Rainfall nowcasting model for early warning systems applied to a case over Central Italy

Rainfall is the main precursor of floods and landslide events that cause major damages every year worldwide. For an early warning system to be useful for risk mitigation and prompt interventions, the use of accurate rainfall forecasting models is crucial. With this aim, a blended model named PRAISE-MET (Prediction of Rainfall Amount Inside Storm Events with METeo) is proposed. The model is based on a coupling of a stochastic model and numerical weather prediction (NWP) model, and its characteristics are best for improving rainfall prediction at catchment scales, where the models generally used are often based on stochastic processing without using information from weather forecasts. PRAISE-MET can be easily adapted to provide input data in models for rainfall-runoff or landslide prediction, as shown for the case study illustrated here that occurred on 1 March 2006 in Montenero di Bisaccia (Central Italy). The application revealed how the use of the PRAISE-MET model would provide useful information on landslide triggering conditions in advance. The obtained results show that it is possible to gain some hours to effectively activate the protection procedures by setting the proper levels of exceedance probability for rainfall thresholds.