Possible future projection of Indian Summer Monsoon Rainfall (ISMR) with the evaluation of model performance in Coupled Model Inter-comparison Project Phase 5 (CMIP5)
Introduction
The Indian Summer Monsoon (ISM) produces around 80% of its rainfall during the months of June-July-August-September (JJAS). The early or late onset of ISM and large spatial and temporal variability of Indian Summer Monsoon Rainfall (ISMR) causes floods and droughts (IPCC, 2007) and greatly affects agriculture and water resources in northern plain of the country.
In recent decades, spatial and temporal variability of rainfall have been supposed to change, but no clear evidence of global warming impact on long term series of All India Summer Monsoon Rainfall (Mooley and Parthasarathy, 1984, Kripalani et al., 2003, Guhathakurta and Rajeevan, 2008) is noticed. However, some significant trend is found at regional levels (Kumar et al., 1992, Goswami et al., 2006). The global model inter-comparison activities began in late 1980s (Cess et al., 1989) and continued with the Atmospheric Model Inter-comparison Project (AMIP) (Gadgil and Sajani, 1998, Gates et al., 1999). Researchers have examined the skill of climate models in simulating rainfall variability and the inherent bias in representation of mean monsoon rainfall and its variability on different time scales (Meehl and Washington, 1993, Kitoh et al., 1997, Gadgil and Sajani, 1998, Hu et al., 2000, Cubasch et al., 2001, Lal et al., 2001, Kang et al., 2002, May, 2002, Wang et al., 2004, Fan et al., 2012). Gadgil and Sajani (1998) have used twenty (20) Atmospheric Global Circulation Models (AGCMs) under AMIP and shows that models have not evolved to a stage where year-to-year variation of ISMR can be represented, since models have their own limitations in capturing the regional rainfall accurately (Turner and Annamalai, 2012). However, modeling studies have been carried out under different emission scenarios for the study of future summer monsoon rainfall. Very little change in All India Summer Monsoon Rainfall is noticed in climate models experiments (Lal et al., 1994, Lal et al., 1995, Mahfouf et al., 1994, Timbal et al., 1995). Climate models experiments have also been carried out under the Coupled Model Inter-comparison Project (CMIP) (Meehl et al., 2000, Kang et al., 2002, Covey et al., 2003, Achuta-Rao et al., 2004, Kucharski et al., 2008). In 1990s, World Climate Research Programme (WCRP) coordinated CMIP to perform control runs and idealized 1% per year CO2 increase experiments in climate models (Meehl, 1997). Several additional phases of the CMIP, termed as CMIP2 and CMIP2 + (Meehl et al., 2000, Meehl et al., 2005, Covey et al., 2003) were also carried out.
Kripalani et al. (2007a) have studied changes in East Asian monsoon mean precipitation and its variability in Coupled Model Intercomparison Phase 3 (CMIP3) and conducted t-test and F-ratio respectively to evaluate their statistical significance. The changes in mean precipitation varied from − 0.6% for CNRM-CM3 to 14% for ECHO-G and UKMO-HadCM3. Kripalani et al. (2007b) also examined South Asian Summer Monsoon precipitation variability in models of International Panel on Climate Change Assessment Report 4 (IPCC AR4). Only nineteen (19) models, out of twenty two (22) models of IPCC AR4, could capture 500–900 mm rainfall during summer monsoon season. This simulated mean precipitation in IPCC AR4 varies from 500 to 900 with coefficient of variation from 3 to 13%. An increase of 8% in mean monsoon precipitation is projected under doubling of CO2 scenario. Sabade et al. (2010) used CMIP3 data set in scenarios B1, A1B, A2 and examined responses of South Asian summer monsoon to a transient increase in future anthropogenic radiative forcing for the period of 2031–2050 and 2081–2100. Selected ten (10) models have been examined for projected changes in seasonal monsoon rainfall and found an increase in precipitation over western equatorial Indian Ocean and southern parts of India. Parth-Sarthi et al. (2012) studied the possible future changes in ISMR in A2, B1 and A1B scenarios in CMIP3 data. Ashfaq et al. (2009) used a high resolution nested model and suggested suppression of ISMR in future time periods due to weakening of the monsoon circulation.
Interannual variability of the monsoon and its interaction with the seasonal processes and teleconnections in the tropics is an important objective of CMIP5 (Cook et al., 2012, Lee and Wang, 2012, Li et al., 2012, Meehl et al., 2012). The CMIP5 Multi-Model Mean (MMM) is more skillful than the CMIP3 MMM with respect to observations (Sperber et al., 2013). Research work is carried in many ways on Indian Summer Monsoon using CMIP5 and CMIP3 data. In CMIP5, the RCP 4.5 experiment (2075–2099), an increase occurs in global mean precipitation of around 3.2%/K (Hsu et al., 2013) and there is a larger increase in annual mean precipitation over the Asian monsoon region with less uncertainty as compared to CMIP3 models (Lee and Wang, 2012). Taylor et al. (2011) studied precipitation in monsoon regions under various radiative forcings in 21st century of CMIP5 and Cherchi et al. (2011) analyzed global monsoons in a fully coupled atmosphere-ocean general circulation model and suggested intensification of summer monsoon in future in response to the increased moisture under CO2 forcings. The Hamburg COSMOS model shows a complex behavior with changing skewness of the rainfall distribution and an associated increase in monsoon failure events (Schewe and Levermann, 2012). Preethi et al. (2012) evaluated performance of climate models in simulating observed variability of ISMR in CMIP5 data and estimated future projections of ISMR. Menon et al. (2013b) studied variability of ISMR in twenty (20) models in CMIP5 for mid 19th century to the end of 21st century and suggested significant increase in ISMR and sub-seasonal variability under unmitigated climate change (Menon et al., 2013a). Kitoh et al. (2013) evaluates global monsoons in CMIP5 historical and climate change simulations, including statistical testing of changes in the pattern of Asian summer monsoon rainfall. Bandgar et al. (2014) demonstrates the importance of Northwest Pacific (NWP) circulation variability in predicting summer monsoon precipitation over South Asia and addressed cyclonic circulation and associated deficit of ISMR. Subodh et al. (2014) tries to extensively assess the capability of CFSv2 in simulating the ISMR and prioritizes areas which require considerable improvement for the better prediction skill of Indian summer monsoon. They have carried out 30 years of forecast system free runs to understand improvements in the prediction skill in CFSv1 and CFSv2 models with present-day initial conditions. Thus, the intraseasonal and interannual variability simulated by the model can be assessed along with the observations.
Previous studies, projected rainfall and wind circulation in CMIP5 data does not require much attention over the homogeneous monsoon regions of India. The present paper is aimed to evaluate the model’s performance in simulating rainfall and wind circulation and their future projected changes over homogenous monsoon regions in RCPs 4.5 and 8.5 of CMIP5. The details of models, data and experiments are given in Section 2. Section 3 describes CMIP5 models performance in simulating rainfall and wind circulation under Historical Experiment while Section 4 deals with possible future projected changes of rainfall and wind circulation in RCP 4.5 and 8.5. Conclusions are discussed in Section 5.
Section snippets
Models, data and experiments
Table 1, Table 2 show list of models used in Historical and RCPs of 4.5 and 8.5 in CMIP5 data. The Historical experiment is equivalent to the 20th century simulation (20C3M) of CMIP3 and models are integrated from 1850 to 2012 with external forcing changing with time. The external forcing includes GHGs, the solar constant, volcanic activity, ozone and aerosols. The forcing data for 1850–2005 is taken from observation. To evaluate model's performance, simulated rainfall in Historical experiment
Evaluation of CMIP5 model’s performance
Fig. 1 depicts homogeneous monsoon regions namely North West India (NWI), Central Northeast India (CNI), North East India (NEI), West Central India (WCI), Peninsular India (PI) and Hilly Regions (HR) (Das, 2009). Averaging of rainfall over the Indian land region is done by averaging rainfall on those grids which represents the land regions. The validation of CMIP5 models in simulating rainfall is done with IMD and GPCP observations.
Fig. 2 depicts the spatial distribution of JJAS rainfall (mm
Future projected changes in rainfall
The future projected percentage changes in JJAS (mm month− 1) rainfall during 2006–2050 in RCPs of 4.5 and 8.5 of BCC-CSM1.1 (m), CCSM4, CESM1 (BGC), CESM1 (CAM5), CESM1 (WACCM), CESM1 (FASTCHEM) and MPI-ESM-MR with respect to Historical experiment (1961–2005) is shown in Fig. 8a-k. As discussed in Section 3, these seven (7) models have shown good agreement with observed rainfall (IMD and GPCP), but the RCPs experiments data of CESM1 (FASTCHEM) is not available (Table 2), therefore is not
Conclusions
CMIP5 models simulated rainfall of Historical experiments is validated with observed rainfall of IMD and GPCP. Models BCC-CSM1.1(m), CCSM4, CESM1(BGC), CESM1(CAM5), CESM1(FASTCHEM), CESM1(WACCM), and MPI-ESM-MR shows closeness with observation (IMD and GPCP) by Taylor diagram. Models CCSM4 and MPI-ESM-MR are able to reproduced JJAS wind at 850 and 200 hPa as shown in NCEP reanalysis data. These six (6) models future projected percentage changes in rainfall for period of 2006–2050 in RCPs of 4.5
Acknowledgement
Authors acknowledge World Climate Research Program’s Working Group on CMIP5 and thank to climate modeling groups (listed in Table 1, Table 2) for producing and making available their CMIP5 model output. Author also acknowledges the use of GPCP Precipitation data and NCEP Reanalysis data provided by the NOAA/OAR/ESRL PSD, Boulder, Colorado, USA (http://www.esrl.noaa.gov/psd/). Thanks to Department of Science & Technology (DST), India for funding this research work and India Meteorological
References (66)
- et al.
An overview of results from the Coupled Model Intercomparison Project
Glob. Planet. Chang.
(2003) - et al.
Possible changes in the characteristics of Indian Summer Monsoon under warmer climate
Glob. Planet. Chang.
(2012) - et al.
Scenarios of long-term socio-economic and environmental development under climate stabilization
Technol. Forecast. Soc. Chang.
(2007) - et al.
An appraisal of coupled climate model simulations
- et al.
The version-2 Global Precipitation Climatology Project (GPCP) monthly precipitation analysis (1979–present)
J. Hydrometeorol.
(2003) - et al.
Suppression of south Asian summer monsoon precipitation in the 21st century
Geophys. Res. Lett.
(2009) - et al.
Indian summer monsoon rainfall predictability and variability associated with Northwest Pacific circulation in a suit of coupled model hindcasts
Theor. Appl. Climatol.
(2014) Interpretation of cloud-climate feedback as produced by 14 Atmospheric General Circulation Models
Science
(1989)- et al.
Effects of increased CO2 levels on monsoons
Clim. Dyn.
(2011) - et al.
Scenarios of the greenhouse gas emission and atmospheric concentrations
Sub-report 2.1A of Synthesis and Assessment Product 2.1 by the U.S. Climate Change Science Program and the Subcommittee on Global Change Research
(2007)