Nowcasting Radar for Hydrological Flood Prediction: applications in the Marche Region, Italy

Accurate flood forecasting is essential to mitigate the impacts of extreme rainfall on communities and infrastructure. Traditional hydrological prediction methods often rely on rain gauge data and numerical models, which can be limited in capturing precipitation's spatial and temporal dynamics, particularly during intense or rapid-onset events. X-band polarimetric radar provides a valuable alternative for quantitative rainfall estimation, offering finer spatial and temporal resolution crucial for hydrological applications.

This study investigates the integration of radar nowcasting into flood forecasting workflows, focusing on data from an X-band polarimetric radar operated by the Civil Protection Service of the Marche Region, Italy. Several case studies have been analyzed considering different precipitation regimes: convective events with a short-time peak of intense rainfall and stratiform events, characterized by several hours of persistent precipitation associated with frontal systems.

The Cetemps Hydrological Model (CHyM) is used to simulate river discharge and assess hydrological stress indices under three scenarios: (1) rain gauge data alone, (2) radar data alone, and (3) radar data integrated with nowcasting outputs to generate 1-hour forecasted rainfall fields. Results demonstrate that radar-based nowcasting significantly improves flood prediction accuracy and lead time, particularly in flash flood scenarios driven by convective systems.

This study highlights the importance of radar nowcasting techniques in improving flood forecasting capabilities for enhancing flood prediction in regions prone to extreme rainfall, emphasizing its role in building more resilient and proactive flood management systems.