Runoff and soil erosion modelling at catchment scale in a tropical volcanic Island (Martinique, Lesser Antilles)

Le résumé fourni par la source

Terrigenous inputs from hillslopes to hydrosystems can alter the quality of Fort-de-France Bay’s waters and sediments and so the balance of ecosystems (notably the degradation of coral reefs). Depending on the intensity of rainfall episodes, terrigenous inputs increase the turbidity the bay’s water and potentially result in the arrival of pollutants (including pesticides) in coastal waters leading to the ban of certain coastal economic activities. Another major impact is land degradation in upstream watersheds, leading to the impoverishment of agricultural soils. Furthermore, there has been much less research in soil erosion and sediment transfer in these tropical volcanic island settings compare to other environments.In order to analyze the rate and the spatio-temporal dynamics of water and suspended particles fluxes, the objective of this study is to model runoff and erosion on the two main watersheds (100 and 65 km²) draining into the Fort-de-France Bay, including a variety of soil type, land use and morphology. The modelling approach was used to perform an in-depth analysis of runoff and erosion processes. Various hypotheses were tested at different scales, particularly on the genesis of runoff (hortonian or soil saturation), soil depth, soil percolation. etc. Our results suggest the importance of saturation processes in controlling flood event occurrence. Compared with more traditional soil erosion modelling studies, this study on large watersheds have also enabled us to analyze upscaling effects’ aspects.This calibrated runoff and erosion model will be used to simulate different management or remediation plan to reduce sediment exports to the bay. Management plan will deal with agricultural practices and soft hydraulic systems such as grass strips, hedges (planted with local or endemic plant species) or fascines.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.