A Numerical Study to Investigate Precipitation Features of Monsoon Deep Depressions over Bay of Bengal: Comparison of Coupled and Control Simulations

The present study is aimed to investigate the rainfall characteristics of Monsoon Deep Depressions (MDD) originating over the Bay of Bengal (BoB) basin using a coupled ocean-atmospheric model (COAWST) and a stand-alone atmospheric (WRF) model with a lead time of up to 72h. It is found that though the tracks of the four MDDs considered in the study have been reasonably simulated, the intensity was overestimated in both sets of simulations compared to India Meteorological Department (IMD) best estimates. Upon decomposition of the contributors to the rainrate for the composite of the storms in the deep depression (DD) phase, it was found that the moisture sources/sinks play a more important role than the cloud sources/sinks in modulating the rainfall processes. Further analysis of the moisture sources/sinks showed that the horizontal and vertical advection are the major drivers in modulating the contribution of the moisture sources/sinks. The validation of rainfall using CMORPH datasets suggested that the coupled simulations had a higher skill in rainfall prediction. Furthermore, the composite of different components of moisture sources/sinks (especially vertical advection) was found to be more realistically simulated in COAWST compared to WRF upon validation with MERRA datasets. Analysis of the composite energetics showed that scarcity of bulk kinetic energy in the later hours of the DD phase in COAWST led to the dissipation of the storm core, which led to better prediction of rainfall. On the other hand, a re-intensification of the storm core by means of condensational heating led to an overestimation of rainfall in WRF, which finally resulted in lower skill in rainfall prediction. In spite of the stand-alone atmospheric model capturing the horizontal moisture incursion in the lower levels significantly, the better representation of the vertical structure enabled the coupled model to capture the precipitation features more realistically, increasing skill in rainfall prediction.