millets (∼7.3% of total Indian produc-
tion of five major cereals in 2015), have not been
assessed. The combined influence of ENSO and IOD
on grains at the sub-regional scale has also not been
quantified. There is a need to better understand the
individual and combined impacts of ENSO and IOD
on India’s agricultural production—including tradi-
tional staples like sorghum and millets—at the sub-
regional scale for advancing agricultural predictions
and planning.
Our study investigates how these two natural cli-
mate variability modes—ENSO and IOD—influence
production metrics of rice (Paddy), maize (Corn),
sorghum (Jowar), pearl millet (Bajra), and finger mil-
let (Ragi) during the Indian summer monsoon sea-
son. India is the second-largest producer of rice and
the largest producer of millets in the world. There
is growing recognition of the benefits of traditional
cereals [27] that are highly nutritious, less resource-
intensive and more climate resilient than rice [28,
29]. In 2017, National Institution for Transforming
India Aayog, the apex public policy think tank of
the Government of India, released the National
Nutrition Strategy for ‘Nourishing India’, and the
Indian Government has implemented initiatives such
as declaring millets as ‘Nutri-cereals’ and celebrating
2018 as the National Year of Millets to boost mil-
let production and promote its increased inclusion
in diets. The United Nations declared 2023 as the
‘International Year of Millets’ to raise awareness about
its health benefits.
This study specifically aims to (a) quantify the
impact of ENSO and IOD on the production, har-
vested area, irrigated area, and yields of rice, maize,
sorghum, and millets, (b) understand the spatial vari-
ations in these impacts across India in the context of
rainfall and temperature variations, and (c) evaluate
the sensitivity of maize and traditional grains relat-
ive to rice. Examining multiple production metrics
provides a more comprehensive perspective, as most
previous literature has focused predominantly on
yields. Further, considering harvested areas alongside
yields better captures the overall production dynam-
ics by accounting for changes in the spatial extent of
crop cultivation in response to climate fluctuations.
Our findings are relevant for assessing the resiliency
of agriculture to natural climate fluctuations, inform-
ing seasonal predictability of crop production, and
understanding cascading risks to other allied sectors
such as public health and food security.
2. Data and methods
2.1. Summer monsoon grain and soil data
District-level crop production metrics (production,
harvested area, irrigated areas, and yields) for 1966–
2017 are obtained from the International Crops
Research Institute for the Semi-Arid Tropics Village
Dynamics of South Asia (ICRISAT-VDSA) dataset
(http://data.icrisat.org/dld/src/crops.html) [30]. Our
study focuses on the production metrics of five grains
during the summer monsoon (or kharif) season
(June–September, JJAS): rice, maize, sorghum, pearl
2

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
millet, and finger millet. While exact planting and
harvest dates vary across districts and crop varieties
within India, June–September broadly captures the
primary growing season for all five crops [31,32].
Yields in the VDSA data are reported as the ratio
of production to harvested area. In some districts
and years, reported irrigated areas exceeded harvested
areas, indicating that harvested areas could be smaller
than planted areas for which data is unavailable.
District-level soil type data are also obtained
from ICRISAT-VDSA (http://data.icrisat.org/dld/src/
crops.html) [30]. Each district is characterized by at
least one of three soil types: a primary, secondary,
and tertiary, representing the highest to lowest per-
centage in that district. We focus on seven soil types
categorized according to the USDA soil taxonomy—
Alfisols, Alfisols-Mollisols-Mix, Alfisols-Inceptisols-
Mix, Vertisols, Aridisols, Inceptisols, and Entisols.
2.2. Climate data
For ENSO, we use the Ni˜
no3.4 index that captures
sea surface temperatures (SST) in the central equat-
orial Pacific. Ni˜
no3.4 SSTs are closely related to
Indian monsoon variability [26]. For IOD, we use the
Dipole Mode Index (DMI), a measure of the differ-
ence in SSTs between the western and eastern Indian
Ocean regions [33]. Ni˜
no3.4 index and DMI timeser-
ies are obtained from NOAA Earth System Research
Laboratory Physical Sciences Division (https://psl.
noaa.gov/gcos_wgsp/Timeseries/) [34–37].
Daily gridded rainfall dataset is from the Indian
Meteorological Department (0.25◦×0.25◦; 1968–
2015) (https://www.imdpune.gov.in/cmpg/Griddata/
Rainfall_25_NetCDF.html) [38,39] and monthly
maximum temperature dataset is from Climatic
Research Unit, CRU TS4.06 (0.5◦×0.5◦; 1968–2015)
(https://crudata.uea.ac.uk/cru/data/hrg/cru_ts_4.06/
cruts.2205201912.v4.06/) [40]. We select these data-
sets for their high spatial resolution and the length
of the record overlapping with agricultural data
(1968–2015).
2.3. Calculating anomalies
Following Iizumi et al [41], we remove the 5-year
moving averages from the seasonal values of each
production metric at the district-level to account
for long-term trends driven by advances in techno-
logy, management, and other external factors. The
choice of the moving average window influences the
trend estimate and the number of datapoints in our
timeseries; a longer window would reduce the num-
ber of years for analysis [41]. These detrended val-
ues, referred to as absolute seasonal anomalies, rep-
resent interannual variations likely associated with
interannual climate variations rather than long-term
trends. Percent anomalies of production metrics are
calculated as the ratio of seasonal anomalies to their
respective 5-year moving averages. We limit our
analysis to districts with at least 40 years of data to
avoid spurious anomalies due to small sample sizes.
We linearly detrend and standardize the JJAS
Ni˜
no3.4 Index, DMI, and the total seasonal rainfall
and average seasonal maximum temperature by sub-
tracting the climatological (1981–2010) mean and
dividing by the standard deviation [42,43]. We define
El Ni˜
no/La Ni˜
na and IOD+/IOD−years based on the
standardized indices exceeding ±0.5 standard devi-
ations (σ) (table S1 in the supplementary material).
2.4. Quantifying the influence of climate modes on
crop production metrics and climate anomalies
To compare the influence of ENSO and IOD phases
at the national-aggregate level, we calculate weighted
averages of anomalies of each production metric
across all districts growing each crop. This approach
ensures that districts with more cropland or higher
production contribute proportionally more to the
overall national-average anomalies. We also exam-
ine the spatial patterns of crop production met-
ric anomalies and climate anomalies using compos-
ite maps during six conditions: (a) El Ni˜
no, (b)
La Ni˜
na, (c) IOD+, (d) IOD−, (e) co-occurring El
Ni˜
no and IOD+, and (f) co-occurring La Ni˜
na and
IOD−. Co-occurrences of El Ni˜
no and IOD−(2004)
and La Ni˜
na and IOD+(2008 and 2011) are rare
and excluded from the analysis. The fraction of dis-
tricts experiencing local deficits (anomalies <0) or
excesses (anomalies >0) in each production metric
are calculated.
Production impacts aggregated to the national-
level reflect multiple factors, including different cli-
mate conditions across various districts in India and
their relative sensitivity to climate anomalies. To
evaluate the relative sensitivity of different pairs of
crops, we compare production metrics under sim-
ilar climate conditions and only in overlapping rain-
fed districts growing both crops. We only consider
rainfed districts for this comparison as irrigation can
buffer climate impacts. Rainfed districts are identi-
fied based on the proportion of irrigated to harves-
ted areas being below 0.5 for each crop. Rice is the
most irrigated crop across districts, while other crops
are mostly rainfed (figure S1). This threshold of 0.5
is somewhat arbitrary but ensures a sufficiently large
number of rainfed districts growing both rice and
alternative grains. We note that production metrics
are available at the district-level. Within-district vari-
ations in climate conditions, soil types, and topo-
graphic effects could also affect the relative sensitivity
that we are unable to assess here.
3. Results
3.1. Historical trends in production metrics
The Indian agricultural landscape has changed since
the start of the Green revolution in the 1960s. Figure 1
shows trends in production, harvested area, irrigated
area, and overall yield for the five grains grown during
the Indian monsoon season, India’s primary growing
3

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
Figure 1. Varying trends in production and harvested area of rice, maize and traditional grains despite increases in yields:
timeseries (1966–2017) of (a) production (million tons), (b) harvested area (million hectares), (c) irrigated area (million
hectares), (d) area-weighted yields (tons/hectare) for five cereals grown in India during the monsoon season, and (e) district maps
of average crop yields from 1966–2017. Grey districts indicate districts with <40 years of yield data.
season. Rice has remained the dominant grain and
its production quadrupled, harvested area increased,
and irrigated area more than doubled since the 1960s
(figures 1(a)–(c)). Rice yields more than doubled
(figure 1(d)) through increased fertilizer use and
irrigation expansion during the Green Revolution.
Maize yields also doubled over this period, surpass-
ing rice yields in the early 2000s, through improve-
ments in technology and agronomic practices to meet
the growing demand as food and feed for livestock
[44]. Yields of millets and sorghum are substan-
tially lower than that of rice and maize and have
increased at a substantially slower pace, reflecting
a shift towards prioritization of the cultivation of
rice and maize and a shift in diets away from tradi-
tional, nutrient-dense grains (figure 1(d)). This is also
reflected in the decline in production and harvested
area of sorghum and finger millet production since
the 1960s (figures 1(a) and (b)). Rice yields are typ-
ically highest in the Indian breadbasket region of
Punjab, while maize yields are the highest in south-
eastern India (figure 1(e)). Sorghum yields are highest
in the arid and semi-arid parts of central and peninsu-
lar India (figure 1(e)). The highest pearl millet yields
are found in western and southeastern India and the
Indo-Gangetic plains, while finger millets thrive in
the southern peninsular regions (figure 1(e)).
3.2. Influence of ENSO and IOD on
nationally-aggregated production metrics of kharif
grains
Interannual variability in total kharif production,
harvested area, irrigated area, and yields of all five
grains are associated with ENSO and IOD-driven
4

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
Figure 2. El Ni˜
no decreases national-average crop yields: Weighted average district-level anomalies of kharif (a) production
(1000 tons), (b) harvested area (1000 ha), (c) irrigated area (1000 ha), and (d) yields (kg/hectares) during opposite phases of each
mode. The red (blue) dots indicate declines (increases) of <−0.5σ(>0.5σ) in each characteristic. (e) Number of years with
declines (<−0.5σ) in red (increases >0.5σin blue) during each mode phase. Numbers in brackets in the first column indicate
numbers of years with each phase.
climate fluctuations (figure 2). We use standardized
anomalies to evaluate the extent of interannual vari-
ations in these production metrics [45]. We find that
El Ni˜
nos are most strongly associated with declines in
harvested area, irrigated area, and yield (figures 2(b)–
(d)). For instance, 8 of the 14 El Ni˜
no years between
1966 and 2017 coincide with negative yield anomalies
exceeding −0.5σ. A majority of years with reduced
(<−0.5σ) production (6 of 12), harvested area (6
of 12), irrigated area (6 of 10), and yields (8 of
15) occur during El Ni˜
no years. Alternatively, La
Ni˜
nas are typically associated with average or higher
than average production, harvested area, irrigated
area, and yields (figure 2(e)). Notably, half of the La
Ni˜
na years have higher than average (>0.5σ) total
production.
Similarly, IOD also influences national-level pro-
duction metrics (figure 2). IOD+years are more
often associated with below average harvested area,
irrigated area, and yields. However, there are an
equal number of years with declines and increases
in production. In contrast, IOD−years are largely
associated with increases in all production metrics.
It is also notable that the magnitude and frequency
of yield declines during El Ni˜
no and IOD+years
are higher than the increases during La Ni˜
na or
IOD−. For instance, yield declines during three recent
El Ni˜
no years—2002, 2009, and 2015—exceeded
100 kg ha−1whereas yield increases during only
one La Ni˜
na year—2007 exceeded 100 kg ha−1
(figure 2(d)). Declines in these production metrics
during El Ni˜
no or IOD+years highlight the risk of
5

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
Figure 3. Simultaneous declines in multiple crops and grains likely to reduce during El Ni˜
no and positive IOD events: (a)
weighted seasonal yield anomalies for each crop. Dots in panel (a) indicate occurrences of each mode phase. Number of grains
with simultaneous declines (anomalies <0; brown) and increases (anomalies >0; green) in absolute (b) production, (c) harvested
area, (d) irrigated area, and (e) yield during ENSO and IOD phases.
food shortages and insecurity, while increases during
La Ni˜
na or IOD−years might not effectively com-
pensate for shortages.
El Ni˜
no or IOD+conditions reduce the yields of
multiple grains simultaneously (figures 3and S2). For
instance, 10 of 14 El Ni˜
no years and 9 of 13 IOD+
years experience simultaneous reductions in the
yields of three or more grains (figure 3(a)). During
El Ni˜
no or IOD+years, a median of four grains
experience simultaneously reduced yields, whereas a
median of two grains experience reduced yields dur-
ing La Ni˜
na or IOD−years (figure 3(e)). For instance,
rice, maize, sorghum and pearl millets experienced
markedly reduced yields during the strong 2015 El
Ni˜
no. Conversely, during La Ni˜
na or IOD−years,
a median of three grains experience simultaneously
higher yields (figure 3(e)). These results indicate that
El Ni˜
no or IOD+events have a relatively stronger
and consistent negative impact on the yields of mul-
tiple grains than La Ni˜
na or IOD−events. In contrast,
La Ni˜
na or IOD−events are more consistently asso-
ciated with higher production, harvested, and irrig-
ated area of multiple crops than the corresponding
declines during El Ni˜
no or IOD+years (figures 3(b)–
(d)).
Compared to yields, production, harvested area,
and irrigated area (figures 3(b)–(d)) are more likely
to experience simultaneous increases than declines
during ENSO and IOD events. This suggests that
adverse climate conditions are less likely to negatively
affect these production characteristics of multiple
crops simultaneously and that farmers might adopt
strategies to minimize production losses during con-
ditions that are likely to negatively impact yields of
multiple crops. For instance, during El Ni˜
no years,
there is a higher likelihood of simultaneous declines
in yields and production than increases but more
crops are likely to experience simultaneous increases
in harvested and irrigated area. This could indic-
ate that farmers increase the area they plant and
irrigate in anticipation of El Ni˜
no. Since El Ni˜
nos
often bode weaker monsoons in India, early warnings
and interventions are often in place to help farmers
mitigate these impacts [46]. Farmers might employ
strategies to mitigate the negative impacts of weaker
monsoonal rains, such as adjusting sowing times
and spacing, adopting new sowing techniques, imple-
menting soil conservation measures, and improving
pest, water, and livestock management. They also
shift to climate-ready crop varieties, change crop sys-
tems, relocate agricultural fields, and modify policy
structures [46]. During La Ni˜
na and IOD−condi-
tions, a combination of higher yields and more har-
vested and irrigated area for multiple crops likely
contributes to greater increases in overall produc-
tion of multiple crops (figure 3). This explains why
we see more simultaneous increases in production
than in yields (figures (b) and (e)). It is also pos-
sible that yields are more directly sensitive to cli-
mate conditions, while anticipated production losses
can be mitigated by adjusting planted or irrigated
area.
6

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
The magnitude of impacts of ENSO and IOD
events vary by crop and also vary depending on
whether we evaluate absolute anomalies in various
metrics or percent anomalies relative to their averages
because of the substantial differences in average pro-
duction metrics of these crops. In absolute and per-
cent terms, we observe that the median nationally-
aggregated production, harvested area, and yield
declines are largest for pearl millets and maize during
El Ni˜
no (figure S3). During La Ni˜
na and IOD−years,
pearl millets experience the largest absolute increases
in harvested areas while rice experiences the largest
increases in production, and finger millets experience
the largest increase in yields. We also find that percent
anomalies in production metrics of traditional grains
during all ENSO and IOD phases experiences larger
variations than that of rice and maize.
Overall, the findings suggest that at the
nationally-aggregated level, the production metrics of
traditional grains such as pearl millet experience lar-
ger fluctutations relative to rice in response to ENSO
and IOD. Rice is the main irrigated crop across many
districts (figure S1) while other grains have limited
irrigation (figure S4), highlighting the importance of
irrigation in buffering crop production from climate
variations. Further, the potential for multiple grains
to be negatively impacted simultaneously under-
scores the agricultural risks posed by natural climate
variations.
3.3. Spatial differences in production metrics and
climate anomalies during ENSO and IOD events
The impacts of ENSO and IOD exhibit substantial
heterogeneity across districts likely associated with
differences in soil types, climate conditions, topo-
graphy, and management practices such as irrigation,
pesticide and fertilizer application. To examine these
spatial variations, we analyze district-level anomalies
in production metrics during El Ni˜
no and IOD+
events, as these phases demonstrate the most sub-
stantial negative impacts (figures 4, S5–8). An under-
standing of the geographical variations in impacts can
help identify the grains and districts most likely to be
affected to inform planning, preparedness, and agri-
cultural decision-making.
Figure 4shows the spatial variations in yield
impacts of all five grains. El Ni˜
no events result
in the largest and most widespread yield reduc-
tions for maize in the climatological drier north-
western, eastern, and semi-arid peninsular regions
of India, where several districts experience negat-
ive anomalies exceeding 200 kg ha−1(figure 4(d)).
Approximately 80.7% of maize-producing districts
experience decreases in maize yields during El Ni˜
no
(table 1). The most substantial yield reductions
occur in districts that experience anomalously drier
and hotter conditions linked to El Ni˜
no events
(figure 5(a)–(b)) that can lead to water stress and soil
moisture deficits associated with reduced photosyn-
thesis, stunted growth, and lower crop yields [47].
Over 61% of maize-producing districts are associ-
ated with hot conditions and dry conditions during
El Ni˜
no (figure 5(c)). Similarly, 70.4% of the rice-
growing districts show declines in rice yields during El
Ni˜
no conditions (table 1), with the largest reductions
in arid and semi-arid districts (figure 4(a)). Most dis-
tricts in the Indo-Gangetic basin and southeastern
India show limited impacts for most crops during El
Ni˜
no (figure 4), likely because of the high rates of
irrigation in those regions that can minimize rainfall
deficits.
Previous research suggests that sorghum and mil-
lets demonstrate greater tolerance to hot and dry
conditions [48]. Our results support this finding —
figure 4shows that the magnitude of absolute yield
anomalies for these traditional rainfed grains is sim-
ilar to or lower than that for irrigated crops like
rice and maize in arid and semi-arid regions during
El Ni˜
nos. However, a higher percentage of districts
cultivating traditional grains experience yield reduc-
tions compared to rice districts during El Ni˜
nos (table
1). Specifically, 85.6% of sorghum districts, 79.3%
of pearl millet districts, and 73.1% of finger millet
districts experience yield reductions (table 1). The
largest declines in sorghum and finger millet yields
occur in districts in peninsular India (figure 4(g) and
(m)), while the largest declines in pearl millet yields
are in northwestern India (figure 4(j)). The relatively
smaller fraction of districts affected for finger mil-
let could be because it is well-suited for cultivation
in the cooler, hilly regions of the western and east-
ern ghats, where it could benefit from anomalously
warm conditions (figure 5(b)). In addition, millets
in the southernmost part of India experience yield
increases rather than declines during El Ni˜
nos and
IOD. Overall, these results suggest that while the mag-
nitude of yield reductions are typically smaller for
traditional grains relative to rice and maize, El Ni˜
no
could adversely affect a larger fraction of districts
growing traditional grains.
During El Ni˜
no events, maize also sees the
largest number of districts with production (79.8%
districts), harvested area (69.8% districts), and
yield (80.7% districts) declines (table 1). The
most substantial absolute reductions in production
(>50,000 tons) and harvested area (>20,000 ha) are
seen for rice in East India and pearl millet in arid
regions (figures S7 and S8), where they are primar-
ily cultivated. The absolute reductions in production
and harvested area for sorghum and finger millet are
not as substantial as those for other grains. Maize,
on the other hand, experiences widespread but relat-
ively small reductions in both production (∼79.8%
of districts) and harvested areas (∼69.8% of districts)
(table 1, figures S7 and S8).
Relative to El Ni˜
no, IOD+events affect fewer dis-
tricts. The absolute yield anomalies during IOD+
7

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
Figure 4. Spatial differences in yield impact anomalies during El Ni˜
no and positive IOD+events: composite maps of yield
anomalies for each crop (a–c) rice, (d–f) maize, (g–i) sorghum, (j–l) pearl millet, and (m–o) finger millet during (left) El Ni˜
no,
(center) positive IOD, and (right) co-occurring El Ni˜
no & positive IOD years.
Table 1. Percent districts with below average (anomalies <0) production, harvested area, and yield during each phase.
Production Harvested area Yield
% Districts
with loss El Ni˜
no IOD+
El Ni˜
no
& IOD+El Ni˜
no IOD+
El Ni˜
no
& IOD+El Ni˜
no IOD+
El Ni˜
no
& IOD+
Rice 74.5% 71.9% 79.8% 66.7% 70% 70.2% 70.4% 64.4% 78.6%
Maize 79.8% 79.4% 81.9% 69.8% 68.7% 73.3% 80.7% 73.7% 80.2%
Sorghum 75.6% 78.1% 83.6% 60.7% 68.2% 68.2% 85.6% 69.7% 80.1%
Pearl millet 81.1% 67.7% 84.8% 70.1% 71.3% 72% 79.3% 62.8% 76.8%
Finger millet 82.1% 88.5% 85.9% 59% 67.9% 75.6% 73.1% 70.5% 71.8%
events show a dipole pattern across several grains,
particularly for rice, sorghum, and pearl millet
(figure 4). Districts across the Indo-Gangetic basin
observe yield increases due to relatively wetter and
cooler conditions (figures 4and 5), while districts
in the southern semi-arid regions that experience
drier and hotter conditions observe yield reduc-
tions (figures 4and 5). Over at least 62% of grain-
producing districts are affected by negative yields in
IOD+years (table 1). During IOD+, pearl millet
has the smallest percent of districts with production
(67.7%) and yield (62.8%) declines and finger mil-
let has the smallest percent of districts with harvested
area (67.9%) declines compared to other grains (table
1). During IOD+events, there is an average reduc-
tion of ∼100,000 tons in the absolute rice production
in the eastern districts, which are the primary rice-
producing regions (figure S7). IOD+events are also
associated with reduction in the absolute harvested
areas of over 10,000 ha for rice in the eastern districts,
8

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
Figure 5. Spatial patterns of climate conditions during El Ni˜
no and positive IOD+events: composite maps of anomalies in (a)
total seasonal rainfall, and (b) maximum temperature during El Ni˜
no, positive IOD, and co-occurring El Ni˜
no & positive IOD
years. The inset histograms illustrate the number of districts experiencing dry or wet (panel a) and hot or cool (panel b)
conditions. (c) Fraction of districts experiencing dry and hot conditions for each phase and crop.
and pearl millet in the arid regions where pearl millets
are primarily cultivated (figure S8).
La Ni˜
na conditions are generally linked to wetter
and cooler climates (figure S9), leading to a major-
ity of districts experiencing increases in production
metrics (table S2). During La Ni˜
na, finger millet
sees the largest percent of districts with production
(>75.6%) and yield (>79.5%) increases, while rice
sees the largest percent of districts with harvested area
(>63.7%) increases compared to other grains (table
S2). Furthermore, while IOD−conditions tend to be
associated with less wet and cool conditions com-
pared to La Ni˜
na, they lead to smaller increases in
production and yields but a larger percent of districts
experience production and yield increases across all
grains except finger millet (table S2, figure S5). IOD−
conditions are associated with smaller increases in
rice and pearl millet production and harvested
area compared to La Ni˜
na, in the primary regions
growing those grains (figures S7–S8). Yield anom-
alies during IOD−show similar spatial patterns to
La Ni˜
na (figure S5).
In addition to absolute yield anomalies, we also
consider percent yield anomalies to compare the rel-
ative impact of climate variability on each crop’s
expected yield, independent of baseline production
levels. While the absolute yield declines in tradi-
tional grains across their primary growing districts
are modest, the relative yield declines are more pro-
nounced in these districts during ENSO and IOD
phases (figures 4and S6). Maize and rice experience
the most widespread reductions, with ∼72% of dis-
tricts seeing percent yield declines during El Ni˜
no
(figure S6 and table S3). Sorghum and pearl mil-
let percent yield declines are comparable to maize
percent yield declines in the northwest, with about
20% reduction during El Ni˜
no (figure S6). During
IOD+events, more districts experience percent yield
9

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
declines in pearl millet (∼63.4% of districts) and
finger millet (∼79.5% of districts), primarily in the
southern peninsular region (table S3). Rice, pearl
and finger millet experience the most widespread
increases in percent yields, with more than 64% of
districts experiencing yield increases during La Ni˜
na
and IOD−(table S3). All crops experience declines
and increases in percent yield in similar regions and
roughly to the same magnitude during ENSO and
IOD events (figure S6).
3.4. Influence of co-occurring ENSO and IOD
events
Since El Ni˜
no and IOD+events individually have sub-
stantial negative impacts on most production met-
rics studied here (figure 4), and their co-occurrences
are increasing [16], we examine the influence of
their co-occurrences on all production metrics. In
our study period, there are 7 co-occurring El Ni˜
no
and IOD+years and 6 co-occurring La Ni˜
na and
IOD−years (table S1). Most co-occurring El Ni˜
no
and IOD+years are associated with overall yield
declines except 1982, with the greatest yield declines
(>80 kg ha−1) in 1972, 1987, 1997, and 2015 (figure
S10(d)). Most co-occurring El Ni˜
no and IOD+years
are also associated with declines in other produc-
tion metrics (figure S10(a)–(c)). For instance, over-
all production declined by 15,000 tons in 1982,
1987, 1997, and 2015 (figure S10(a)) and harves-
ted and irrigated areas declined by 10,000 ha in
1982 and 1987 (figure S10(b)–(c)). In contrast, all
co-occurring La Ni˜
na and IOD−years are associ-
ated with increases in absolute overall production
(>15,000 tons) (figure S10(a)). Most co-occurring
La Ni˜
na and IOD−years are also associated with
increases in harvested areas and yields (except 1971
and 1985) (figure S10 (b) and (d)).
Co-occurring El Ni˜
no and IOD+years lead to
more severe and widespread reductions in absolute
production and harvested area for all grains except
finger millet, relative to the impacts during individual
events (figures S7–8 and table 1). This is likely because
co-occurring El Ni˜
no and IOD+conditions are asso-
ciated with stronger drier and hotter conditions in
the Indo-Gangetic and peninsular India, and stronger
wetter and hotter conditions in central and eastern
India than during the occurrences of the individual
modes (figure 5(a)–(b)). These intensified conditions
result in larger reductions in rice yields and produc-
tion compared to traditional grains (figures 4and S7)
and also affect the harvested areas of rice and pearl
millet (figure S8). During co-occurring El Ni˜
no and
IOD+years, over 71% of districts experience reduc-
tions in absolute production and yields for all grains
(table 1). However, during co-occurring El Ni˜
no
and IOD+years, absolute yield reductions are more
severe but less widespread for most grains except
rice, while percent yield declines for rice, maize, and
pearl millet are more widespread compared to the
impacts observed during El Ni˜
no alone (figure 4and
table 1). Similarly, co-occurring La Ni˜
na and IOD−
years are associated with stronger but less widespread
increases in absolute and relative production met-
rics for most grains corresponding to the stronger
wetter and more widespread cooler conditions across
India than individual La Ni˜
na and IOD−years
(figures S5–9 and tables S2–3). All crops show sim-
ilar regional patterns and magnitudes of percent yield
changes during co-occurring ENSO and IOD events
(figure S6).
3.5. Relative sensitivity of grains to ENSO and IOD
in rainfed areas
To understand the spatial patterns of crop impacts
under similar climate conditions and their relative
differences among crops, we analyze the sensitivity of
alternative grains to rice, the dominant kharif grain
in India, during ENSO and IOD events in predom-
inantly rainfed districts. We find that there is greater
yield variability during ENSO and IOD conditions in
rainfed districts than irrigated districts (figure S4).
This greater variability indicates that rainfed districts
are more susceptible to the erratic nature of ENSO
and IOD events. Therefore, we compare the sensitiv-
ity of maize, sorghum, pearl millet, and finger mil-
let relative to rice, during individual ENSO and IOD
phases in overlapping rainfed districts where both rice
and the alternative grains are grown (figure 6). This
approach allows us to discern the impacts on pro-
duction metrics of different crops under similar rain-
fall and temperature conditions. We define sensitivity
as the median anomaly in production metrics influ-
enced by natural climate variations associated with
ENSO and IOD. We focus the discussion here on the
sensitivity to ENSO but figures 6and S11 also show
the sensitivity to IOD.
In predominantly rainfed districts that grow both
rice and maize, the median negative production and
yield anomalies of maize are larger than rice dur-
ing El Ni˜
no conditions (figure 6). We find similar
results comparing the median production and yield
anomalies of rice and pearl millets in overlapping dis-
tricts, suggesting that maize and pearl millets have
higher sensitivity than rice. In contrast, comparing
the median production and harvested area anom-
alies of rice and sorghum and rice and finger mil-
lets in overlapping districts, suggests that sorghum
has lower sensitivity than rice and finger millets
are equally or less sensitive than rice (figures 6(a)–
(c)). Contrastingly, during La Ni˜
nas, the median
rice production and yield increases surpass those
of other grains. These findings remain consistent
even when evaluating sensitivity in terms of percent
median anomalies in production and harvested areas.
However, median percent production, harvested area,
and yields of sorghum and finger millet surpass rice
in overlapping rainfed districts during El Ni˜
no. This
implies that, given the lower climate sensitivity of
10

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
Figure 6. Comparison of the sensitivity of alternative grains to climate variability relative to rice: boxplots of absolute
anomalies for (a) production, (b) harvested area, and (c) yield for rice and each alternative grain during ENSO and IOD phases in
overlapping rainfed districts. For instance, the first set of boxplots represent weighted anomalies for rice and maize only in the
subset of districts growing both grains. The corresponding inset maps show overlapping rainfed districts for rice and the
alternative grain in purple. Vertical lines separate boxplots for each pair of grains. Note: rainfed districts are defined as districts
with irrigated area fraction of harvested areas below 50%.
sorghum and finger millet compared to rice, shifting
investments from rice cultivation to sorghum and fin-
ger millet could reduce the sensitivity of grain pro-
duction to natural climate fluctuations.
4. Discussion
Our analysis examines the influence of ENSO,
IOD and their co-occurrences on four produc-
tion metrics—yields, production, harvested areas,
and irrigated areas—for five key monsoon cer-
eals across India and their geographic variations.
Our key findings are [1] El Ni˜
no and IOD+are
associated with reductions in national-level yield,
production, harvested area, and irrigated area of
kharif grains, whereas La Ni˜
nas and IOD−are
associated with increases; [2] El Ni˜
no and IOD+
simultaneously reduce yields for a median of four
grains while La Ni˜
na and IOD- simultaneously
enhance yields for a median of three grains; [3] co-
occurring El Ni˜
no and IOD+years are associated
with stronger and more widespread reductions in
rice yields than alternative grains compared to the
impacts of individual El Ni˜
no and IOD+events;
and [4] Compared to rice, maize and pearl mil-
let production and yields have higher sensitivity to
11

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
similar climate variations, while sorghum and fin-
ger millet have lower sensitivity to El Ni˜
nos than
rice.
Through this analysis, we make several unique
contributions to the literature. First, we document
the effects of ENSO and IOD on relatively under-
studied characteristics such as harvested and irrig-
ated areas, providing a more comprehensive under-
standing of grain production dynamics. Second, we
assess the spatial pattern of production metrics to gain
insights into the geographical distribution of impacts
and identify specific sub-regions and grains that are
the most and least sensitive to ENSO and IOD-driven
climate variations. Such spatial information can help
agricultural planners and farmers prioritize and tailor
interventions to mitigate climate-related risks for spe-
cific grains in specific regions. Third, we character-
ize the yield response of the 5 grains to individual
and co-occurring ENSO and IOD events, which has
not yet been done. Lastly, we compare the sensitiv-
ity of traditional grains relative to rice to evaluate
their resilience to natural climate variations, which
can inform recent efforts to incentivize the produc-
tion and consumption of these nutrious cereals.
Our results are consistent with Bhatla et al [25]
and Pandey et al [49], who found declines in rice
and maize production during El Ni˜
no events in parts
of the Indo-Gangetic basin, and Cherian et al [50],
who found that El Ni˜
no events reduce rice produc-
tion in Karnataka (peninsular India) due to below-
normal rainfall. Our finding that ∼81% of maize-
producing districts experience yield reductions dur-
ing El Ni˜
no and co-occurring El Ni˜
no & IOD+
years helps contextualize the results of Anderson
et al [17], which shows that ENSO accounts for 25%
of maize production variability in India. Consistent
with Davis et al [28], we find that sorghum and
finger millet are less sensitive to climate variations
compared to rice. Differences between our stud-
ies arise because our analysis examines the average
impacts from large-scale climate fluctuations driven
by ENSO and IOD while Davis et al [28] define
sensitivity to the locally hottest and driest years.
Furthermore, consistent with our finding, Heino et al
[23] show that irrigation has reduced the effects of
natural climate variability on global yields of several
crops.
We note a few caveats of this study. Although we
used the longest available dataset of crop data (1966–
2017), the relatively short time period limits the
sample size of ENSO and IOD events. Additionally,
we define predominantly rainfed areas as districts
with the proportion of irrigated to harvested areas
<0.5, which is an arbitrary threshold aiming to cap-
ture a sufficient number of rainfed districts grow-
ing both rice and alternative grains for the sensit-
ivity analysis. However, the rainfed versus irrigated
yields and production of each grain are not explicitly
available, limiting our ability to characterize the effect
of irrigation. Finally, we use district-level climate
and yield data, which may not capture finer-scale
variations in temperature and precipitation within a
district, limiting the accuracy of climate-crop produc-
tion relationships.
Overall, our study provides geographically expli-
cit information on how key cereals are affected by
two key modes of natural climate variability that
are a source of seasonal predictability [51,52]. This
knowledge can inform short-term planting decisions
and management interventions to minimize negat-
ive impacts on agricultural production and farmer
livelihoods when ENSO or IOD events are fore-
cast. Studies have indicated that ENSO events can
affect how forecasts influence farmer decision mak-
ing. For instance, Maggio and Sitko [53] found
that seasonal forecast information could influence
farmers’ adaptive practices in response to fore-
casted drought. However, Guido et al [54] high-
lighted the challenges farmers face when forming
expectations about upcoming seasonal rainfall, par-
ticularly when these expectations are disconnected
with observed trends in climate data. These stud-
ies emphasize the significance of providing local cli-
mate information to empower farmers to make well-
informed agricultural decisions. Reducing sensitiv-
ity of agricultural production to such known climate
variations can help better prepare for and minim-
ize food security risks from climate variability and
change.
Data and materials availability
The crop and soil data, climate data, and climate
variability indices used in this study are available
for download at http://data.icrisat.org/dld/src/crops.
html;https://www.imdpune.gov.in/cmpg/Griddata/
Rainfall_25_NetCDF.html;https://crudata.uea.ac.
uk/cru/data/hrg/cru_ts_4.06/cruts.2205201912.v4.
06/, https://psl.noaa.gov/gcos_wgsp/Timeseries/. The
source code used to perform the analyses and gen-
erate the publication figures can be accessed at a
dedicated GitHub repository maintained by the cor-
responding author: https://github.com/madhulikag/
climatevariability_monsooncrops_india. All data
necessary to evaluate the conclusions of this study
are provided within the paper and/or in the
Supplementary Materials. The data supporting the
findings of this study are openly available at the fol-
lowing URL/DOI: https://github.com/madhulikag/
climatevariability_monsooncrops_india.
Acknowledgments
We acknowledge the International Crops Research
Institute for the Semi-Arid Tropics (ICRISAT) Village
Dynamics in South Asia for providing mesolevel crop
and soil data for India. We also thank the Indian
Meteorological Department, Climatic Research Unit,
12

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
and NOAA PSL for providing public access to the
climate data and natural climate variability indices
used in this study. We acknowledge support from the
National Science Foundation (NSF AGS-1934383)
and from Washington State University.
Funding
M G and D S were supported by Washington State
University and NSF award AGS-1934383.
Conflict of interest
The authors declare that they have no competing
interests.
Author contributions
All authors contributed to designing the research. M
G conducted the analyses. M G and D S drafted the
paper with feedback from all coauthors.
ORCID iDs
Madhulika Gurazada https://orcid.org/0009-
0000-4215-4583
Sonali McDermid https://orcid.org/0000-0002-
4244-772X
Ruth DeFries https://orcid.org/0000-0002-3332-
4621
Kyle F Davis https://orcid.org/0000-0003-4504-
1407
Jitendra Singh https://orcid.org/0000-0001-9132-
0269
Deepti Singh https://orcid.org/0000-0001-6568-
435X
References
[1] Kakoti M, Paul A, Sarma H H and Sonowal S 2023 El Ni˜
no
Southern oscillation (ENSO): an overview on Indian
monsoon and food security Frontiers in Agricultural
Sustainability (Integrated Publications TM) ch1, pp 1–18
[2] Yun K S, Lee J Y, Timmermann A, Stein K, Stuecker M F,
Fyfe J C and Chung E-S 2021 Increasing ENSO–rainfall
variability due to changes in future tropical temperature–
rainfall relationship Commun. Earth Environ. 21–7
[3] Chand R 2022 Changes in Labour Force and Employment in
Rural and Urban India: 2017–18–2020–21 (Indian Economic
Association)
[4] Kroeger C and Reeves A 2022 Extreme heat leads to short-
and long-term food insecurity with serious consequences for
health Eur. J. Public Health 32 521
[5] Mishra V, Smoliak B V, Lettenmaier D P and Wallace J M
2012 A prominent pattern of year-to-year variability in
Indian summer monsoon rainfall Proc. Natl Acad. Sci. USA
109 7213–7
[6] Shukla J and Mooley D A 1987 Empirical prediction of the
summer monsoon rainfall over India Mon. Weather Rev.
115 695–704
[7] Sikka D R 1980 Some aspects of the large scale fluctuations
of summer monsoon rainfall over India in relation to
fluctuations in the planetary and regional scale circulation
parameters Proc. Indian Natl Sci. 89 179–95
[8] Gadgil S and Gadgil S 2006 The Indian monsoon, GDP and
agriculture Econ. Polit. Wkly. 41 4887–95 (available at: http://
www.jstor.org/stable/4418949)
[9] Lenka S, Devi R, Joseph C M and Gouda K C 2022 Effect of
large-scale oceanic and atmospheric processes on the Indian
summer monsoon Theor. Appl. Climatol. 147 1561–76
[10] Ashok K, Guan Z and Yamagata T 2001 Impact of the Indian
Ocean dipole on the relationship between the Indian
monsoon rainfall and ENSO Geophys. Res. Lett. 28 4499–502
[11] Cherchi A, Terray P, Ratna S B, Sankar S, Sooraj K P and
Behera S 2021 Indian Ocean Dipole influence on Indian
summer monsoon and ENSO: a review Indian Summer
Monsoon Variability ed J Chowdary, A Parekh and
C Gnanaseelan (Elsevier) ch 8, pp 157–82 (available at:
www.sciencedirect.com/science/article/pii/B97801282–
24021000119)
[12] Ratna S B, Cherchi A, Osborn T J, Joshi M and Uppara U
2021 The extreme positive Indian Ocean Dipole of 2019 and
associated Indian summer monsoon rainfall response
Geophys. Res. Lett. 48 e2020GL091497
[13] Hrudya P H, Varikoden H and Vishnu R 2021 A review on
the Indian summer monsoon rainfall, variability and its
association with ENSO and IOD Meteorol. Atmos. Phys. 133
1–14
[14] Chakraborty A 2018 Preceding winter La Ni˜
na reduces
Indian summer monsoon rainfall Environ. Res. Lett.
13 054030
[15] Mohanty P and Padhi S R 2022 Role of ENSO and IOD in
the Indian summer monsoon variability: a review
[16] Xiao H M, Lo M H and Yu J Y 2022 The increased frequency
of combined El Ni˜
no and positive IOD events since 1965s
and its impacts on maritime continent hydroclimates Sci.
Rep. 12 7532
[17] Anderson W B, Seager R, Baethgen W, Cane M and You L
2019 Synchronous crop failures and climate-forced
production variability Sci. Adv. 5eaaw1976
[18] Anderson W et al 2023 Climate variability and simultaneous
breadbasket yield shocks as observed in long-term yield
records J. Agric. Meteorol. 331 109321
[19] Stuecker M F, Tigchelaar M, Kantar M B and Magar V 2018
Climate variability impacts on rice production in the
Philippines PLoS One 13 e0201426
[20] Iizumi T, Luo J J, Challinor A J, Sakurai G, Yokozawa M,
Sakuma H, Brown M E and Yamagata T 2014 Impacts of El
Ni˜
no Southern Oscillation on the global yields of major
crops Nat. Commun. 53712
[21] Ray D K, Gerber J S, MacDonald G K and West P C 2015
Climate variation explains a third of global crop yield
variability Nat. Commun. 65989
[22] Heino M, Puma M J, Ward P J, Gerten D, Heck V, Siebert S
and Kummu M 2018 Two-thirds of global cropland
area impacted by climate oscillations Nat. Commun.
91257
[23] Heino M, Guillaume J H A, Müller C, Iizumi T and
Kummu M 2020 A multi-model analysis of teleconnected
crop yield variability in a range of cropping systems Earth
Syst. Dyn. 11 113–28
[24] Najafi E, Pal I and Khanbilvardi R 2019 Climate drives
variability and joint variability of global crop yields Sci. Total
Environ. 662 361–72
[25] Bhatla R, Varma P, Verma S and Ghosh S 2020 El Nino/La
Nina impact on crop production over different agro-climatic
zones of Indo-Gangetic Plain of India Theor. Appl. Climatol.
142 151–63
[26] Nageswararao M M, Dhekale B S and Mohanty U C 2018
Impact of climate variability on various Rabi crops over
Northwest India Theor. Appl. Climatol. 131 503–21
[27] Davis K F, Chiarelli D D, Rulli M C, Chhatre A, Richter B,
Singh D and DeFries R 2018 Alternative cereals can improve
water use and nutrient supply in India Sci. Adv. 4eaao1108
[28] Davis K F, Chhatre A, Rao N D, Singh D and DeFries R 2019
Sensitivity of grain yields to historical climate variability in
India Environ. Res. Lett. 14 064013
13

Environ. Res. Lett. 19 (2024) 104059 M Gurazada et al
[29] Davis K F, Chhatre A, Rao N D, Singh D, Ghosh-Jerath S,
Mridul A, Poblete-Cazenave M, Pradhan N and DeFries R
2019 Assessing the sustainability of post-Green revolution
cereals in India Proc. Natl Acad. Sci. 116 25034–41
[30] ICRISAT International Crops Research Institute for the
Semi-Arid tropics 2015 village dynamics in South Asia Meso
level data for India: 1966–2011 (available at: http://data.
icrisat.org/dld/src/crops.html)
[31] Rao V U M, Kumar P V, Rao B B, Khandgonda I R and
Rao C S District level crop weather calendars of major crops
in India
[32] Rao C A R et al 2022 Impact of climate change on
productivity of food crops: a sub-national level assessment
for India Environ. Res. Commun. 4095001
[33] Saji N and Yamagata T 2003 Possible impacts of Indian
Ocean Dipole mode events on global climate Clim. Res.
25 151–69
[34] Trenberth K (National Center for Atmospheric Research
Staff) (ed) The climate data guide: nino SST Indices (Nino
1+2, 3, 3.4, 4; ONI and TNI) (available at: https://
climatedataguide.ucar.edu/climate-data/nino-sst-indices-
nino-12-3-34-4-oni-and-tni)
[35] National Center for Atmospheric Research Staff (ed) The
climate data guide: SST data: hadiSST v1.1 (available at:
https://climatedataguide.ucar.edu/climate-data/sst-data-
hadisst-v11)
[36] NOAA Physical Sciences Laboratory (available at: https://psl.
noaa.gov/gcos_wgsp/Timeseries/DMI/)
[37] Rayner N A, Parker D E, Horton E B, Folland C K,
Alexander L V, Rowell D P, Kent E C and Kaplan A 2003
Global analyses of sea surface temperature, sea ice, and night
marine air temperature since the late nineteenth century J.
Geophys. Res. 108
[38] Rajeevan M, Bhate J, Kale J and Lal B 2006 High resolution
daily gridded rainfall data for the Indian region: analysis of
break and active monsoon spells Curr. Sci. 91 10
[39] Pai D, Rajeevan M, Sreejith O, Mukhopadhyay B and
Satbha N 2014 Development of a new high spatial resolution
(0.25◦×0.25◦) long period (1901–2010) daily gridded
rainfall data set over India and its comparison with existing
data sets over the region MAUSAM 65 1–18
[40] Harris I, Osborn T J, Jones P and Lister D 2020 Version 4 of
the CRU TS monthly high-resolution gridded multivariate
climate dataset Sci. Data 7109
[41] Iizumi T, Takaya Y, Kim W, Nakaegawa T and Maeda S 2021
Global within-season yield anomaly prediction for major
crops derived using seasonal forecasts of large-scale climate
indices and regional temperature and precipitation Weather
Forecast. 36 285–99
[42] Singh J, Ashfaq M, Skinner C B, Anderson W B and Singh D
2021 Amplified risk of spatially compounding droughts
during co-occurrences of modes of natural ocean variability
npj Clim. Atmos. Sci. 41–14
[43] Singh J, Ashfaq M, Skinner C B, Anderson W B, Mishra V
and Singh D 2022 Enhanced risk of concurrent regional
droughts with increased ENSO variability and warming Nat.
Clim. Change 12 163–70
[44] Eliazer Nelson A R L, Ravichandran K and Antony U 2019
The impact of the Green revolution on indigenous crops of
India J. Ethn. Foods 68
[45] Ben-Ari T and Makowski D 2014 Decomposing global crop
yield variability Environ. Res. Lett. 9114011
[46] Datta P, Behera B and Rahut D B 2022 Climate change and
Indian agriculture: a systematic review of farmers’
perception, adaptation, and transformation Environ. Chall.
8100543
[47] Seleiman M F, Al-Suhaibani N, Ali N, Akmal M, Alotaibi M,
Refay Y, Dindaroglu T, Abdul-Wajid H H and Battaglia M L
2021 Drought stress impacts on plants and different
approaches to alleviate its adverse effects Plants 10 259
[48] Chaturvedi P, Govindaraj M, Govindan V and Weckwerth W
2022 Editorial: sorghum and pearl millet as climate resilient
crops for food and nutrition security Front. Plant Sci.
13 851970
[49] Pandey V, Misra A K and Yadav S B 2019 Impact of El-Nino
and La-Nina on Indian climate and crop production Climate
Change and Agriculture in India: Impact and Adaptation ed
S S Mahdi (Springer International Publishing) pp 11–20
[50] Cherian S, Sridhara S, Manoj K N, Gopakkali P, Ramesh N,
Alrajhi A A, Dewidar A Z and Mattar M A 2021 Impact of El
Ni˜
no Southern oscillation on rainfall and rice production: a
micro-level analysis Agronomy 11 1021
[51] Barnston A, Tippett M, L’Heureux M, Li S and DeWitt D
2012 Skill of real-time seasonal ENSO model predictions
during 2002–11: is our capability increasing? Bull. Am.
Meteorol. Soc. 93 631–51
[52] Shi L, Hendon H H, Alves O, Luo J J, Balmaseda M and
Anderson D 2012 How predictable is the Indian Ocean
Dipole? Mon. Weather Rev. 140 3867–84
[53] Maggio G and Sitko N 2019 Knowing is half the battle:
seasonal forecasts, adaptive cropping systems, and the
mediating role of private markets in Zambia Food Policy
89 101781
[54] Guido Z, Zimmer A, Lopus S, Hannah C, Gower D,
Waldman K, Krell N, Sheffield J, Caylor K and Evans T 2020
Farmer forecasts: impacts of seasonal rainfall expectations
on agricultural decision-making in Sub-Saharan Africa Clim.
Risk Manage. 30 100247
14