Numerical Simulations of Severe Precipitation Events over Liguria (Italy) with the WRF Model and Analysis of the Sensitivity to Different Cloud Microphysics Parameterizations

Abstract:

Mediterranean coastal regions are regularly affected by sudden heavy precipitation events leading to very dangerous flash floods. Due to its position and its topographical peculiarities, one of the most affected areas is Liguria, a very complex region located in Northwestern Italy. Three different case studies relative to severe rainfall events recently occurred in Liguria have been considered in the present study. In all selected cases, the formation of a quasi-stationary mesoscale convective system over Liguria Sea interacting with local dynamical effects (orographically-induced low-level wind and temperature gradients) played a crucial role in the generation of severe precipitations.

Different sets of simulations of the aforementioned events have been performed using the Advanced Research core of the Weather Research and Forecasting (WRF) model, to investigate the sensitivity of the predicted precipitation field to model resolution and different microphysics parameterization approaches. Specifically, eight microphysics schemes available in WRF have been compared in very high-resolution (1.1 km), convection-permitting simulations. The data set used to evaluate model performances has been extracted from the official regional observing network, composed by about 150 professional WMO-compliant stations. Two different strategies have been exploited to assess the model skill in predicting precipitation: a traditional approach, where matches between forecast and observations are considered on a point-by-point basis, and an object-based method where model success is based on the correct localization and intensity of precipitation patterns. This last method allows to overcome the known fictitious models performance degradation for increasing spatial resolution.

As remarkable results of this analysis, a clear role of horizontal resolution on the model performances accompanied by the identification of a family of best-performing parameterization schemes emerge. The outcomes of the study offer important suggestions for operational weather prediction systems under potentially dangerous heavy precipitation events.