On the Lateral Inflows Assessment Within a Real-Time Stage Monitoring Addressed to Flood Forecasting

The role of the lateral inflows component for an on-line stage forecasting model Muskingum based is investigated. In particular, the original empirical formulation is replaced by a physically based approach incorporating the error on lateral contribution that is updated using the stage observations available in real-time. The updating procedure for lateral inflows assessment is different depending on the weight of the lateral contribution in the downstream flood evolution. The model requires the estimation of four parameters if the downstream rating curve is unknown, otherwise only two parameters have to be determined. The accuracy of the new model formulation is tested by its application to different flood events observed along an equipped river reach of the Upper Tiber River basin, in Central Italy, characterized by a significant intermediate drainage area. Assuming the rating curve known at the downstream site, the forecast stage hydrographs are found in good agreement with the observed ones for all the investigated events and more accurate than those provided by the original model. A preliminary analysis has highlighted that the forecast reliability is not significantly affected by the rainfall spatial variability in the intermediate basin, whereas it is strictly linked to the ratio between the rainfall depth in the upstream and intermediate drainage area (RRI). Moreover, a threshold of RRI has been defined in order to identify the forecast stages which could be affected by a not negligible degree of uncertainty. On this basis, the proposed model can be conveniently employed as part of a real-time flood forecasting system involving the RRI continuous monitoring during critical storm events.