Abstract:
Flood forecast and crisis management are part of real time operational management. Including the representation and analysis of river hydraulic response in the forecast seems to improve the anticipation of an event and could add valuable information in the operational management. An experimental real-time system has been implemented in two sectors in France to evaluate the capacity of such a system as a decision-aid tool for flood crisis management. One of the pilot sites is a catchment in South of France with the challenge to represent karstic cavities. The full system includes 4 data streams of discharge and water level measurements, 9 rain gauges and the MΓ©tΓ©o FranceΒ forecastΒ AROME model. It is based on hydrological and hydraulic models, and an automated system runs simulations of the models, post process the results and provides operational data. The latter can be visualized into a dedicated interface for the operator. Furthermore, alerts are triggered and can concern any observed or calculated data. These alerts are used to trigger new runs, provide flood forecast information useful in crisis management, and automatically send emails or SMS. This system allows the anticipation of floods in key places and add a view of the entire situation, even in unmonitored sectors, which could be valuable to the river management authorities. This experiment highlights several challenges facing the engineers such as the availability and accuracy of data streams inputs, and the accuracy and reliability of models, which must be improved in a cyclic approach for each new flood event. Β© 2020, Springer Nature Singapore Pte Ltd.
