Chronic flooding occurring at high tides under calm weather conditions is an early impact of sea-level rise. This hazard is a reason for concern on tropical islands, where coastal infrastructure is commonly located in low-lying areas. We focus here on the Guadeloupe archipelago, in the French Antilles, where chronic flood events have been reported for about 10 years. We show that the number of such events will increase drastically over the 21st century under continued growth of CO₂ emissions.

A model downscaling approach is used to investigate the effects of sea-level rise (SLR) on local tides. Results indicate that SLR induces larger increases in tidal amplitude and stronger nonlinear tidal distortion in the bay compared to the adjacent shelf sea. SLR can also change shallow-water tidal asymmetry and influence the direction and magnitude of bed-load sediment transport. The model downscaling approach is widely applicable for local SLR projections in estuaries and coastal bays.

In a context of rising sea levels linked to global warming, the issue of marine flood risk is becoming central to the management of low-lying coasts in the decades to come. The CRISSIS research program aims to propose a multidisciplinary, integrated and operational approach of marine flood risk, involving geographers, modellers, geomaticians and specialists in risk and crisis management. This work is dedicated to understand and simulate the hazard through historical and statistic approaches.

A model has been developed in order to estimate insurance-related losses caused by coastal flooding in France. It aims to identify the potential flood-impacted sectors and the subsequent insured losses a few days after the occurrence of a storm surge event on any part of the French coast. This system shows satisfactory results in the estimation of the losses related to Xynthia storm surge, which was used for the model's calibration.

Extreme value analyses of sea-level using tide-gauge measurements usually suffer from limited effective duration of observation which can result in large uncertainties, especially when outliers are present. To tackle this issue, a Bayesian MCMC method is developed integrating historical data in extreme sea-level analyses. A real case study shows a significant improvement in return values estimation and the usefulness of the Bayesian framework to predict future annual exceedance probabilities.