Context of the added value in coupled atmosphere-land RegCM4–CLM4.5 in the simulation of Indian summer monsoon

Abstract

The potential of the RegCM4–CLM4.5 model for value addition over the MIROC5 GCM for Indian Summer Monsson (ISM) simulation for the present climate (1975–2005) has been examined. The added value has been investigated in terms of the added value index (AV) and the Modified Brier Skill Score (MBSS). In general, a strong dry (wet) bias > 5 mm/day in the RegCM4 (MIROC5) dataset was prevalent over different parts of India. Additional precipitation statistics such as the area-averaged mean, standard deviation, correlation coefficient, and percentage bias over the homogeneous region of precipitation suggest that the RegCM4 occasionally improves over the MIROC5, depending upon the region and the metric under consideration. Further, a comparable tropospheric temperature gradient and the vertical wind shear have been found for both the dataset. However, this does not hold for the large-scale moisture transport as the RegCM4 shows an advantage in representing the magnitude and the location of the low-level jet. The dominant unrealistic moisture transport and associated stronger Bay of Bengal branch of circulation causes heavy precipitation (> 12 mm/day), thus leading to improved precipitation statistics for MIROC5. A quantifiable added value over the GCM in terms of the positive indices have been found in the RegCM4. The AV (MBSS) suggests positive values i.e. 0.6–0.8 (0.2–0.5) over major parts of India, signifying added information and/or the improvement upon the downscaling. The experiment certainly provides additional information in the precipitation simulation over the major part of India however the two indices do not necessarily agree over certain areas.