Climatic Change (2022) 172:37

https://doi.org/10.1007/s10584-021-03068-2

Precipitation variability over India

during the 20th and 21st centuries: investigating
natural and anthropogenic drivers

P. Kishore 1 & Ghouse Basha 2 & M. Venkat Ratnam 2 & Amir AghaKouchak 1,3 &
Isabella Velicogna 1,4 & M. Rajeevan 5

Received: 16 January 2019 / Accepted: 16 March 2021/

# The Author(s), under exclusive licence to Springer Nature B.V. 2022

Abstract
Changes in precipitation pattern can lead to widespread impacts across natural and human
systems. This study assesses precipitation variability as well as anthropogenic and natural
factors responsible for the precipitation variability over India in the twentieth and twenty-first
centuries using the Coupled Model Intercomparison Project (CMIP5) and observational data
sets. We specifically quantify the impact of natural (volcanic and solar radiation) (NAT),
greenhouse gas (GHG), anthropogenic aerosol (AA), and land use (LU) forcings on
precipitation changes over India. In the observational record, a decrease in precipitation is
observed during monsoon and winter seasons, whereas an increase is noticed in pre- and
post-monsoon seasons. These observed changes are likely dominated by changes to GHGs
and a reduction in AA during the twentieth century compared to those of the LU and NAT
forcings. Anthropogenic aerosol plays a dominant role in reducing rainfall in the summer
monsoon, leading to drying trends over the second half of the twentieth century in India. The
projected simulations indicate an increasing trend during the twenty-first century. The
temporal and spatial variability of precipitation displays substantial changes over India
during different seasons at different periods under the Representative Concentration Path-
way (RCP) 2.6 and 8.5 emission scenarios. Overall, model simulations based on RCP 8.5
show increases in future precipitation throughout the twenty-first century over India.

Keywords India . Precipitation . GHG . AA . CMIP5 . Climate change

1 Introduction

Precipitation plays a significant role in regional climate and represents a fundamental link
between the atmosphere, hydrosphere, and biosphere. Precipitation over India, which occurs

* P. Kishore
kishore1818@gmail.com

Extended author information available on the last page of the article

 37 Page 2 of 20 Climatic Change (2022) 172:37

mostly during the monsoon season (June, July, August, and September), plays a major role in
the Indian economy. More than 75% of India’s annual rainfall falls during this season. Apart
from the monsoon precipitation over India, winter (December, January, and February) precip-
itation also plays an important role over the Western Himalaya (WH), particularly in Himachal
Pradesh and other regions in North Eastern India. Winter precipitation in the WH occurs under
the influence of extra-tropical systems known as western disturbances (WDs) (Dimri and Dash
2012).
Several studies have analyzed spatial trends of precipitation and temperature in India during
the twentieth century (Rahman et al. 2009; Kumar et al. 2010; Kishore et al. 2016; Basha et al.
2017; Mazdiyasni et al. 2017; Mishra et al. 2014; and Ali and Mishra 2018). Temporal
variations of annual rainfall show significant increasing trends over different parts of India
(Guhathakurta and Rajeevan 2007). Kishore et al. (2016) analyzed spatial trends using gridded
precipitation data from 1901 to 2007 and reported significant changes in precipitation across
part of India. An increase in precipitation was noticed over the northeastern part of India in all
the seasons, whereas a decrease was reported over the western Himalayas, north central
eastern, and southern parts of India.
The need for understanding potential changes in precipitation patterns in India caused by
climate change was realized decades ago. Climate change, natural variability, and local
anthropogenic activities (e.g., land use change) interact at different spatial and temporal scales.
Quantifying these effects and identifying their interactions and impacts on climatic variables
remains a scientific challenge. Climate models offer an opportunity to investigate the under-
lying mechanisms of historical climate change using different forcings. Tuner and Annamalai
(2012) discussed the impact of climate change on South Asian summer monsoons. They
concluded that increases in surface temperature and CO2 concentrations do not explain the
corresponding increase in rainfall over India due to the decadal variability of monsoon, and the
effect of aerosols, which result from industrialization and land use land cover change. Using
climate models, several studies have pointed out that greenhouse gases (GHGs) have influ-
enced changes in global precipitation (Polson et al. 2013a; Wu et al. 2013). Other studies have
argued that a weakening of the Asian summer monsoon circulation is primarily due to the
influence of GHGs (Kitoh et al. 1997; Cherchi et al. 2011; and May 2011). Both GHGs and
anthropogenic aerosol (AA) can affect the South Asian summer monsoon and the global
ecosystem (Bollasina et al. 2011; Ramanathan et al. 2005; Zhang et al. 2016; Guo et al. 2015;
Li et al. 2015). Recent studies show that anthropogenic aerosol (AA) can potentially play a
major role in the Northern Hemisphere (e.g., Polson et al. 2014). Climate models that include
AA forcing generate pronounced drying trends over the Asian summer monsoon region
(Bollasina et al. 2011; Salzmann et al. 2014; Santer et al. 1995). Xie et al. (2013) used CMIP5
climate simulations that consider AA and GHG separately. They also observed that regional
changes in GHG and AA are similar over ocean temperature and precipitation. Furthermore,
they found that AAs play a more dominant role, leading to drying trends in several regions
located in the Northern Hemisphere than those located in the Southern Hemisphere. Recently,
Zhang et al. (2016) reported a drying trend in East Asia during the boreal summer over the
period 1850–2005 using CMIP5 simulations, which is caused mainly by the AA effect and to a
lesser extent by GHGs.
Chaturvedi et al. (2012) and Menon et al. (2013) demonstrated an increasing trend in annual
precipitation over India with a limited number of CMIP5 models. Paul et al. (2016) reported
the weakening of the Indian summer monsoon rainfall due to the changes in land use and land
cover. The CMIP5 models (MPI-ESM-LR, INM-CM4 and MRI-CGCM3) represent the
Climatic Change (2022) 172:37 Page 3 of 20 37

spatial variation of rainfall when compared to the observational data during monsoon (Jena
et al. 2016). One of the key gaps in the literature is lack of a comprehensive analysis on the
relative influence of natural (volcanic and solar radiation) (NAT), GHG, AA, and land Use
(LU) forcings on the precipitation variability over India. In this study, we investigate precip-
itation variability using a large number of simulations from the Coupled Model Intercompar-
ison Project phase 5 (CMIP5) with different forcings.
The main goals of this article are as follows: (1) to evaluate the performance of CMIP5
models during the twentieth century using ground-based observations; (2) to quantify the
impact of natural (NAT) and anthropogenic (GHG, AA and LU) forcings on precipitation
changes over India; and (3) to analyze the temporal and spatial variability of projected
precipitation in the twenty-first century for different time periods under RCP 2.6 and 8.5.
The structure of this paper is as follows: Section 2 provides background on the CMIP5 models
and observational data along with the methodology used in this paper; Section 3 presents the
key results and discusses the performance of CMIP5 models and other factors responsible for
the observed changes in precipitation. Section 4 provides a brief discussion on projected
changes in precipitation during the twenty-first century. Finally, the conclusion and future
work are summarized in Section 5.

2 Data

2.1 Observational datasets

The India Meteorological Department (IMD) has developed high-resolution daily gridded
(0.25ox0.25o) precipitation data, which is archived with the National Data Centre, Pune, for the
period of 1901–2015 (Pai et al. 2014). This dataset draws daily rainfall records from 6955 rain
gauges. An average of ~ 500 observations is available for each grid (Rajeevan et al. 2008; Pai
et al. 2014). The minimum number of observations available for each grid is ~ 50. The data
density varies from year to year, with about 1450 stations in the first year (1901) and 3950
stations during the years 1991–1994. The data density was relatively higher (> 3100 stations
per day) from 1951. More stations are available over the west coast of India. This data was
evaluated rigorously against other observational and reanalysis data sets (Kishore et al. 2016).

2.2 CMIP5 simulations

We have analyzed precipitation simulations from CMIP5 models, which include historical
GHG, AA, LU, and NAT (see Table 1) (Taylor et al. 2012). This study uses historical
simulations in which all forcings have been applied to the models for the period 1850–2005,
including anthropogenic greenhouse gas concentrations and volcanic aerosols. These simula-
tions are performed by different modeling groups that participated in CMIP5, organized by the
World Climate Research Programme (WCRP) Working Group on Coupled Modeling
(WGCM) and addressed in the 5th Assessment Report (AR5) of the IPCC. The CMIP5 long
term experiments include a coupled control run for the twentieth century with all forcings
(historical run) and different emission and mitigation scenarios in which a Representative
Concentration Pathway (RCP) leads to different approximate radiative forcings at the end of
the twenty-first century (RCP2.6, RCP4.5, RCP6.0, and RCP8.5). Some institutes provide
historical simulations with natural forcings only (historical Nat run). The period 1901–2005
 37 Page 4 of 20 Climatic Change (2022) 172:37

was derived from the historical experiment, and the period 2006–2100 was extracted from the
RCP 2.6, 4.5, 6.0, and 8.5 scenarios (Taylor et al. 2012). We have used all four RCPs from the
year 2006–2100. The concentrations of radiatively active gases and aerosols for RCP 2.6, 4.5,
6.0, and 8.5 are as follows: 490, 650, 850, and 1370 ppm CO2by the end of the twenty-first
century, respectively. In 2100, CO2-equivalent concentrations are greater than 1370 ppm in the
RCP8.5 and almost four times the present-day atmospheric CO2concentration. All the GCM
datasets have been re-gridded to a uniform resolution of 1ox1o through cubic spline interpo-
lation. Precipitation over India is not homogeneous, so we have divided the whole India into 7
different regions as categorized by IMD (Rao 1976), which is shown in Fig. S1. Precipitation
datasets indicate the long-term changes in regional patterns at temporal and spatial time scales.
Therefore, to describe the regional trend patterns over time, the robust regression analysis was
performed with a 95% confidence interval (Shepard 1968). This technique is based on iterative
reweighted least-squares (IRLS) regression, an improvement to ordinary least-squares regres-
sion and is less affected by outliers (Holland and Welsch 1997). These weights are based on
the residuals, which measure how far the observation is from its predicted value (Kutner et al.
2004). Further, we applied the maximum covariance analysis (MCA) to find the relationship
between two variables. This evaluation was to isolate the most coherent pairs of spatial
patterns and identify a linear relationship between two geographical fields that were most
closely related to each other (Von Storch and Zwiers 1999). The MCA method is commonly
based on the Monte Carlo method and evaluates the expected rectangular covariance to that of
a randomly scrambled ensemble (Wallace et al. 1992). Considering all this information, we
have taken the ensemble mean of the best performing models for studying projected future
changes.

3 Results

3.1 Twentieth century precipitation analysis

The monthly mean climatology of precipitation from the IMD gridded precipitation data is
plotted in Fig. 1a. During the monsoon (June, July, August, and September), India receives
approximately 85% of its annual precipitation. The maximum precipitation is during the month
of July, followed by August, September, and June (Fig. 1a). The IMD divided India into seven
geographical and climatologically coherent zones accounting for precipitation variability (see
Fig. S1) (Rao 1976). Figures 1(b–f) illustrate the spatial mean climatology of precipitation
during the four seasons along with the annual variation. Spatial climatology maps confirm that
most of the precipitation occurs during the monsoon season (Fig. 1d). During the monsoon,
Northeast (NE) India receives the most amount of precipitation followed by the West Coast
(WC), and small parts of Central India (CI), while other regions receive relatively less
precipitation. During the pre-monsoon season (March, April and May), NE India receives
the most precipitation whereas during post-monsoon (October and November) east central
(EC) India receives the highest precipitation due to the NE monsoon. The lowest rainfall is
observed over the northwestern part of India throughout the whole year. During the winter and
pre-monsoon seasons, northern India receives the most amount of rainfall.
Figures 1 (g–k) display the spatial distribution of precipitation trends during the period
1901–2015 for different seasons, using the robust regression technique (Shepard 1968).
Statistical significance is assessed using the t test at 0.05 significance level (95% confidence).
Table 1 The186 CMIP5 simulations from RCP26, RCP45, RCP60 and RCP85 scenarios used in this study with different forcings including His: Historical. AA: anthropogenic
Climatic Change

aerosols, GHG: well-mixed greenhouse gases, LU: land-use change, NAT: natural forcing (all simulations were gathered from the CMIP5 website: http://cmip-pcmdi.llnl.gov/index.
html.). The last row represents the ensemble members of each model

Model ID Scenarios Model ID Scenarios Model ID Scenarios

ACCESS1–0 RCP2.6,4.5,6.0, 8.5 GFDL-ESM2G RCP2.6,4.5,6.0, 8.5 FGOALS_g2 AA, GHG, NAT, RCP2.6,4.5,6.0,
(2022) 172:37

8.5
ACCESS1.3 RCP8.5 GISS-E2-H-CC RCP8.5 FIO-ESM RCP2.6,4.5,6.0, 8.5
BNU-ESM GHG, NAT, RCP2.6,4.5,6.0, 8.5 GISS-E2-H His, AA, GHG, NAT, GFDL-CM3 His, AA, GHG, NAT,
RCP2.6,4.5,6.0, 8.5 RCP2.6,4.5,6.0, 8.5
CCSM4 His, AA, GHG, LU, NAT, GISS-E2-R-CC RCP8.5 MIROC-ESM GHG, NAT, RCP2.6,4.5,6.0, 8.5
RCP2.6,4.5,6.0, 8.5
CESM1-BGC RCP2.6,4.5,6.0, 8.5 GISS-E2-R His, AA, GHG, NAT, MIROC5 His, RCP2.6,4.5,6.0, 8.5
RCP2.6,4.5,6.0, 8.5
CESM1-CAM5 AA, GHG, LU, NAT, RCP8.5 HadGEM2-AO RCP8.5 MPI-ESM-LR His, RCP2.6,4.5,6.0, 8.5
CMCC-CESM RCP8.5 HadGEM2-CC RCP8.5 MPI-ESM-MR RCP2.6,4.5,6.0, 8.5
CMCC-CMS RCP 2.6,4.5,6.0, 8.5 HadGEM2-ES GHG, NAT, RCP 2.6,4.5,6.0, 8.5 MRI-CGCM3 His, GHG, NAT, RCP 2.6,4.5,6.0,
8.5
CMCC-CM RCP2.6,4.5,6.0, 8.5 IPSL-CM5A-LR His, AA, GHG, NAT, NorESM1-ME RCP8.5
RCP2.6,4.5,6.0, 8.5
CNRM-CM5 His, GHG, NAT, RCP2.6,4.5,6.0, and 8.5 IPSL-CM5A-MR GHG, NAT, RCP2.6,4.5,6.0, 8.5 NorESM1-M His, AA, GHG, NAT,
RCP2.6,4.5,6.0, 8.5
CSIRO-Mk3–6-0 His, AA, GHG, NAT, RCP2.6,4.5,6.0, IPSL-CM5B-LR RCP2.6,4.5,6.0, 8.5 BCC-csm1–1-m His, RCP2.6,4.5,6.0, 8.5
8.5
CanESM2 His, AA, GHG, LU, NAT, MIROC-ESM-CHEM RCP8.5 BCC-csm1–1 GHG, NAT, RCP8.5
RCP2.6,4.5,6.0, 8.5
His:14; AA:10; GHG:17; LU:5; NAT:17; inmcm4 His, RCP2.6,4.5,6.0, 8.5
RCP2.6:27; RCP4.5:27; RCP6.0: 27; RCP8.5: 27
Page 5 of 20
37
 37 Page 6 of 20 Climatic Change (2022) 172:37

Areas in which the trends are statistically significant are hatched with an “x” mark. The main
apparent pattern in the IMD data is an increasing precipitation trend over WH. This increasing
trend is most significant during the winter and pre-monsoon seasons, followed by monsoon
and post-monsoon. The trends vary from season to season and across different regions of
India. Aside from WH, other regions of India show a moderate decrease (increase) during
winter (pre-monsoon) seasons. During the winter (pre-monsoon) the average precipitation over
India exhibits an increasing (decreasing) trend during the twentieth century. North central and
NE India show a drastic reduction in precipitation (~6 mm/dec), whereas other regions show
an increase (~2 mm/dec) during monsoon season. Jin and Wang (2017) observed a significant
reduction of rainfall over northern central India during monsoon season using various obser-
vational datasets during 1951–2002. A decrease in precipitation during the post-monsoon

Fig. 1 (a) Box-and-whisker plot for the IMD monthly precipitation climatology over India. The boxes indicate
the interquartile model spread (the range between the 25th and 75th quantiles); the green circles indicate the
outliers. (b–f) Spatial variation of annual and seasonal precipitation climatology. (g–k) Spatial trends (mm/dec)
of the annual and seasonal precipitation at 0.05 significance level (hatched regions)
Climatic Change (2022) 172:37 Page 7 of 20 37

season is noticeable in terms of spatial variation. The annual precipitation shows a clear
increasing trend except for the NE, North central, and southern tip of India.
We have also estimated the seasonal variation of precipitation for different regions of India,
as shown in Table 2 using daily IMD precipitation datasets. During the winter (pre-monsoon),
the WH region shows a strong increase in precipitation compared with other regions, with the
rate of 2.6 (3.3) mm/dec during 1901–2015. Only the WH region exhibits an increase in
precipitation trends, while other regions show decreasing trends in winter. This increase is very
large compared with the precipitation in all of India during the winter. Both winter and pre-
monsoon trends show similar patterns. All regions, except for WH and IP, show decreasing
precipitation trends during the monsoon season. Precipitation over India as a whole shows a
decreasing trend of about 0.6 mm/dec during the monsoon season. During the post-monsoon
season, the WH, WC, and NW regions show increasing trends, while other regions show
decreasing trends. The annual rainfall over India shows an increasing trend, except over the
NE region. Considering the entire record, the whole India shows an increase in precipitation
with a rate of 0.3 mm/dec, whereas the monsoon precipitation decreases.

3.2 Assessment of CMIP5 models

At first, we have evaluated 31 CMIP5 model precipitation datasets with the observed IMD
gridded precipitation in India to investigate which models capture spatial, temporal, and
seasonal precipitation variability during the period 1970–2000. Previous studies have
highlighted substantial variability in seasonal performance of climate models (Liu et al.
2014), and hence, the mean bias and correlation coefficients of all the models relative to the
observations have been estimated at each grid point for all seasons separately. The mean bias
of IMD and individual models and their correlation values during the monsoon are shown in
Fig. 2a. The other seasons have also been evaluated and the results are depicted in Fig. S2 in
Supplementary Information. The spatial bias and correlation between IMD and ACESS1–0
model precipitation for monsoon are shown in Figs. 2b and c. The positive bias is observed in
most regions in India as seen from Fig. 2b. In some regions, a negative bias is observed over
western Himalaya and northeastern part of India. The spatial correlation between the IMD and
ACCESS1–0 model shows good relationships in most of the regions, despite some negative
correlations observed over the southwestern part of India (Fig. 2c). The average correlation
coefficient for India is 0.8. Figure 2d shows the power spectral density estimated using cross-
spectral analysis during the same period (1997–2000) between IMD and model datasets. This
technique can help identifying the common wavelet structures and investigating characteristics

Table 2 IMD Seasonal precipitation trends over India for different geographical regions at 0.05 significance
level (95% confidence) from 1901 to 2015

Units: mm/dec Winter Pre-monsoon Monsoon Post-monsoon Annual

India −0.011 0.282 −0.587 0.049 0.325

EC −0.275 −0.048 0.776 0.083 0.421
IP −0.327 −0.201 0.134 −0.238 0.476
NC −0.439 −0.053 −1.836 −1.027 −0.325
NE −0.312 −0.224 −2.891 −0.197 −0.859
NW −0.109 0.142 0.511 0.381 0.708
WC −0.127 0.017 0.183 1.231 1.873
WH 2.557 3.251 1.941 1.932 2.320
 37 Page 8 of 20 Climatic Change (2022) 172:37

of the underlying signals (Stott et al. 2016). The power spectra illustrate several peaks in IMD
precipitation that are coherent in phase with the same period of oscillations in the models,
indicating significant coherence between both time-series. Note that all these analyses have
been performed for each historical data from CMIP5 models for each season separately during
the period 1970–2000. For the rest of the analysis, we have considered only models that have a
correlation coefficient greater than 0.6 and also exhibit a relatively low bias with respect to the
observations. Therefore, for each season a set of representative models have been chosen as
follows: winter (MIROC5, NorESM1-ME, IPSL-CM5A-LR, and FGOALS_g2), spring
(CESM1-CAM5, GISS-E2-H, GISS-E2-R, ACCESS1–0, and MIROC-ESM), monsoon
(NorESM1-M, GFDL-CM3, BNU-ESM, inmcm4, ACCESS1–0, and FGOALS_g2), fall
(GFDL-ESM 2G, HadGEM2-CC, IPSL-CM5A-LR, HadGEM2-ES, and HadCM3), and for
annual we considered most of the models except IPSL-CM5B-LR and MRI-CGCM3 due to
their very high mean bias and low correlation (Fig. 2). Ramesh and Goswami (2014)
mentioned that the use of better models rather than conducting a multimodel average is more
beneficial for certain applications. Jena et al. (2016) also studied the warming impact on
rainfall over Indian regions and found that only a few models can capture the monsoon season
rainfall while a few others can capture the winter season rainfall well. They concluded that no
ideal model exists which is consistent with our finding reported in Fig. 2.

3.3 Influence of anthropogenic and natural forcings on precipitation

In this section, we briefly discuss the factors responsible for the changes in precipitation over
India. Temporal evolution of precipitation anomalies and the corresponding standard

Fig. 2 (a) Mean bias of monsoon season precipitation between IMD and individual CMIP5 models over the
entire India—the corresponding correlation coefficients are included within the boxes. (b) Example spatial bias
between IMD and ACCESS1–0, (c) example spatial correlation between IMD and ACCESS1–0, and (d) cross-
spectral power spectrum corresponding to monsoon precipitation time series based on IMD data and different
climate models during1975–2000
Climatic Change (2022) 172:37 Page 9 of 20 37

deviations (the shaded regions) for IMD along with historical (ALL), AA, GHG, LU, and
NAT experiments are depicted in Fig. 3. The observed precipitation anomalies vary signifi-
cantly with respect to time over India. The precipitation deficiencies over India caused by
historical droughts are shown by the vertical shaded lines (light red color) (Gadgil and Gadgil
2006). Major droughts are observed in 1966 and 2002 with an Indian summer monsoon
rainfall deficit of ~ 23% and large negative anomalies occurring in the western and north-
western parts of India. An increase in precipitation is observed from 1901 to 1960 followed by
a decreasing trend thereafter in observed data. The trend in observational monsoon rainfall
over the entire time frame (1901–2005) shows a decreasing trend of 0.4 mm/year. For the
country as a whole, the AA forcing demonstrates a negative trend in the monsoon precipita-
tion, whereas GHG (LU) exhibits a strong increasing (decreasing) trend. To further investigate
the temporal variability, we have investigated the correlation between the simulations and
observations during 1901–1960, 1961–2005, and 1901–2005 (Table 3). Considering monsoon
precipitation, it appears that NAT followed by AA and GHG exhibit higher correlations with
observations as displayed in Table 3, though they vary at different study periods.
Noteworthy changes in precipitation due to anthropogenic forcings including aerosol
emissions, occurred after the year 1960 (Moorthy et al. 2013; Acharya and Sreekesh 2013).
Aerosol emissions have been rising in India since the 1950s due to rapidly increasing
population, industrial growth, and also cooking fires. Guo et al. (2015) used 24 available

Fig. 3 Temporal variations of monsoon precipitation anomalies for the period 1901–2005. ‘All’ indicates the
historical forcing including natural and anthropogenic. Individual forcings include Anthropogenic Aerosol (AA),
Greenhouse Gases (GHG), Land Use change (LU), Natural forcings (solar radiation + volcanic eruptions) (NAT).
The precipitation anomaly corresponding to each forcing is also plotted (blue line). The vertical lines in the figure
indicate drought periods over India
 37 Page 10 of 20 Climatic Change (2022) 172:37

Table 3 Correlation coefficients between IMD and precipitation simulations from various models and forcings
over different periods

Anthropogenic vs IMD correlation coefficient

1901–1960/1961–2005/1901–2005

Model Winter Pre-monsoon Monsoon Post-monsoon Annual

ALL −.084/0.48/ 0.37/−0.22/ −0.11/0.54/ −0.12/−0.17/ 0.30/0.57/

0.411 0.486 0.354 −0.044 0.367
AA −.10/−0.07/ −0.23/0.35/ 0.72/−0.32/ −0.22/−0.39/ 0.49/0.44/
−0.012 −0.044 0.461 −0.162 0.412
GHG −0.02/0.62/ 0.31/0.33/ −0.12/−0.14/ 0.52/0.39/ 0.38/0.47/
−0.067 −0.032 0.362 −0.012 0.518
LU 0.45/−0.06/ 0.33/0.61/ −0.12/−0.09/ −0.16/0.41/ 0.38/0.62/
0.321 −0.182 0.126 0.328 0.382
NAT −0.06/−0.14/ 0.06/−0.011/ 0.53/0.34/ 0.33/−0.082/ 0.42/0.26/
−0.067 0.423 0.478 −0.082 0.332

historical experiments with all-forcings and concluded that aerosol’s indirect effects play a
dominant role in reducing rainfall over South Asia under the rising anthropogenic aerosol over
the second half of the twentieth century. Polson et al. (2014) mentioned that the Northern
Hemisphere monsoon precipitation expresses substantial changes during the second half of the
twentieth century and concluded that anthropogenic aerosol is the dominant external factor. In
addition to the AA effect, the GHGs play an important role on the large-scale circulation and
precipitation changes (Zhang and Li 2016). The AA and GHG forcings individually lead to
clear decreasing and increasing trends, respectively (see Figs. 3b and d). The discrepancies
between Fig. 3b and d are probably because of the nonlinear behavior of climate models in
response to AA and GHGs simultaneously, which has been stated in the past studies (Ming
and Ramaswamy 2009; Bollasina et al. 2011).
The time-series of precipitation shows reasonable agreement with AA and NAT forcing
simulations. Some discrepancies still remain, and hence, we used the MCA technique to
determine which forcing is more influential on precipitation. From this technique, we can
identify the pairs of coupled spatial patterns that explain the largest covariance between the
IMD and AA precipitation simulations (see Fig. 4). The primary mode of MCA shows similar
trend patterns relative to IMD data across India except for a few locations (e.g., Western
Himalayas) indicating opposite signs (see Figs. 4a and b), which means the relative lower
influence of the AA forcing in those areas. The opposite sign in the Western Himalayas may
be due to limited ground-based precipitation data and a complex topography. Charles et al.
(2016) and Das et al. (2017) pointed out the combined effect of multiple drivers of the South
Asian monsoon during summer and winter, which makes the estimation of precipitation
through the general circulation models (GCMs) challenging. The time-series of IMD and
AA-forced precipitation clearly illustrate negative trends since 1950 (Fig. 4c). The squared
covariance factor (SCF) is estimated to be about 56% between IMD and AA (Fig. 4c). SCF is
basically the ratio of the total covariance explained by the squared sum of all the patterns
(modes) in percent (Bretherton et al. 1992). SFC can be used to comprehend the relative
prominence of different drivers on precipitation. For example, the first leading mode of
covariance between GHG and IMD precipitation is lower (43%) than that of AA-IMD
precipitation (56%) when comparing Fig. S3 to Fig. 4. The spatial patterns of IMD and
GHG look similar with the exception of the western Himalaya region, the southern tip of India,
Climatic Change (2022) 172:37 Page 11 of 20 37

and the northeast regions, where IMD shows an opposite sign. From these results, we observed
that both AA and GHG have a relatively significant impact on regional changes in the Indian
monsoon precipitation over the second half of the twentieth century. This is in line with
Ganguly et al. (2012) who found that reduction in summer monsoon precipitation over most
parts of India is strongly influenced by the anthropogenic and biomass burning aerosols. If the
relationship does not change in the future, rising GHG emissions would lead to increase in
monsoon rainfall for future projections.
The spatial distribution of trends in the observed precipitation and simulated precipitation
by different anthropogenic and natural forcings are shown in Fig. 5. The historical “ALL”
experiment shows evidence of decreasing annual precipitation over the NE and NW regions in
India as well as an increasing trend over EC. The drying pattern over some regions in CI is
likely to be due to AA, given that the mean trend of NAT is too weak to produce the observed
pattern (Bollasina et al. 2011). The AA trends show significant negative trends along the west
coast, Interior Peninsula, and central India, but modest decreases can also be seen in the
western Himalaya region. This is likely due to the southward shift of tropical rainfall in
response to regional forcings from Asian aerosols (Rotstayn et al. 2015). Rainfall trends in Fig.

Fig. 4 Spatial patterns of the first MCA mode associated with (a) anthropogenic aerosol (AA) and (b) IMD
precipitation during the period 1901–2005. (c) The corresponding covariance coefficients associated with AA
(pink curve) and IMD precipitation (green curve). The squared coefficient factor (SCF) and correlation
coefficients are indicated on the top of the panel
 37 Page 12 of 20 Climatic Change (2022) 172:37

5c are strongly positive due to GHG forcings. Consistent with the time-series analysis, the
spatial patterns of AA and GHG forcings exhibit opposite precipitation trends over India.
Wang et al. (2009) conducted three aerosol climate simulations with absorbing aerosols alone,
scattering aerosols alone, and both types together to investigate the impact of AA on the Indian
monsoon. They reported that scattering aerosols have limited impact, whereas absorbing
aerosols have a strong influence on the Indian monsoon. An increase/decrease of absorbing
aerosols (or AA) result in a decrease (increase) in monsoon rainfall over India, which is
consistent with our finding. Our results show that LU forcing and precipitation display
negative trends. Similar results are obtained by Paul et al. (2016) that show deforestation
leads to weakening of the Indian summer monsoon (ISM). The NAT forcing (combined effect
of both solar changes and volcanic eruptions) shows a positive trend over northern and
southern parts of India. However, the aggregated impact across the entire India is relatively
small (see Fig. 3).
It is noted that the negative (positive) precipitation trend observed over India in AA (GHG),
and even other forcings, may be due to the offset and interactions between different forcings as
observed in the East Asian summer monsoons (Zhang and Li 2016). In addition, we have also
estimated the combined anthropogenic (AA+GHG + LU) and natural (NAT) forcings
(Figure not shown for brevity). The influences of anthropogenic forcings are significantly
higher compared to that of the natural forcings over India (also apparent from Fig. 3). Overall,

Historical precipitation trends: 1901-2005

Historical_All Historical_AA Historical_GHG
40
(a) (b) (c)
35

30
Latitude (deg)

25
4.65
20
3.75
15 2.85
1.95
10
(mm/decade)

1.12
5
70 80 90 100 70 80 90 100 70 80 90 100 0.38
Historical_LU Historical_NAT -0.38
40
(d) (e) -1.12
35 -1.95

30 -2.85
Latitude (deg)

-3.75
25
-4.65
20

15

10
5
70 80 90 100 70 80 90 100
Longitude (deg) Longitude (deg)
Fig. 5 Geographical distribution of precipitation trends (mm/decade) for historical data along with different
forcings. (a) Historical surface temperature, (b) AA, (c) GHG, (d) LU, and (e) NAT. Trends were estimated at the
95% confidence level (hatched regions) for each grid
Climatic Change (2022) 172:37 Page 13 of 20 37

significant (moderate) negative trends are observed under anthropogenic (natural) forcings.
This implies that anthropogenic forcings play a more significant role in precipitation than
natural forcings.

4 Future projections of precipitation

Figure 6 shows the time-series of seasonal and annual precipitation over India from the
selected CMIP5 models (1850 to 2100), along with observational data (1901–1960). The
figure shows historical and projected precipitation based on four RCPs, all of which indicate
significant increases in precipitation in all seasons except winter. The increasing trend in
RCP8.5 is more pronounced than the other RCPs during the twenty-first century. The RCP8.5
precipitation projections show an increase of about ~ 16% by the end of the twenty-first
century during the monsoon season. The highest projected increase in precipitation is expected
in the monsoon season followed by the post-monsoon and pre-monsoon.
Figure 7 displays the relative change in precipitation for the monsoon and winter seasons
during 2080–2099 relative to 1986–2005. The relative percent change in precipitation based
on RCP8.5 (and similarly for other RCPs) is estimated as 100.*((RCP8.5-historical)/histori-
cal). Significant relative changes are observed in the precipitation based on RCP2.6, RCP4.5,
and RCP8.5 during the monsoon season, whereas the changes are relatively small in winter.
By the end of the twenty-first century, RCP2.6 shows a decrease in precipitation in both
seasons, whereas, during the monsoon, RCP8.5 exhibits a significant increase in the monsoon
precipitation. The RCP 4.5 shows increase in precipitation during both the seasons of southern
part of India, Northeast and northern part of India. During the monsoon season, the RCP8.5
precipitation projections exhibit around 5–8% higher precipitation in the western Himalayas
compared with other regions in India.
The spatial patterns of precipitation trend during the twenty-first century for different
periods, derived using robust regression analysis, are shown in Fig. 8. The hatched pixels
indicate significant trends at the 95% confidence interval. We have divided the twenty-first
century into several periods: near future (2016–2035), mid-future (2046–2065), far future
(2080–2099), and the entire period (2006–2099). Under the RCP2.6, at the end of the twenty-
first century, Central India (CI), WH, and NE India show increases in precipitation, whereas
the southern part of India exhibits decrease in precipitation. Under RCP8.5, the WC shows a
relatively small decreasing trend in the initial periods but shows increasing trends in the later
periods. There are some discrepancies across different RCPs during different projection
periods. However, different RCP projections for the entire period (2006–2099) are consistent
with respect to the sign of change, although the magnitudes of change vary significantly across
RCPs (see the last column in Fig. 8). Overall, increasing trends are observed across most India
except in small regions in northwest and western Himalaya (see RCP8.5 in Fig. 8). The largest
projected increase appears to be in the northeastern (6.5 mm/dec) region of India. Our results
are consistent with Singh and AchutaRao (2018) who reported the largest increase of over 20%
in the northeast of India at the end of the twenty-first century. While both RCP2.6 and RCP8.5
scenarios indicate that there will be an increase in precipitation at the end of the twenty-first
century (Fig. 8), there are more discrepancies between near- and mid- future projections.
We have also estimated the seasonal variation in precipitation trends for the twenty-first
century during different periods such as near-, mid-, and far-future variability under RCP2.6
and 8.5 scenarios (see Table 4). For example, the results display positive trends in winter
 37 Page 14 of 20 Climatic Change (2022) 172:37

Fig. 6 Projected seasonal and annual mean precipitation output from CMIP5 models for the years 1850–2100
relative to the base period 1901–1960. The period 2006–2099 represents the future projection scenarios from
different RCPs. The shaded region represents the standard deviation of the ensemble of the selected CMIP5
models. Observational data from IMD (1901–2005) is also plotted

precipitation simulations under RCP2.6 and monsoon and post-monsoon precipitation under
RCP8.5 in all three projections (near-, mid-, and far-future) periods. The maximum positive
trend was observed in monsoon season during the 2075–2099 period under RCP8.5. The
maximum negative trend was observed in post-monsoon season during the 2046–2099 period
under RCP2.6. Table 4 shows varying positive and negative trends in the three projection
periods (near-, mid-, and far-future). However, considering the entire period (2006–2099),
there is an overall positive trend in precipitation across all seasons and under all RCPs, except
for winter precipitation under RCP8.5. Further, the highest trend rate across the entire period
corresponds to monsoon precipitation under RCP8.5.

5 Discussion and conclusion

The results presented in this study provide an assessment of different anthropogenic and
natural forcings on precipitation variability over India. We investigated the climatology of
seasonal precipitation based on ground-based observations, the factors responsible for precip-
itation variability and change during the twentieth century, performance of CMIP5 simulations
Climatic Change (2022) 172:37 Page 15 of 20 37

Precipitation change (2080-2099)-(1986-2005)

Low Pathway (RCP2.6) Low Pathway (RCP4.5) High Pathway (RCP8.5)
40 Monsoon Monsoon Monsoon
(a) (b) (c)
35

30
Latitude (deg)

25

20 11.00

15

Percentage change (%)

7.50
10

5
3.75

Low Pathway (RCP2.6) Low Pathway (RCP4.5) High Pathway (RCP8.5)

40 Winter Winter Winter
(d) (e) (f) 0.00
35

30 -3.75
Latitude (deg)

25
-7.50
20

15

10

5
70 80 90 100 70 80 90 100 70 80 90 100
Longitude (deg) Longitude (deg) Longitude (deg)

Fig. 7 (a–d) Projected changes in precipitation from RCP 2.6 and 8.5 at the end of twenty-first century (2080–
2099) relative to the recent past (1986–2005) during monsoon and winter seasons

during the historical period and also future precipitation projections under different RCPs.
Specifically, we quantified the impact of natural (volcanic and solar radiation) (NAT),
greenhouse gas (GHG), anthropogenic aerosol (AA), and land use (LU) forcings on precip-
itation changes over India. Our conclusions can be summarized as follows:

& The average winter precipitation over the entire India exhibits an increasing trend, whereas
the post-monsoon precipitation shows a decreasing trend during the twentieth century.
Furthermore, monsoon precipitation displays an increasing trend expect over North
Central and North East parts of India.
& A significant increase in precipitation is observed in WH compared with the other regions
of India during the twentieth century.
& Precipitation changes during the twentieth century can primarily be explained by decreases
in AA and increases in GHG forcings. The overall change likely depends on the relative
strengths of these two opposing drivers.
& The CMIP5 models illustrate large uncertainty and inter-model variability in simulated
precipitation over India. Only a few models provide simulations consistent with observa-
tions with regard to bias and correlation. The uncertainties may be due to a wide range of
factors including resolution (Mishra et al. 2014), seasonal biases (Liu et al. 2014) and
unreliable convective parameterization (Kharin et al. 2007). For this reason, future pro-
jections were investigated using the best models relative to ground-based observations.
 37 Page 16 of 20 Climatic Change (2022) 172:37

Fig. 8 Geographical distribution of precipitation trends (mm/decade) at the 95% confidence level (hatched
regions) from RCP2.6, RCP4.5, RCP6.0 and RCP8.5 for different periods over India for the twenty-first century.
The periods correspond to near (2016–2035), mid (2046–2065) and far (2080–2099) future time periods, as well
as the whole period (2006–2099)

& During the twenty-first century, the seasonal and annual precipitation are projected to
increase, with the exception of winter precipitation.
& While there are discrepancies across different RCPs and projection periods, different RCP
projections for the entire projection period (2006–2099) are consistent with respect to the
sign of change in precipitation. The magnitude of change, however, varies significantly
across different RCPs.
& Climate models suggest that precipitation will substantially change in the twenty-first
century. Both RCP2.6 and RCP8.5 scenarios indicate that there will be an increase in
precipitation by the end of the twenty-first century. However, there are substantial
discrepancies between near-future (2016–2035) and mid-future (2046–2065) projections.
& Our results show a significant increase in precipitation events over WH, which extends to
the Indo-Gangetic basin. Dry winters and summers often cause large crop failures (e.g.,
wheat) in northern part of India and cotton in the south, resulting in substantial economic
Climatic Change (2022) 172:37 Page 17 of 20 37

Table 4 Seasonal precipitation trends (mm/year) over India during different periods in twenty-first century based
on RCP 2.6, 4.5, and 8.5 at 0.05 significance level

RCP2.6

Seasons 2006–2030 2046–2070 2075–2099 2006–2099

Winter 0.2810 0.132 0.345 0.287

Monsoon 2.256 0.431 −0.487 0.548
Pre-monsoon −0.327 0.623 −0.437 0.098
Post-monsoon −0.013 −2.489 1.840 0.103
Annual 0.851 −0.189 0.078 0.326
RCP4.5
Winter −0.285 0.598 −0.476 0.102
Monsoon 1.054 1.437 1.031 1.571
Pre-monsoon −0.077 0.849 −0.295 0.301
Post-monsoon 0.498 −0.108 −2.059 1.005
Annual 0.251 0.830 −0.158 0.762
RCP8.5
Winter −0.185 −0.512 −0.732 −0.032
Monsoon 1.783 2.245 3.612 2.345
Pre-monsoon 0.267 −0.039 0.289 0.306
Post-monsoon 1.034 2.361 3.419 1.589
Annual 0.791 0.707 1.725 1.213

losses. While precipitation is projected to increase in the long-run the actual impacts on
seasonal droughts and agricultural production deserves more in-depth future studies.

We acknowledge that there are uncertainties associated with future model projections and that
there is substantial inter-model variability. As models improve, future generations of the same
models may exhibit different precipitation trends. However, this paper provides a path forward for
evaluating the relative contributions of different anthropogenic and natural forcings on precipita-
tion, which is important for understanding the underlying drivers of precipitation change.

Supplementary Information The online version contains supplementary material available at https://doi.org/
10.1007/s10584-021-03068-2.

Acknowledgements We acknowledge the GCM modeling groups, the Program for Climate Model Diagnosis
and Inter-comparison (PCMDI), and the WCRP’s Working Group on Coupled Modeling for their roles in
making available the WCRP CMIP5 multi-model datasets. The authors would like to thank the National Center
for Atmospheric Research (NCAR) for providing the CRU data. The authors are also thankful to the National
Climate Centre, India Meteorological Department for providing the gridded precipitation data.

References

Acharya P, Sreekesh S (2013) Seasonal variability in aerosol optical depth over India: a spatio-temporal analysis
using the MODIS aerosol product. Int J Remote Sens 34:4832–4849. https://doi.org/10.1080/01431161
Ali H, Mishra V (2018) Increase in subdaily precipitation extremes in India under 1.5 and 2.0o C warming
worlds. Geophys Res Lett 45(14):6972–6982
Basha G, Kishore P, Venkat Ratnam M, Jayaraman A, AghaKouchak A, Ouarda TBMJ, Velicogna I (2017)
Historical and projected surface temperature over India during the 20th and 21st century. Sci Rep 7:2987.
https://doi.org/10.1038/s41598-017-02130-3
 37 Page 18 of 20 Climatic Change (2022) 172:37

Bollasina MA, Ming Y, Ramaswamy V (2011) Anthropogenic aerosols and the weakening of the south Asian
summer monsoon. Science 334:502–505
Bretherton CS, Smith C, Wallace JM (1992) An intercomparison of methods for finding coupled patterns in
climate data. J Clim 5(6):541–560
Charles SP, Chiew F, Zeng H (2016) Climate change and water in South Asia: overview and literature review.
CSIRO Sustainable Development Investment Portfolio Project Rep. 36pp
Chaturvedi RK, Joshi J, Jayaraman M, Bala G, Ravindranath NH (2012) Multi-model climate change projections
for India under representative concentration pathways. Curr Sci 103:791–802
Cherchi A, Alessandri A, Masina S, Navarra A (2011) Effects of increased CO2 levels on monsoons.
ClimDynam 37:83–101. https://doi.org/10.1007/s00382-010-0801-7
Das L, Dutta M, Mezghani A, Benestad RE (2017) Use of observed temperature statistics in ranking CMIP5
model performance over Western Himalayan region of India. Int J Climatol. https://doi.org/10.1002/joc.
5193
Dimri AP, Dash SK (2012) Wintertime climatic trends in the western Himalayas. Clim Chang 111:775–800
Gadgil S, Gadgil S (2006) The Indian monsoon, GDP and agriculture. Econ Political weekly 41:4887–4895
Ganguly D, Rasch PJ, Wang H, Yoon JH (2012) Climate response of the south Asian monsoon system to
anthropogenic aerosols. J Geophys Res 117:D13209. https://doi.org/10.1029/2012JD017508
Guhathakurta P, Rajeevan M (2007) Trends in the rainfall pattern over India. Int J Climatol 28:1453–1469.
https://doi.org/10.1002/joc.1640
Guo L, Turner AG, Highwood EJ (2015) Impacts of 20th century aerosol emissions on the south Asian monsoon
in the CMIP5 models. Atmos Chem Phys 15:6367–6378
Holland PE, Welsch RE (1997) Robust regression using iteratively reweighted least-squares. Commun Stat A6:
813–827
Jena P, Azad S, Rajeevan MN (2016) CMIP5 projected changes in the annual cycle of Indian monsoon rainfall.
Climate 4:14. https://doi.org/10.3390/cli4010014
Jin Q, Wang C (2017) A revival of Indian summer monsoon rainfall since 2002. Nat Clim Chang. https://doi.org/
10.1038/NCLIMATE3348
Kharin VV, Zwlers FW, Zhang X, Hegerl GC (2007) Changes in temperature and precipitation extremes in the
IPCC ensemble of global coupled model simulations. J Clim 20(8):1419–1444
Kishore P, Jyothi S, Basha G, Rao SVB, Rajeevan M, Velicogna I, Sutterley TC (2016) Precipitation climatology
over India: validation with observations and reanalysis datasets and spatial trends. ClimDyn. https://doi.org/
10.1007/s00382-015-2597-y
Kitoh A, Yukimoto S, Noda A, Motoi T (1997) Simulated changes in the Asian summer monsoon at times of
increased atmospheric CO2. J Meteorol Soc Japan 75:1019–1031
Kumar V, Jain SK, Singh Y (2010) Analysis of long-term rainfall trends in India. Hyd Res 55:484–496
Kutner M, Nachsteim CJ, Neter J (2004) Applied linear regression models, 4th edn. McGraw-Hill, New York
Li X, Ting M, Li C, Henderson N (2015) Mechanisms of Asian summer monsoon changes in response to
anthropogenic forcing in CMIP5 models. J Clim 28(10):4107–4125. https://doi.org/10.1175/JCLI-D-14-
00559.1
Liu Z et al (2014) Seasonal and regional biases in CMIP5 precipitation simulations. Clim Res 60:35–50. https://
doi.org/10.3354/cr01221
May W (2011) The sensitivity of the Indian summer monsoon to a global warming of 2o C with respect to pre-
industrial times. Clim Dyn 37(9–10):1843–1868. https://doi.org/10.1007/s00382-010-0942-8
Mazdiyasni O et al (2017) Increasing probability of mass-mortality during Indian Heatwaves. Sci Adv 3(6):
e1700066. https://doi.org/10.1126/sciadv.1700066
Menon A, Levermann A, Schewe J, Lehmann J, Frieler K (2013) Consistent increase in Indian monsoon rainfall
and its variability across CMIP-5 models. Earth System Dynamics 4:287–300
Ming Y, Ramaswamy V (2009) Nonlinear climate and hydrological responses to aerosol effects, J Clim, 22,
1329-1339.
Mishra V, Kumar AR, Ganguly J, Sanjay M, Mujumdar R, Krishnan M, Shah RD (2014) Reliability of regional
and global climate models to simulate precipitation extremes over India. J Geophys Res Atmos 119:9301–
9323. https://doi.org/10.1002/2014JD021636
Moorthy KK, Babu SS, Manoj MR, Satheesh SK (2013) Buildup of aerosol over the Indian region. Geophys Res
Lett 40:1011–1014. https://doi.org/10.1002/grl50165
Pai DS, Sridhar L, Rajeevan M, Sreejith OP, Satbhai NS, Mukhopadhyay B (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):1–18
Climatic Change (2022) 172:37 Page 19 of 20 37

Paul S, Ghosh S, Oglesby R, Pathak A, Chandrasekharan A, Ramsankaran R (2016) Weakening of Indian

summer monsoon rainfall due to changes in land use land cover. Nature Publ Gr 6:1–10. https://doi.org/10.
1038/srep32177
Polson D, Bollasina M, Hegerl GC, Wilcox LJ (2014) Decreased monsoon precipitation in the northern
hemisphere due to anthropogenic aerosols. Geophys Res Lett 41:6023–6029. https://doi.org/10.1002/
2014GL060811
Polson D, Hegerl DC, Allan RP, and Sarojini (2013a) Have greenhouse gases intensified the contrast between
wet and dry regions?. Geophys Res Lett 40:4783–4787. https://doi.org/10.1002/grl50923.
Rahman SH, Sengupta D, Ravichandran M (2009) Variability of Indian summer monsoon rainfall in daily data
from gauge and satellite. J Geophys Res 114:D17113. https://doi.org/10.1029/2008JD011694
Rajeevan M, Bhate J, Jaswal AK (2008) Analysis of variability and trends of extreme rainfall events over India
using 104 years of gridded daily rainfall data. Geophys Res Lett 35(L18707). https://doi.org/10.1029/
2008GL035143
Ramanathan V, Chung C, Kim D, Bettge T, Buja L, Kiehl JT, Washington WM, Fu Q, Sikka DR, Wild M
(2005) Atmospheric brown clouds: impacts on South Asian climate and hydrological cycle, PNAS 102:
5326–5333. https://doi.org/10.1073/pnas.0500656102.
Ramesh KV, Goswami P (2014) Assessing reliability of regional climate projections: the case of India monsoon.
Sci Rep 4:4071. https://doi.org/10.1038/srep04071
Rao YP (1976) Southwest monsoon. Meteorological monogrpahs, synoptic meteorology 1/1976. India meteo-
rological department, New Delhi
Rotstayn LD, Collier MA, Luo JJ (2015) Effects of declining aerosols on projections of zonally averaged tropical
precipitation. Env Res Lett 10:044018
Salzmann M, Weser H, Cherian R (2014) Robust response of Asian summer monsoon to anthropogenic aerosols
in CMIP5 models. J Geophys Res Atmos 119:11321–11337. https://doi.org/10.1002/2014JD021783
Santer BD, Taylor KE, Wigley TM, Penner JE, Jones PD, Cubasch U (1995) Towards the detection and
attribution of an anthropogenic effect on climate. ClimDyn 12(2):77–100
Shepard DA (1968) Two-dimensional interpolation function for irregularly spaced data. In: Proceedings of 1968
ACM National Conference. N Y, pp 517–524
Singh R, AchutaRao K (2018) Quantifying uncertainty in twenty-first century climate change over India.
ClimDyn 52:3905–3928. https://doi.org/10.1007/s00382-018-4361-6
Stott PA, Christidis N, Otto FEL, Sun Y, Vanderlinden JP, Oldenborgh GJV, Vautard R, Storch HV, Walton P,
Yiou P, Zwiers FW (2016) Attribution of extreme weather and climate-related events. Wiley Interdiscip Rev
Clim Chang 7(1):23–41
Taylor KE, Stouffer RJ, Meehl GA (2012) An overview of CMIP5 and the experiment design. Bull Am Meteorol
Soc 93:485–498. https://doi.org/10.1175/BAMS-D-11-00094.1
Von Storch H, Zwiers FW (1999) Statistical analysis in climate research. Cambridge University press, UK, p 494
Wallace JM, Smith C, Bretherton CS (1992) Singular value decomposition of wintertime sea surface temperature
and 500 mb height anomalies. J Clim 5:561–576
Wang C, Kim D, Ekman AML, Barth MC, Rasch PJ (2009) Impact of anthropogenic aerosols on Indian summer
monsoon. Geophys Res Lett 36:L21704. https://doi.org/10.1029/2009GL040114
Wu L, Su H, Jiang JH (2013) Regional simulation of aerosol impacts on precipitation during the east Asian
summer monsoon. J Geophys Res 118:6454–6467. https://doi.org/10.1002/jgrd.50527
Xie SP, Lu B, Xiang B (2013) Similar spatial patterns of climate responses to aerosol and greenhouse gas
changes. Nat Geosci 6(10):828–832. https://doi.org/10.1038/ngeo1931
Zhang L, Li T (2016) Relative roles of anthropogenic aerosols and greenhouse gases in land and oceanic
monsoon changes during past 156 years in CMIP5 models. Geophys Res Lett 43:5295–5301

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations.
 37 Page 20 of 20 Climatic Change (2022) 172:37

Affiliations

P. Kishore 1 & Ghouse Basha 2 & M. Venkat Ratnam 2 & Amir AghaKouchak 1,3 & Isabella
Velicogna 1,4 & M. Rajeevan 5
1
Department of Earth System Science, University of California, Irvine, CA 92697, USA
2
National Atmospheric Research Laboratory, Gadanki, Tirupati, India
3
Department of Civil and Environmental Engineering, University of California, Irvine, CA 92697, USA
4
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, USA
5
Ministry of Earth Sciences, Prithvi Bhavan, New Delhi, India