Cloud model simulation of heavy rainfall event using ensemble initialization

Abstract

An attempt has been made in the present research to simulate a heavy rainfall flash-flood event over Skopje, Macedonia on 6 August 2016. A cloud-resolving model CRM has been used to simulate a super-cell storm’s initiation and evolution. Initialization has been performed with an ensemble approach, avoiding the artificial initiation of convection. The initial conditions are taken from the upper air sounding data, derived from a triple nested WRF model forecast. The cloud model is then configured over a small domain and run with a very fine horizontal grid resolution of 250-m, with an open lateral boundary conditions LBC. This preliminary study showed that the ensemble approach has some potential advantages in the generation of the initial data and the model initialization. The applied method minimizes the uncertainties to some level and provides a more qualitative-quantitative assessment of supercell storm initiation, cell structure, evolutionary properties, and intensity. A high-resolution 3-D run is capable to resolve detailed aspects of convection, including high-intensity convective precipitation. The results are also significant from the aspect of a better understanding of storm vortex dynamics and microphysical processes, responsible for heavy rainfall production. After a series of experiments and verification, such a system could be a reliable tool in weather forecast offices for very short-range forecasting (nowcasting) and early warning of weather extremes.