The RAMSES weather forecast modeling, one year of results

A. Tiesi1, S. Laviola1, M.M. Miglietta2, S. Gabriele3, 2018, The RAMSES weather forecast modeling, one year of results, 1st AISAM meeting, Bologna, Italy, 10-13/09/2018,

The project RAMSES (RAilway Meteorological SEcurity System) is a pilot project developed by CNR and Rete Ferroviaria Italiana S.p.A., focusing on the mitigation of the hydrological risks for railways in Calabria (Italy). In the framework of the modelling part of the project, the work presents the results based on one year of weather forecasting at the CNR-ISAC. The activity has focused on the set up of an operative numerical system, performing high-resolution short-range forecasts and provide alerts in case of intense precipitation. The WRF (Weather and Research Forecasting) model is implemented in order to perform weather forecasts over two nested domains with the finest grid spacing of 3 km. A WRF run starts every day, and represents the background fields (WRFBG) on which data assimilation is performed. The analysis of the meteorological data is performed by LAPS (Local Area and Prediction System, NOAA, USA). LAPS is a mesoscale analysis tool adaptable for any source of data and its implementation does not require high CPU performances, which makes it very useful for operative meteorological scopes. In the present project, LAPS performs analyses using data from METAR stations, soundings, radar (CAPPI at 3 levels), and SEVIRI/MSG (Eumetsat/Geostationary). A short-range forecast of WRF (WRFHR) starts every 3 hours, assuming as initial condition the real-time LAPS analysis, and covers the following 6 hours. After more than one year of forecasting activity, the statistical results show an improvement of the WRFHR forecasts compared to the WRFBG forecasts due to data assimilation, and a good level of robustness of the operative system chain WRF-LAPS-WRF.

Data from