High-Resolution Weather Forecasting over Complex Orography: Sensitivity to the Assimilation of Conventional Data

Abstract

Weather forecasting for regions with complex orography, as the Alps, presents several challenges and the task becomes even more difficult when high resolution is required. Moreover, for the Alpine region, some of the problems are due to the lack of observations especially over the Mediterranean Sea. A possibility for improving forecasts is to reuse assimilation techniques locally. In this paper, results obtained through data assimilation are presented: objective analysis (OA) of observations and data analyses from the European Centre for Medium-Range Weather Forecasts (ECMWF) are used together to generate a new set of mesoscale initial (ICs) and boundary conditions (BCs). In particular, OA is applied to surface data and radiosoundings using two methods: Cressman and multiquadric. The sensitivity of the weather forecast to the number of upper-air stations assimilated by OA is tested using data from the Piedmont flood (4?6 November 1994). At first, a comparison is made between ICs, obtained through the data assimilation, and the surface data; then a few weather forecast experiments, using the fifth-generation Pennsylvania State University?NCAR Mesoscale Model (MM5), are performed to assess the impact of the data assimilation on the forecast. The results show a measurable improvement in the high-resolution precipitation forecast. It is also shown that this technique can be used for high-resolution real-time forecasts.