Comparative analysis between satellite-derived nowcasting product and severe weather event records for regional risk assessment

Extreme convective events represent a major challenge for meteorological risk assessment and rapid‑alert systems. This thesis investigates the potential of the Rapidly Developing Thunderstorm (RDT) satellite product, developed within the EUMETSAT–NWC SAF programme, as support for monitoring and evaluating severe convective hazards in Emilia‑Romagna. The main objective is to assess whether satellite‑detected convective cells can be related to observed ground impacts. An original methodology integrates heterogeneous information from RDT and the European Severe Weather Database (ESWD), establishing a spatio‑temporal association between convective cells and severe‑weather reports. Classification and predictive modelling techniques, including Logistic Regression and Random Forest, are used to estimate cell hazardousness and evaluate performance. Results indicate that RDT has limited capability in reliably discriminating ground impacts, mainly due to intrinsic product constraints and incomplete observations, yet meaningful relationships emerge between specific cell characteristics and impact likelihood. The second part of the work examines territorial exposure through a land‑use‑based exposure map for Emilia‑Romagna. The integration of RDT cells, ESWD reports, and exposure levels shows that impact distribution depends not only on convective intensity but also on territorial features and reporting capacity. Overall, the thesis outlines a coherent and adaptable framework for combining satellite‑based detection with impact information, contributing to innovative approaches for weather‑risk evaluation under substantial observational uncertainty.