Performance of cloud microphysical parameterizations in simulating Indian summer monsoon depressions

Abstract:

Numerical Weather Prediction (NWP) is one of the vital components of meteorology and has direct implications for society. However, the most significant contributions to the forecast errors have sources in the parameterization schemes of the NWP model. This study assesses the impact of five cloud microphysical parameterization schemes (MP) on simulations of fourteen monsoon depressions (MDs) using the Weather Research and Forecasting (WRF version 3.8.1) model, where the MDs are important rain-bearing synoptic disturbance which accounts for an ample amount of monsoonal rainfall over the Indian subcontinent. The simulations are carried out with a lead time up to 96 hours at 27, 9, and 3km horizontal resolution for fourteen MDs for composite study. The selected five MP schemes are WRF single moment 6 class (WSM6), WRF double moment 6 class (WDM6), Milbrandt (MIL), Thompson (THOM), and Aerosol Aware Thompson (AAT). These fourteen MDs formed over the Bay of Bengal and moved towards India and have different stages of the storm (developed as ‘low’ and attained maximum category of ‘deep depression’). It is found that the choice of MP significantly impacts the key characteristics of the MDs, such as rainfall, wind, temperature, and associated convective processes. In general, MPs have underestimated the precipitation compared to Tropical Rainfall Measuring Mission (TRMM), and WDM6 has the least errors; WDM6 produced the lowest amount of wind underestimation in the south-western sector of the storm up to 96 hours of simulation. The lowest biases in moisture convergence and absolute vorticity are found in WDM6 at the surface, which leads to lower moisture flux from ground and thus better rainfall correlation. Further, inter-comparisons of simulations of MPs are carried out using WDM6 as the benchmark, and the detailed results will be discussed.