Evaluation of season long rainfall simulated by WRF over the NWH region: KF vs. MSKF

Abstract

Abstract This study is aimed to assess the performance of Numerical Weather Predication (NWP) model Weather Research and Forecasting (WRF) in simulating the season long rainfall pattern over the North-West Himalayan (NWH) region. For this purpose, two experiments are designed with identical initial conditions but varying cumulus physics options viz.; Kain-Fritsch (KF) and Multi-scale Kain-Fritsch (MSKF). WRF simulations have been carried out for multiple nested domains (45, 15 and 5 km) configuration for monsoon season of 2017. Model derived rainfall corresponding to two experiments has been evaluated against the gridded rainfall data prepared by India Meteorological Department (IMD) and satellite based GPM-IMERGE and INSAT-3D HE-rain. Present results unveiled that the KF scheme overestimates the amount of rainfall over the entire study domain; whereas, MSKF scheme is capable to simulate the magnitude of rainfall comparable to the observations. Model simulated rainfall corresponding to both the experiments yields stronger correlation with satellite based GPM-IMERG as compared to the IMD data. It is noted based on statistically robust Chow test that the WRF simulation with KF scheme estimates break point in the rainfall pattern at 1200 m whereas WRF simulation with MSKF scheme resolves break point at 2000 m which is in close agreement with the previous studies. Categorical statistics metrics and qualitative analysis of WRF simulations suggest that the scale aware cumulus scheme MSKF performs better in resolving monsoon season rainfall pattern over the rugged topography of the NWH region. Present study has implications for improved understanding of spatial pattern of monsoon rainfall over the NWH region.