The objective is to assess the causes of exceptional floods in the Ouémé basin using the HEC-RAS model. The results of the calculation made it possible to characterize: the losses and damage due to human settlement on the banks and agricultural production in the flood zone, the flooded extent and the height of submersion depend on the return period, most of the Flood waters converge towards the west of the basin (low risk) and towards the east around the Damè-Wogon depression (high risk).

A scientific model to simulate river discharges in un-measured locations is made available via a service on the web for end-users. It is shown how this allows an increasing uptake by non-modelers, for the benefit of hydrological assessments and management.

Groundwater-induced flooding, a rare phenomenon that is increasing in low-elevation coastal cities due to higher water tables, is often neglected in flood risk mapping due to its sporadic frequency and considerably lower severity with respect to other flood hazards. A loosely coupled flood model is used to simulate the interplay between surface and subsurface flooding mechanisms simultaneously. This work opens new horizons on the development of compound flood models from a holistic perspective.

In this work, we proposed a multi-source data assimilation framework for near-real-time flood mapping. We used a quasi-2D hydraulic model to update model states by injecting both stage gauge observations and satellite-derived flood extents. Results showed improvements in terms of water level prediction and reduction of flood extent uncertainty when assimilating both stage gauges and satellite images with respect to the disjoint assimilation of both observations.

This short paper is the preface of the PIAHS volume of the IAHS/UNESCO FRIEND-Water conference of Cotonou in November 2021.