Understanding Hydroclimatic Extremes in Changing Monsoon Climates with Daily Bias Correction of CMIP5 Regional Climate Models over South Asia

Abstract:

The assessment of hydroclimatic and hydrometeorological extremes in changing climates has gathered special attention in the latest IPCC 5^thAssessment Report (AR5). In monsoon regions such as South Asia, hydrologic modeling (i.e., stream flow assessment, water budget analysis, etc.) needs to incorporate such extremes to simulate retrospective and future scenarios. For information of past and future climate, Regional Climate Models (RCMs) are preferred over global models due to their higher resolution and dynamic downscaling capabilities. Although the models perform well in representing the mean climate, they still possess significant biases, especially in daily hydrometeorological extremes over monsoon regions. Therefore, modification and correction of RCM results while preserving the extremes are crucial for hydrologic modeling in changing monsoon climates such as in South Asia. In this context, we generate a gridded observed product that preserve the hydroclimatic and hydrometeorological extremes for the Ganges-Brahmaputra-Meghna (GBM) basin region in South Asia. A recent approach to bias correction is also proposed for correcting regional climate data in currently available future projections. The 30 year dataset (1971-2010) is used for comparing hydroclimatic and hydrometeorological extremes with APHRODITE and ERA-Interim Reanalysis products. The assessment has revealed that the new gridded data set provides much accurate maximum rainfall intensity, number of dry days, number of wet days and number of rainy days with greater than 500mm rainfall than any other available gridded data products. Using the gridded data sets, bias correctionis applied on CMIP5 multi-model historical datasets to evaluate RCM data performance over the region, which show great improvement in regional climate data for future hydrologic modeling scenarios and analyzing impacts of climate extremes.