CORDEX-SA Regional Climate Model Ensemble Performance Evaluation and Future Climate Change over India

Abstract

<p>The present study aims to analyse future changes in mean maximum and extreme temperature over India using 17 regional climate simulations from  CORDEX-SA experiment RCM ensembles at a spatial resolution of 0.5° x 0.5° for mid-term (2041-2060) and long-term (2081-2099) future under RCP 4.5 and RCP 8.5 scenarios. The regional climate simulations of the three RCM ensembles namely IITM-RegCM4, SMHI-RCA4 and MPI-CSC-REMO2009 were first evaluated against observed seasonal maximum temperature (Mar-Jun) during historical period (1971-2005). The datasets were obtained from CCCR-IITM ESGF data node for the study period. The model performance was assessed using standard performance measure statistics such as mean absolute error, root mean square error, mean bias, percentage bias. The RCMs show warm and cold bias in simulating the climatological mean maximum temperature with mean bias ranging from to 8.43°C(warmest) to -37.29°C (coldest) by CNRM-CM5 RCM of IITM-RegCM4 ensemble and by CSIRO-Mk3.6 RCM of SMHI-RCA4 ensemble respectively over the country. Variance scaling bias correction method was applied to correct the bias associated with the RCMs which significantly reduced the RMSE of RCMs from 11.03 (SMHI-RCA4) and 9.17 (IITM-RegCM4) to around zero after bias correction. Future changes assessed in mean maximum temperature show an increase in the range of 0.5°C to 4.5°C during mid-term (2041-2060) and 0.6°C to 5.84°C long-term (2081-2099) future period while under RCP 8.5 the increase ranges from 0.88° C to 5.40° C for mid-term (2041-2060) and 1.31° C to 12.87° C for long-term (2081-2099) period. The most pronounced increase is observed in the northern, eastern and north-eastern region of the country in which highest rise was simulated by IPSL-CM5A-MR(SMHI-RCA4) in northern region of the country for both the scenarios. The study also analyses changes in different ET-SCI indices as a measure of extreme temperature which have increased multi-fold in the future over the country. The study identifies the regions and magnitude of significant climate change signal expected in the mean maximum and extreme temperature in the future. It enhances the understanding and quantification of inter-model uncertainties within CORDEX-SA experiment RCM simulations. The outcomes of the study have both scientific and societal values in building resilience and informed efforts to avert the severe implications posed by increasing extreme temperature and heat events over India.</p><p> </p><p>Keywords: Future Climate Change, CORDEX-SA, Regional Climate Model, Bias Correction, Extreme weather events</p>