Cascade sensitivity tests to model deep convective systems in complex orography with WRF

Abstract

The goal of the present study is to improve the analysis of the performance of the Weather Research and Forecasting (WRF) model, Version 4.3.3, under different settings, in the prediction of precipitation in case of deep convective events. This is done through a wide cascade sensitivity test involving parameterizations expected to play a major role in the description of precipitation for deep convective events: cloud microphysics (CM), planetary boundary layer (PBL), surface layer (SL) and land-surface (L-S) model. Four significant precipitation events, which were associated to deep convective systems and occurred between 2019 and 2021 in Lombardia and Liguria regions (Northern Italy), have been simulated according to 45 different WRF settings. High resolution simulations are required when dealing with small scale deep convective systems as well as a suitable verification method. So, traditional statistical indexes have been integrated with new-generation verification methods, based on the identification of precipitation patterns in forecast and observed fields and on the evaluation of their similarity through the calculation of a coupling index. This analysis method has been exploited to evaluate the quality of the high-resolution simulations performed, and to individuate the best-performing WRF setting in simulating intense convective events. From the present work emerges how the unique CM parameterization characterized by a triple-moment treatment for cloud ice significantly outperforms all other CM schemes underlying the main role of cloud ice in the description of the precipitation case studies. Furthermore, PBL and L-S schemes show a leading role, at least as much as CM, in the description of the events.