LETTER • OPEN ACCESS You may also like

- Characterisation and mitigation of

Unprecedented drought in South India and recent renewable droughts in the Australian
National Electricity Market
water scarcity Andy Boston, Geoffrey D Bongers and
Nathan Bongers

- Study of electrical breakdown and

To cite this article: Vimal Mishra et al 2021 Environ. Res. Lett. 16 054007 secondary pull-in failure modes for NEM
relays
M Ramezani, S Severi, H A C Tilmans et
al.

- Monolithic three-dimensional tunnel

View the article online for updates and enhancements. FET–nanoelectromechanical hybrid
reconfigurable logic circuits
Seung Kyu Kim and Woo Young Choi

This content was downloaded from IP address 203.110.242.40 on 11/01/2024 at 13:34

 Environ. Res. Lett. 16 (2021) 054007 https://doi.org/10.1088/1748-9326/abf289

LETTER

Unprecedented drought in South India and recent water scarcity

OPEN ACCESS
Vimal Mishra1,2,∗, Kaustubh Thirumalai3, Sahil Jain1 and Saran Aadhar1
RECEIVED 1
25 November 2020
Civil Engineering, Indian Institute of Technology (IIT) Gandhinagar, Gandhinagar, Gujarat 382355, India
2
Earth Sciences, Indian Institute of Technology (IIT) Gandhinagar, Gandhinagar, Gujarat 382355, India
REVISED 3
Department of Geosciences, University of Arizona, 1040 E. 4th Street, Tucson, AZ 85721, United States of America
11 March 2021 ∗
Author to whom any correspondence should be addressed.
ACCEPTED FOR PUBLICATION
26 March 2021 E-mail: vmishra@iitgn.ac.in
PUBLISHED Keywords: drought, GRACE, climate variability, water scarcity
16 April 2021
Supplementary material for this article is available online
Original content from
this work may be used
under the terms of the
Creative Commons
Abstract
Attribution 4.0 licence. Peninsular Indian agriculture and drinking water availability are critically reliant on seasonal
Any further distribution winter rainfall occurring from October to December, associated with the northeastern monsoon
of this work must
maintain attribution to (NEM). Over 2016–2018, moderate-to-exceptionally low NEM rainfall gave rise to severe drought
the author(s) and the title
of the work, journal conditions over much of southern India and exacerbated water scarcity. The magnitude and
citation and DOI.
dynamics of this drought remain unexplored. Here, we quantify the severity of this event and
explore causal mechanisms of drought conditions over South India. Our findings indicate that the
3-year cumulative rainfall totals of NEM rainfall during this event faced a deficit of more than
40%—the driest 3-year period in ∼150 years according to the observational record. We
demonstrate that drought conditions linked to the NEM across South India are associated with
cool phases in the equatorial Indian and Pacific Oceans. Future changes in these teleconnections
will add to the challenges of drought prediction.

1. Introduction India, where certain parts of South India receive a

majority of their annual rainfall totals during the
Deficiency of the summer monsoonal precipitation is OND season (Rajeevan et al 2012). Despite lesser pre-
one of the main drivers of meteorological drought in cipitation totals compared to the ISM, the NEM is
India, which if prolonged, can transform into more critically important for water availability, agriculture,
impactful agricultural and hydrological droughts and the livelihood of millions of people residing in
(Mishra and Singh 2010, Mishra et al 2010, Mo 2011). peninsular India.
Agricultural and hydrological droughts can pose last- Previously, studies have indicated that both mon-
ing impacts on food production and water availab- soon seasons have experienced profound changes
ility, respectively (Van Loon 2015, Samaniego et al over the past few decades (Mishra et al 2012, Rajeevan
2018, Mishra 2020). India experiences two major et al 2012, Roxy et al 2015, Singh et al 2019). For
monsoon seasons—the Indian summer monsoon instance, seasonal mean precipitation associated with
(ISM), also known as the southwestern monsoon the ISM has shown a declining trend leading to
and the lesser-studied northeastern monsoon (NEM) more frequent monsoon-season deficits (Mishra et al
or the winter monsoon (Gadgil and Gadgil 2006, 2012, Christensen et al 2013, Roxy et al 2015). Sim-
Rajeevan et al 2012). The ISM is the major source ilarly, the increase in precipitation associated with
of precipitation for much of India over the period the NEM over the last few decades has been attrib-
of June to September (hereafter JJAS) and has been uted to the warming of the Indian Ocean (Mishra
the focus of extensive study (Gadgil and Gadgil 2006, et al 2012, Roxy et al 2015). Furthermore, the El
Singh et al 2019). On the other hand, the NEM is more Niño Southern Oscillation (ENSO) and Indian Ocean
important in selected parts of India and is associated Dipole (IOD) phenomena are well-known drivers
with rainfall during the period between October and of deficits in monsoon rainfall (Ashok et al 2001,
December (hereafter OND) (Kripalani and Kumar Kumar et al 2007) and are also expected to undergo
2004, Zubair and Ropelewski 2006, Yadav 2012). In changes with ongoing increases in greenhouse gases
particular, the NEM significantly impacts peninsular (Cai et al 2018, Timmermann et al 2018). Addressing

© 2021 The Author(s). Published by IOP Publishing Ltd

Environ. Res. Lett. 16 (2021) 054007 V Mishra et al

the mechanisms of why and how these monsoon sea- (GRACE) and GRACE follow on (GRACE-FO)
sons are shifting under a warming world is critical for missions. TWS is available for the period April
improving predictions of drought conditions in India. 2002 to June 2017 from the GRACE satellites. The
Whereas previous studies have shown strong link- GRACE-FO mission provides the data from June
ages between summer monsoon droughts in India 2018 to present. Therefore, the TWS data for July
and sea surface temperature (SST) variability in the 2017 to May 2018 is not available. We obtained
equatorial Indian and Pacific Oceans (Kripalani and TWS from GRACE and GRACE-FO from NASA’s
Kulkarni 1997, Barlow et al 2002, Niranjan Kumar Jet Propulsion Lab (JPL: https://podaac.jpl.nasa.gov/
et al 2013, Roxy et al 2015), few studies have focused dataset/TELLUS_GRAC-GRFO_MASCON_CRI_
on the causes of rainfall deficits associated with the GRID_RL06_V2) for the 2002–2019 period. The
NEM (Dimri et al 2016). From 2016 to 2018, South GRACE mascon product (RL06 V2) contains grid-
India witnessed severe drought conditions, which sig- ded monthly global water storage anomalies relative
nificantly impacted agriculture and water availabil- to mean, which is available at 0.5◦ spatial resolution
ity in the region (‘Chennai water crisis: City’s reser- (Wiese et al 2016). To remove the seasonal cycle from
voirs run dry,’ BBC 2019). The densely populated TWS, monthly mean TWS was removed from each
states of Andhra Pradesh, Karnataka, and Tamil Nadu month, and scale factors were applied.
continuously declared drought in 2016, 2017, and To assess the influence of SSTs on the 2016–18
2018 related to the deficits in NEM precipitation. The drought, we used monthly data from the HadSST
drought caused water crises in both urban and rural dataset (Hadley Centre) for the period 1870–2018
areas (Aguilera 2019). Despite the profound impacts at 2.0◦ spatial resolution (Rayner et al 2003). We
of the 2016–18 drought in South India, its magnitude, obtained surface air temperatures (SATs) from
drivers, and mechanisms remain unexplored. In this Berkley Earth (Rohde et al 2013) to analyze anomal-
study, we focus on the 2016–2018 drought, quantify ous temperature conditions during NEM droughts.
its severity, and investigate its causes and relation- Since SST data has a strong warming trend, we used
ships with regional and global ocean–atmosphere Ensemble Empirical Mode Decomposition (EEMD;
variability. We place this extreme event in the con- (Wu and Huang 2009)) to remove the secular trend
text of the previous droughts and conclude that its (Wu et al 2011) from SST time series as in Mishra
severity was unprecedented over the observational (2020). The EEMD method has an advantage over
record. conventional detrending as it removes both linear and
non-linear trends (Mishra 2020). We estimated SST
2. Data and methods and precipitation anomalies for the NEM (October–
December) to diagnose the linkage between precipit-
The NEM (October–December) is a dominant source ation and SST. To examine the coupled variability of
of rainfall in South India (Rajeevan et al 2012). South precipitation and SST, we use maximum covariance
India (Latitude: 8◦ N–15◦ N; Longitude: 74◦ E–81◦ E)) analysis (MCA; (Bretherton et al 1992)). In addi-
comprises of five Indian states and three union territ- tion, we used empirical orthogonal function (EOF)
ories. The region encompasses nearly 19% of India’s analysis to obtain the dominant modes of variabil-
area and harbors around 250 million people, which is ity in rainfall during the NEM when SST was not
one-fifth of the total population of India (Census of used. The MCA, performed on two fields (here pre-
India 2011). South India is an agriculturally rich part cipitation and SST) together, identifies the leading
of the country, with over 60% of its rural population modes of variability in which the variations of the
engaged in agriculture (Aulong et al 2012). The pop- two fields are strongly coupled (Mishra et al 2012).
ulation depends largely on the NEM for agricultural Sea level pressure (SLP) and wind fields (horizontal, u
production. We used gridded daily precipitation data and meridional, v) were obtained from the European
available at 0.5◦ spatial resolution for the period of Centre for Medium-Range Weather Forecasts Reana-
1870–2018 (Mishra et al 2019). Mishra et al (2019) lysis version 5 (ERA-5; (Hersbach and Dee 2016)) for
used station observations from India Meteorology the period 1979–2018 to understand the mechanism
Department (IMD) to develop the gridded precipita- of the northeast monsoon. Further, SLP and wind
tion for the pre-1900 (1870–1900) period, which was fields were regridded to 2◦ to make them consistent
merged with the gridded data available for the post- with SST.
1900 period (1901–2018; (Pai et al 2014)) from IMD. Towards predictability of NEM rainfall, we
More details on the gridded precipitation data and employed univariate and multivariate techniques. We
evaluation of its quality can be obtained from Mishra use the lagged relationship between SST anomalies
et al (2019). The gridded data capture orographic pre- and rainfall over South Asia during the NEM as a pre-
cipitation along the Western Ghats, Northeast, and dictor of OND rainfall. We used SST anomalies from
the foothills of Himalaya (Pai et al 2014, Mishra et al the Nino 3.4 region and over the northern Indian
2019). Ocean (NIO; 6◦ –24◦ N, 40◦ –100◦ E) as a predictor of
Total water storage (TWS) data were obtained monthly NEM precipitation using the following three
from the Gravity Recovery and Climate Experiment equations:

2
Environ. Res. Lett. 16 (2021) 054007 V Mishra et al

Figure 1. Three-year cumulative precipitation anomalies (mm) during the Northeast monsoon (NEM, October–December).
(a), (b) The spatial pattern of 3 year cumulative precipitation anomalies (mm) during 1874–1876 and 2016–2018 periods,
respectively, in southern India (denoted by the green box). (c) Area-averaged (over the green box) 3 year moving-mean
precipitation anomalies (%) for the period 1870–2016. Red dots in (c) demarcate the two periods of interest, and show that the
2016–18 was the 1st and 1874–76 was the 2nd worst drought in last 150 years. Long-term precipitation data is based on station
observations from the Indian Meteorological Department (IMD).

ROND = a∗ Nino3.4AMJ + ε (1) investigate the long-term observational history of

NEM rainfall in the region, we used rainfall obser-
vations from the IMD (Pai et al 2014), spanning
ROND = a∗ SST-NIOOND + ε (2) from 1870 to 2018. Domain-averaged precipitation
anomalies associated with the NEM indicate that
ROND = a∗ Nino3.4AMJ + SST-NIOOND + ε. (3) most of South India experienced exceptional (>40%)
precipitation deficits during 1874–1876 and 2016–
where ROND is rainfall during October-December 2018 (figure 1). We calculated precipitation anom-
(OND) over South India, Nino3.4AMJ is April-June alies during the NEM for one, two, and three con-
(AMJ) SST anomalies over the Nino 3.4 region, SST- secutive year durations over the 1870–2018 period
NIOOND is SST anomalies during OND over the to estimate abnormal deficit-years in the long-term
northern Indian Ocean (NIO), a and b are the regres- record (figures 1, 2 and S4, table 1). There are
sion coefficients, and ε is the residual. five pronounced periods of drought (>29% deficits)
in the overall record including the recent drought
3. Results of 2016–2018, the droughts during 2001–03, 1949–
1951, 2002–04, and the well-known Great Drought
3.1. Unprecedented recent failure of northeast of 1876–78 (Cook et al 2010, Singh et al 2018),
monsoon rainfall which was associated with the Great Madras Fam-
South India receives more than 40% of its total annual ine (Blanford 1884, Mishra et al 2019). Among these
precipitation during the NEM season (figure S1 events, our analysis indicates that the Great Drought
(available online at stacks.iop.org/ERL/16/054007/ and the recent event of 2016–18 are the most severe
mmedia)), and thus deficits in NEM rainfall pose (figure 1). During 2016–18, South India experienced
significant water-related challenges in the region. To the worst NEM drought over the last 150 years with

3
Environ. Res. Lett. 16 (2021) 054007 V Mishra et al

Table 1. Top five driest years for one, two, and three-year cumulative northeast monsoon (OND).

2 year cumulative 3 year cumulative

Precipitation Precipitation Precipitation
Drought year anomaly (%) Drought year anomaly (%) Drought year anomaly (%)

1876 −68.82 1875–76 −54.37 2016–18 −45.40

2016 −62.84 2016–17 −51.73 1874–76 −37.52
1938 −59.71 1988–89 −41.29 2001–03 −29.56
1988 −53.74 2002–03 −39.50 1949–51 −29.55
1974 −49.31 1908–09 −37.91 2002–04 −29.55

a precipitation deficit of 45%, whereas the 1874–76 rains later that year. We also note that 12-month
drought was the second-worst, with a deficit of 37% precipitation anomalies and TWS anomalies are
(table 1). We note that the 1-year and 2-year dura- well-correlated (r = 0.63), where local observations
tion NEM deficits for 1876 (69%) and 1876–77 (54%) indicate that rainfall is the major contributor of TWS
were comparable to the deficits during 2016 (63%) (Asoka et al 2017). Thus, we attribute the loss in
and the 2016–17 (52%) durations (table 1, figures regional TWS to the long-term 3-year drought, which
S2–S4). However, the consecutive 3-year NEM defi- was precipitated by the lack of NEM rainfall.
cit for 2016–18 was more significant than the Great Total water loss in South India estimated from
Drought. We find that annual rainfall anomalies addi- the GRACE satellite was 79 km3 in December 2016
tionally indicate drought conditions in 2016, 2017, (figure 2(a)). Similarly, GRACE–FO data reveal that
and 2018 (figure S5). Moreover, 2 and 3-year annual total water loss in June 2017 and 2019 was 46.5 and
rainfall anomalies for 2016–17 and 2016–18 also show 41.7 km3 , respectively (figures 2(b) and (c)). Recov-
a major rainfall deficit in South India (figure S5). ery in TWS occurred in late 2019 due to improved
Thus, we conclude that the 2016–18 drought caused NEM rainfall over the region. The 2016–2018 drought
by the failure of the NEM also contained severe caused a significant loss in TWS, which also likely res-
annual rainfall deficits. ulted in a significant depletion in groundwater across
Over individual NEM seasons, the two most South India. We caveat that we did not estimate the
extreme dry events occurred in 1876 and 2016 with overall loss in groundwater due to uncertainty in soil
precipitation deficits of 69% and 63%, respectively moisture (Long et al 2013, Castle et al 2014)—an
(table 1). The rainfall deficit in 2016 was more severe estimate outside the scope of this work—however we
in comparison to the lack of precipitation in 2017 and suspect that the groundwater depletion was driven
2018 (figure S2). The failure of the NEM in 2016 as by the drought in addition to increased groundwater
well as relatively low rainfall totals over the consec- extraction (Thomas et al 2017) during the drought
utive years were the main causes behind the 2016–18 (Asoka et al 2017). Despite the uncertainty in the
drought in South India (table 1). Overall, the 3-year estimation of total water loss from GRACE satellites
NEM drought of 2016–2018 was more severe than the (Long et al 2013), the combined influence of deple-
Great Drought of 1874–1876. Infamously, the 1876 tion in surface-water and groundwater during this
drought resulted in famine and the deaths of mil- event led to unprecedented water scarcity in South
lions of people (Mishra et al 2019, Mishra 2020). India (Aguilera 2019, ‘Chennai water crisis: City’s
The more recent 2016–18 NEM drought consider- reservoirs run dry,’ BBC 2019).
ably influenced water availability in the region and
caused a water crisis across South India (‘Chennai 3.2. Mechanism of deficit during the Northeast
water crisis: City’s reservoirs run dry,’ BBC 2019). monsoon
Furthermore, the 2016–2018 NEM drought in We examined circulation patterns to understand
South India was unprecedented in the last 150 years mechanisms behind variability in NEM rainfall. To
and had severe implications for water availability. do so, we first examined climatological surface tem-
TWS from the GRACE and GRACE–FO satellites peratures (SAT and SST), sea-level pressure (SLP),
showed a considerable loss in South India due to and wind fields at 850 hPa during the OND season
the recent (2016–2018) drought (figure 2). Twelve (figure 3). SLP and wind fields were taken from the
months moving precipitation anomalies pinpoint ERA-5 reanalysis dataset (Hersbach and Dee 2016)
the onset of drought in South India during Octo- whereas SSTs and SATs were taken from HadSST
ber 2016 and show that it continued till October (Rayner et al 2003) and Berkley Earth (Rohde et al
2018 (figure 2). Although there was a weak recovery 2013), respectively. Climatologically during boreal
from drought conditions for two months in Novem- fall, cooling SATs over the northwestern Pacific and
ber and December 2018, these rainfall totals were northern latitudes alongside comparatively warmer
not enough to negate the influence of the overall mean-annual SSTs over the northern Indian Oceans
event (2016–2018), which continued till August 2019 set up easterly wind flow across the Bay of Bengal
(figure 2), and was only alleviated by stronger NEM (figures 3(a) and (b)). In particular, warm SSTs in the

4
Environ. Res. Lett. 16 (2021) 054007 V Mishra et al

Figure 2. Total water storage (TWS) anomalies from the GRACE and GRACE–FO during 2002–2019. (a)–(c) TWS anomalies
(cm) during December 2016, June 2017, and June 2019. (d) 12-month moving-sum precipitation anomalies (cm, in blue) and
monthly TWS anomalies (cm, in red) aggregated over South India (south of 15◦ N). Note that the July 2017 to May 2018 period
contains missing data as the GRACE-FO dataset is only available from June 2018 onwards. The Pearson correlation coefficient
between TWS anomalies and precipitation anomalies is 0.63.

western Indian Ocean can elicit easterlies across the SAT patterns, gave rise to anomalous westerlies in the
Indian Ocean and favor moisture transport from the equatorial Indian Ocean, which weakened moisture
Bay of Bengal into peninsular India. These moisture- transport from the Bay of Bengal during the NEM sea-
bearing winds, which become northeasterly before son of both events (figures 3(c)–(f)). Moreover, both
landfall, bring NEM rainfall to South India (Rajeevan years were associated with anomalously low SLP and
et al 2012). Strong winds from across the South cooler surface temperatures across the Indian sub-
China Sea, driven by the underlying SAT and SLP continent and Bay of Bengal, sustaining an anom-
patterns ultimately facilitate NEM rainfall. Thus, alous anticyclonic pattern which inhibited moisture
El-Niño-like conditions in the Pacific with cooler transport into South India (figures 3(c)–(f)). In 2018,
SSTs in the northern portion of the western trop- the rainfall deficit conditions were slightly allevi-
ical Pacific Ocean, juxtaposed with cooler SSTs in ated due to favorable warm conditions in the west-
the eastern Indian Ocean and warmer SSTs in the ern tropical Indian Ocean and cooling in the East
west (i.e. resembling positive IOD-like conditions), (development of a positive IOD event) alongside the
all serve to enhance NEM rainfall over South India. development of El-Niño-conditions in the Pacific.
It is to be expected that circulation patterns which However, it should be noted that western Indian
weaken these processes ought to yield diminished Ocean warming was not particularly pronounced that
NEM rainfall. year and alongside cooler temperature anomalies in
To better understand the causes of rainfall defi- the northern Indian Ocean, resulted in an overall defi-
cits, we investigated anomalous patterns during the cit in NEM rainfall that year.
NEM season for 2016, 2017, and 2018 (figure 3). In Next, we analyzed surface temperature and pre-
2016 and 2017, as expected, cool SST anomalies pre- cipitation anomalies for the five most severe dry
vailed in the tropical Indo-Pacific and were associ- events in South India over the 1870–2018 period dur-
ated with La Niña conditions in the central Pacific ing the NEM season (figure 4). The major droughts
along with negative IOD-like conditions in the Indian in South India occurred in 1876, 2016, 1938, 1988,
Ocean (figures 3(c)–(f)). Both years witnessed anom- and 1974 (in order of severity). Out of these five
alously cooler SSTs in the eastern tropical Indian droughts, four occurred during La Niña conditions.
Ocean and western tropical Pacific, and warmer SSTs In contrast, the well-studied drought of 1876 dur-
in the western Indian Ocean and central Pacific. These ing the NEM was linked with El Niño (figure 4)—a
SST patterns, alongside SLP and adjacent continental finding reported previously (Cook et al 2010, Singh

5
Environ. Res. Lett. 16 (2021) 054007 V Mishra et al

Figure 3. Atmospheric and oceanic patterns during the 2016–18 drought in South India. (a), (b) Climatological mean surface-air
temperature (SAT, ◦ C) and sea-surface temperature (SST, ◦ C), mean sea-level pressure (SLP, Pa) and wind at 850 hPa (in (b))
during the October–December (OND) season. (c), (d) SST, SLP, and wind anomalies associated with the NEM during the OND
season of 2016, (e), (f) 2017, and (g), (h) 2018. Mean SLP and wind fields were obtained from ERA-5 whereas SST was taken from
HadSST and SAT from BEST.

et al 2018, Mishra et al 2019). However, it should be patterns responsible for co-variability between South
noted that cool SST conditions prevailed in the Pacific Indian NEM rainfall and tropical SSTs. The first
Ocean over the 1870–1876 period and the transition leading mode exhibits typical ENSO-like patterns
from the cool to warm phase occurred during the of covariance and explains 77.2% of total variance
NEM season of 1876 (Singh et al 2018). Additionally, (figure 5(a)). As demonstrated above with patterns
the western Indian Ocean was not anomalously warm of the major droughts (figure 4), MCA also indic-
as it typically is during El Niño years (figure 4(a)). ates that negative SST anomalies over the central
Nevertheless, temperature and SLP anomaly com- Pacific (i.e. La Niña) and Indian Oceans (negative
posites for the most severe dry and wet NEM years IOD) result in below normal NEM precipitation
reveal a general propensity for cooler SSTs in the over South India (figure 5(b)). The second leading
Indo–Pacific (i.e. La Niña conditions) to be asso- mode of MCA exhibits a relatively weaker relation-
ciated with precipitation deficits over South India ship between precipitation and SST anomalies during
(figures S6 and S7). On the other hand, warming the NEM (figure 5). The second mode fingerprints
in the central Pacific and Indian Oceans is associ- the role of SST warming in the Indian Ocean as a
ated with a stronger NEM and surplus precipitation driver of increased NEM precipitation in South India
(figure S7). Overall, OND cooling in the Indian and (Roxy et al 2015). We also note that there appears to
central Pacific oceans results in lower SLP and weaker be a slight dichotomy between northern and southern
wind fields, which ultimately drive rainfall deficits in South India, where NEM precipitation in the latter
South India. region is more strongly linked with ENSO (figure 5).
On the other hand, precipitation over the north-
3.3. SST variability during Northeast Monsoon ern parts of South India is more strongly associ-
To clarify the relationship between SST and ated with the second leading mode (figure 5). This
precipitation anomalies associated with the NEM, we finding might help explain some of the ambiguity
performed MCA, which helps delineate the leading surrounding the mechanisms of the impact of the

6
Environ. Res. Lett. 16 (2021) 054007 V Mishra et al

Figure 4. Sea surface temperature (SST)/surface air temperature (SAT) and precipitation (P) anomalies for the top five droughts
that occurred in South India during the northeast monsoon for 1870–2018 period. SST and SAT datasets were obtained from
Hadley Center and Berkley Earth, respectively. SAT data over few regions are not available for 1876.

1876–78 Great Drought on South Indian rainfall. variance (figure 6(a)). The second leading mode
Overall, the leading mode of SST and precipita- reveals a bipolar rainfall pattern across the north-
tion variability during the NEM shows that cold SST ern and southern parts of South India and explains
anomalies in the Indo-Pacific facilitate drought con- 11% of the total variance (figure 6). We note that the
ditions over South India. characteristics of rainfall variability derived from the
We performed EOF analysis to identify the dom- first and second modes of EOF analysis are consist-
inant patterns of NEM rainfall in South India ent with the leading modes obtained from the MCA
(figure 6). The first leading mode from the EOF (figure 5). Taken together, our findings inferred from
analysis picks out rainfall variability across the both EOFs and MCA show that the first leading mode
entirety of South India and explains 50% of total affects rainfall across South India, whereas the second

7
Environ. Res. Lett. 16 (2021) 054007 V Mishra et al

Figure 5. Links between South Indian precipitation and sea surface temperature (SST) during the Northeastern Monsoon season.
(a), (b) Correlation patterns obtained from the first leading mode of maximum covariance analysis (MCA) performed between
precipitation across South India (8◦ N–15◦ N and 74◦ E–81◦ E; see Green Box in figure 1) and SST during the
October–November–December (OND) season over 1870–2018. (c), (d) Same as in the above panels but for the second leading
mode of MCA. Rainfall was obtained from the IMD dataset whereas SST was retrieved from HadSST.

leading mode delineates opposing rainfall trends in 4. Summary and conclusions

the North versus the southern parts of South India
(figure 6). South India faced a severe water crisis during
We calculated principal components (PCs) asso- 2016–2018. In June 2019, a ‘day zero’ was declared in
ciated with the leading modes of variability derived Chennai, Tamil Nadu, due to groundwater depletion
from the EOF analysis (PC1 and PC2) to exam- and drying of four major reservoirs that supply water
ine the predictability of NEM rainfall using SST (Murphy and Mezzofiore 2019), largely induced by
anomalies (figure S8). We also computed the correl- this event. We have shown that this extreme defi-
ation between PC1 and SST anomalies in addition to cit was brought about by one of the worst droughts
oceanic indices (table S1) at different time lags (tables in the last 150 years. The 2016–2018 drought was
S2 and S3). We find that the first principal component worse than the 1874–1876 Great Drought, which was
(PC1) is strongly correlated (r = 0.23, P-value < 0.05) linked to the Great Madras famine and the deaths of
to SSTs from April–June (AMJ) in the Nino 3.4 region several million in South India (Mishra et al 2019).
(figure 6). However, PC2 is more appropriately delin- The severity of the 2016–18 event during the NEM
eated by (r = 0.33, P-value < 0.05) SST anomalies season peaked in 2016—the second singular driest
from OND in Nino 3.4 and in the NIO (figure 6). We year on record (after 1876). Dynamically, our study
use this lagged relationship between oceanic indices implicates negative IOD and La Niña conditions as
and SST anomalies with PCs to establish a predict- facilitators for NEM rainfall deficits, where land-
ive model for NEM rainfall (as in Zhou et al 2019). ward moisture transport from the Bay of Bengal into
Focusing on the first mode of variance, we used clima- peninsular India is inhibited. The prevalence of La
tological Nino 3.4 SSTs from AMJ to predict rainfall in Niña throughout 2016 and 2017 (DiNezio et al 2017)
South India during OND (figure S9). We find that the further worsened the drought that started in 2016.
OND rainfall is more skillfully predicted using AMJ Such rainfall deficits over consecutive years can res-
Nino 3.4 anomalies in comparison to SST anomalies ult in multi-year drought, which have substantial and
over OND NIO (figure S9). We also note that there is adverse impacts on surface and groundwater storage,
no significant increase in prediction skill when both and profoundly affect water availability and agricul-
AMJ Nino 3.4 and OND SST anomalies were used as ture in the densely populated South Indian region.
opposed to Nino 3.4 SST anomalies alone (figure S9) Although the intensity and timing of this recent event
due to high year-to-year variability between Nino 3.4 raise the possibility of anthropogenic forcing influen-
and NIO (figure S10). Overall, our analysis shows that cing NEM droughts, future work focusing on detec-
SST anomalies at Nino 3.4 and over NIO can be used tion and attribution is required to separate the influ-
to predict rainfall during the NEM over South India ence of natural variability (Thirumalai et al 2017,
with limited prediction skill. Williams et al 2020, Winter et al 2020). Moreover,

8
Environ. Res. Lett. 16 (2021) 054007 V Mishra et al

Figure 6. The leading modes obtained from the empirical orthogonal function (EOF) analysis of rainfall during the NEM for the
1870–2018 period. (a) The first leading EOF mode of NEM, which explains 50.6% of the total variance in NEM rainfall in South
India. (b) Lagged correlation between the first leading principle component (PC 1) and 3-month mean SST anomalies over
different regions (Nino 3.4 (5◦ S–5◦ N, 120–170◦ W), North Indian Ocean (NIO; 6◦ –24◦ N, 40–100◦ E), North Pacific Ocean
(NPO; 30◦ N–50◦ N, 120◦ E–175◦ W), North Atlantic Ocean (NAO; 6◦ –24◦ N, 10–60◦ W), Pacific Decadal Oscillation (PDO), and
Southern Oscillation Index (SOI)). (c) and (d) same as (a) and (b) but for the second leading EOF mode and the corresponding
PC 2. Year − 1, Year + 0, and Year + 1 represent the previous, current, and next year of the NEM season, respectively.

potential changes in future patterns of SST variabil- Grant No. OCE-1903482 and acknowledges the Uni-
ity in the Indian Ocean and tropical Pacific will add versity of Arizona and the Department of Geosciences
substantial uncertainty to projections and prediction for support.
of NEM rainfall.
ORCID iDs
Data availability statement
Vimal Mishra  https://orcid.org/0000-0002-3046-
The data that support the findings of this study are 6296
available upon reasonable request from the authors. Kaustubh Thirumalai  https://orcid.org/0000-
0002-7875-4182
Saran Aadhar  https://orcid.org/0000-0003-1645-
Acknowledgments 4093

We acknowledge the India Meteorological Depart- References

ment for providing the precipitation data. The
last author appreciates financial assistance from the Aguilera J 2019 Water is running out in this Indian City, causing
Indian Ministry of Human Resource Development crisis Time (available at: https://time.com/5611385/india-
chennai-water-crisis/)
(MHRD). The study is partially funded by the Ashok K, Guan Z and Yamagata T 2001 Impact of the Indian
Ministry of Earth Sciences and Ministry of Water Ocean dipole on the relationship between the Indian
Resources forum projects. KT was supported by NSF monsoon rainfall and ENSO Geophys. Res. Lett. 28 4499–502

9
Environ. Res. Lett. 16 (2021) 054007 V Mishra et al

Asoka A, Gleeson T, Wada Y and Mishra V 2017 Relative future climate change in the Midwestern United States
contribution of monsoon precipitation and pumping to J. Hydrometeorol. 11 46–68
changes in groundwater storage in India Nat. Geosci. Mishra V, Smoliak B V, Lettenmaier D P and Wallace J M 2012 A
10 109–17 prominent pattern of year-to-year variability in Indian
Aulong S, Chaudhuri B, Farnier L, Galab S, Guerrin J, Summer Monsoon Rainfall Proc. Natl Acad. Sci. USA
Himanshu H and Prudhvikar Reddy P 2012 Are South 109 7213–7
Indian farmers adaptable to global change? A case in an Mishra V, Tiwari A D, Aadhar S, Shah R, Xiao M, Pai D S and
Andhra Pradesh catchment basin Reg. Environ. Change Lettenmaier D 2019 Drought and famine in India,
12 423–36 1870–2016 Geophys. Res. Lett. 46 2075–83
Barlow M, Cullen H and Lyon B 2002 Drought in Central and Mo K C 2011 Drought onset and recovery over the United States
Southwest Asia: La Niña, the warm pool, and Indian Ocean J. Geophys. Res. Atmos. 116 20
precipitation J. Clim. 15 697–700 Murphy P P and Mezzofiore G 2019 Chennai water shortage:
Blanford H F 1884 II. On the connexion of the Himalaya snowfall India’s 6th-largest city is almost out of water. Satellite images
with dry winds and seasons of drought in India Proc. R. Soc. show bone-dry reservoirs (available at: https://
37 3–22 edition.cnn.com/2019/06/20/world/chennai-satellite-
Bretherton C S, Smith C and Wallace J M 1992 An images-reservoirs-water-crisis–trnd/index.html) (Accessed
intercomparison of methods for finding coupled patterns in 27 April 2020)
climate data J. Clim. 5 541–60 Niranjan Kumar K, Rajeevan M, Pai D S, Srivastava A K and
Cai W et al 2018 Increased variability of eastern Pacific El Niño Preethi B 2013 On the observed variability of monsoon
under greenhouse warming Nature 564 201–6 droughts over India Weather Clim. Extremes 1 42–50
Castle S L, Thomas B F, Reager J T, Rodell M, Swenson S C and Pai D S, Sridhar L, Rajeevan M, Sreejith O P, Satbhai N S and
Famiglietti J S 2014 Groundwater depletion during drought Mukhopadyay B 2014 Development of a new high spatial
threatens future water security of the Colorado River Basin resolution (0.25◦ × 0.25◦ ) long period (1901–2010) daily
Geophys. Res. Lett. 41 5904–11 gridded rainfall data set over India and its comparison with
Census of India 2011 Website: office of the registrar general & existing data sets over the region Mausam 65 1–18
census commissioner, India (available at: http:// Rajeevan M, Unnikrishnan C K, Bhate J, Niranjan Kumar K and
censusindia.gov.in/2011-Common/CensusData2011.html) Sreekala P P 2012 Northeast monsoon over India: variability
(Accessed 27 April 2020) and prediction Meteorol. Appl. 19 226–36
Chennai water crisis: City’s reservoirs run dry BBC News Rayner N A et al 2003 Global analyses of sea surface temperature,
(available at: www.bbc.com/news/world-asia-india- sea ice, and night marine air temperature since the late
48672330) (Accessed 18 June 2019) nineteenth century J. Geophys. Res. D 108 14
Christensen J H et al 2013 IPCC 2013 chapter 14: climate Rohde R, Muller R A, Jacobsen R, Muller E and Wickham C 2013
phenomena and their relevance for future regional climate A new estimate of the average earth surface land temperature
change supplementary material Climate Change 2013: The spanning 1753–2011 Geoinfor. Geostat.: Overview 1
Physical Science Basis. Contribution of Working Group I to the Roxy M K, Ritika K, Terray P, Murtugudde R, Ashok K and
Fifth Assessment Report of the Intergovernmental Panel on Goswami B N 2015 Drying of Indian subcontinent by rapid
Climate Change p 62 (https://doi.org/10.1017/ Indian Ocean warming and a weakening land-sea thermal
CBO9781107415324.028) gradient Nat. Commun. 6 1–10
Cook E R, Anchukaitis K J, Buckley B M, D’Arrigo R D, Samaniego L et al 2018 Anthropogenic warming exacerbates
Jacoby G C and Wright W E 2010 Asian monsoon failure and European soil moisture droughts Nat. Clim. Change
megadrought during the last millennium Science 328 486–9 8 421–6
Dimri A P, Yasunari T, Kotlia B S, Mohanty U C and Sikka D R Singh D, Ghosh S, Roxy M K and McDermid S 2019 Indian
2016 Indian winter monsoon: present and past Earth–Sci. summer monsoon: extreme events, historical changes, and
Rev. 163 297–322 role of anthropogenic forcings Wiley Interdiscip. Rev. Clim.
DiNezio P N et al 2017 A 2 year forecast for a 60–80% chance of Change 10 e571
La Niña in 2017–2018 Geophys. Res. Lett. 44 11624–35 Singh D, Seager R, Cook B I, Cane M, Ting M, Cook E and
Gadgil S and Gadgil S 2006 The Indian monsoon, GDP and Davis M 2018 Climate and the Global Famine of 1876–78
agriculture Econ. Political Wkly. 41 4887–95 (www.jstor.org/ J. Clim. 31 9445–67
stable/4418949) Thirumalai K, DInezio P N, Okumura Y and Deser C 2017
Hersbach H and Dee D 2016 ERA5 reanalysis is in production Extreme temperatures in Southeast Asia caused by El Ninõ
ECMWF Newslett. 147 5–6 and worsened by global warming Nat. Commun.
Kripalani R H and Kulkarni A 1997 Rainfall variability over 8 1–8
South-East Asia—connections with Indian monsoon and Thomas B F, Caineta J and Nanteza J 2017 Global assessment of
Enso extremes: new perspectives Int. J. Climatol. 17 1155–68 groundwater sustainability based on storage anomalies
Kripalani R H and Kumar P 2004 Northeast monsoon rainfall Geophys. Res. Lett. 44 11445–55
variability over south peninsular India vis-à-vis the Indian Timmermann A et al 2018 El Niño–Southern Oscillation
Ocean dipole mode Int. J. Climatol. 24 1267–82 complexity Nature 559 535–45
Kumar P, Rupa Kumar K, Rajeevan M and Sahai A K 2007 On the Van Loon A F 2015 Hydrological drought explained Wiley
recent strengthening of the relationship between ENSO and Interdiscip. Rev. 2 359–92
northeast monsoon rainfall over South Asia Clim. Dyn. Wiese D N, Landerer F W and Watkins M M 2016 Quantifying
28 649–60 and reducing leakage errors in the JPL RL05M GRACE
Long D, Scanlon B R, Longuevergne L, Sun A Y, Fernando D N mascon solution Water Resour. Res. 52 7490–502
and Save H 2013 GRACE satellite monitoring of large Williams A P et al 2020 Large contribution from anthropogenic
depletion in water storage in response to the 2011 drought warming to an emerging North American megadrought
in Texas Geophys. Res. Lett. 40 3395–401 Science 368 314–8
Mishra A K and Singh V P 2010 A review of drought concepts Winter J M, Huang H, Osterberg E C and Mankin J S 2020
J. Hydrol. 391 202–16 Anthropogenic impacts on the exceptional precipitation of
Mishra V 2020 Long-term (1870–2018) drought reconstruction in 2018 in the Mid-Atlantic United States Bull. Am. Meteorol.
context of surface water security in India J. Hydrol. Soc. 101 S5–S10
580 124228 Wu Z and Huang N E 2009 Ensemble empirical mode
Mishra V, Cherkauer K A and Shukla S 2010 Assessment of decomposition: a noise-assisted data analysis method Adv.
drought due to historic climate variability and projected Adapt. Data Anal. 1 1–41

10
Environ. Res. Lett. 16 (2021) 054007 V Mishra et al

Wu Z, Huang N E, Wallace J M, Smoliak B V and Chen X 2011 On Zhou Z, Xie S and Zhang R 2019 Variability and predictability of
the time-varying trend in global-mean surface temperature indian rainfall during the monsoon onset month of June
Clim. Dyn. 37 759–73 Geophys. Res. Lett. 46 14782–8
Yadav R K 2012 Why is ENSO influencing Indian northeast Zubair L and Ropelewski C F 2006 The strengthening relationship
monsoon in the recent decades? Int. J. Climatol. between ENSO and northeast monsoon rainfall over Sri
32 2163–80 Lanka and southern India J. Clim. 19 1567–75

11