Performance evaluation of RegCM4 in simulating temperature and precipitation climatology over India

Abstract

In this study, the performance of the Regional Climate Modeling system version 4 (RegCM4) is evaluated over Indian regions through a 30-year climate simulation. The main focus is given on the model bias, correlation between the model simulated and observed results, and the standard error in the model simulations towards the 30-year mean temperature and precipitation climatology and their inter-annual and inter-decadal variability over India and the subregions. We find that the model-simulated temperature climatology has ~ 2 °C cold bias and the precipitation climatology has ~ 0.8 mm/day dry bias over India. Major cold bias (~ 3 °C) is noticed over South Peninsular India in the temperature climatology and a significant dry bias (~ 3 mm/day) is seen over Northeast India in the precipitation climatology. The performance skill of the model is found best over Northwest India for both temperature and precipitation, while it is found poor over South Peninsular India for temperature, and Northeast India for precipitation. Our overall results indicate that the RegCM4 reproduced the temperature climatology reasonably well over Indian regions, while it underestimated the precipitation climatology over many regions particularly during monsoon season. We find that less availability of moisture in the atmosphere and perhaps a weaker convection due to weaker upward motion of air may lead to decrease the precipitation in the model.