Evaluation of selected global climate models for extreme temperature events over India

Abstract

Extreme temperature events derived from the global climate models (GCMs) are used for climate change impact studies in several sectors, especially in agriculture. Reducing the uncertainty in simulated crop yields based on extreme temperatures is an important task for reliable and relevant adaptation decisions. This study compared the performance of selected six global climate models in simulating temperature extreme events over Indian region for the 1976–2005 period. For this, performance statistics such as root mean square error, correlation coefficient, and agreement index were compared spatially and spatiotemporally. The study reveals that all the six models overestimate minimum and maximum temperature extremes for most parts of Central India, which resulted in hot bias. However, these models show a cold bias in simulating low temperature extremes over the Himalayan region. Further the study indicates that GFDL-ESM2M and MIROC-ESM-CHEM have performed relatively better in capturing temperature extreme events over Indian region among the six models that were found to closely simulate the observed climatology.