Impact analysis of dynamical downscaling on the treatment of convection in a regional NWP model – COSMO: a case study during the passage of a very severe cyclonic storm OCKHI

Abstract. A significant source of uncertainty in Numerical Weather Prediction (NWP) models arises from the parameterization of sub-grid scale convection, whose inherent nature of complexity is amplified while applied to tropical regions where weather systems are controlled by many intricate factors. However, as the model resolution becomes finer, it is possible to switch off the convection parameterization, although it is still unclear at what resolution this can be achieved. Ambiguity arises due to the inter-linking of various parameterization schemes within a model, and efficiency of one scheme depends on the output of another. In order to explore these issues, an intense convective episode with very heavy precipitation over the coastal Arabian Sea associated with the passage of OCKHI, one of the very severe cyclonic storms, is chosen as a case study. A set of distinct numerical simulations are carried out using Consortium for Small-scale Modelling (COSMO) to assess the direct and indirect impacts of dynamical downscaling on the treatment of convection. Results obtained from the present investigation indicate dynamical downscaling together with switching off the convection parameterization could simulate the magnitudes of CAPE, one of the proxies for characterizing the occurrence of tropical convection, more realistically. But the downscaling did not improve the rainfall prediction, which were seen to deteriorate in the absence of convection parameterization.