LAPS data analysis and model initialization at ISAC-CNR

The forecast of extreme or high impact meteorological events faces with difficulties related to the chaotic nature of the atmosphere and to the limitations of the physical parameterizations adopted in meteorological models. Forecast quality is also limited by the lack of sufficient data sources at the mesoscale. Remote sensed data (for example, satellite and radar data) are very important to improve forecast accuracy in the short range. The assimilation of meteorological data (DA) is normally performed to produce forecast analyses (IFS-ECMWF, GFS-NOAA) used to initialize medium range models at the synoptic scale. However, high space resolution and short time range forecasts require local scales analyses (e.g. convective scales) for proper model initialization. The Local Analysis and Prediction System (LAPS) is a tool designed at NOAA with the two-fold goal of providing analyses suitable for both nowcasting purposes and for initializing high-resolution meteorological models [1-4]. LAPS allows the exploitation of data coming from several conventional and non-conventional sources, including remote-sensed data as those provided by SEVIRI, Meteosat Second Generation (MSG). An implementation of the LAPS system has been set up for the purpose of providing analyses for two different models: the hydrostatic limited area meteorological model BOLAM [5-6] and the non-hydrostatic high-resolution meteorological model MOLOCH [7-8] (generally, BOLAM outputs are used as initial/boundary conditions (IC/BC) for MOLOCH). Both models have been developed at ISAC-CNR (http://www.isac.cnr.it/dinamica/projects/forecasts/index.html). BOLAM runs over a European domain while MOLOCH runs over a local domain covering the north west of Italy. By comparing BOLAM and MOLOCH simulations both forced with LAPS analyses, the effect of the inclusion of explicit convective scales in the forecasts will be analysed. The case study considered here to test the LAPS analyses as initial conditions for both models is a heavy rain episode occurred in Genoa on the 4th of November 2011, when the precipitation amount exceeded 500 mm in 24 hours and floods caused several casualties. Limited area models generally underestimated the rainfall amount for this event [11]. The LAPS analysis is expected to improve quantitative precipitation forecast accuracy at the convective-permitting resolution, thus to increase the reliability of warnings in case of extreme events. In this paper, aspects on the possible benefits coming from the initialization with LAPS analyses will be presented, with particular attention to the impact of the assimilation of space-born data.