Abstract EN:

The coast of French Guiana is characterized by the northwestward migration of large mud banks alongshore and by high concentration of suspended particulate matter (SPM) resulting from the strong influence of the Amazon River outflow. Mud banks govern the regional coastal morphology. A better understanding of its extension and dynamics is crucial for the coastal ecosystems management and activities at a regional scale. Several studies have been performed over French Guiana coastal area providing descriptive information about mud-bank morphology and migration speed. However, most studies have centred on the description of intertidal areas, whereas only few have focused on the subtidal zone, the biggest part of mud banks. The overarching aim of this thesis it to adapt two different approaches, such as high resolution remote sensing data and numerical models developed based on in-situ local measurements, to gather better insights into the subtidal part morphology and the influence of external forcings on its dynamics. The latter approaches are complementary and formed an interlinked feedback system that allowed the construction of a hydro-sedimentary model regionally adapted.Optimal exploitation of the information provided by recent high spatial resolution sensors such as Landsat 8-OLI is strongly conditioned by the quality of the water reflectance signal retrieval. This implies specific methodological development. One main issue in French Guiana stands in the ability to correct marine reflectance data for sun glint contamination. Thus, an automated SWIR based method has been developed using a 4-year OLI archive gathered over the very turbid waters of French Guiana. Satisfying results based on in-situ Rrs measurements in sun glint affected areas emphasize the relevant performance of the proposed methodology. Sun glint correction allowed the recovery of the marine reflectance signal and the estimation of surface OLI SPM concentration. A method to estimate the location of mud banks footprint has been developed based on SPM recurring values. A good performance was observed when the results were compared to those obtained by locating the limit of wave damping over the fluid mud of the subtidal part of mud banks. Mud-bank migration rates in French Guiana were calculated according to the delimitation of the subtidal part, and showed slightly higher values (2.31 km/year) than suggested by earlier studies. After obtaining remote sensing data and in-situ measurements, a 2D locally adapted hydro-sedimentary model was constructed based on the information provided by the latter approaches. The model takes into account the combined action of tide, currents, river discharge, wind and waves. Besides, the model also considered hydro-sedimentary processes related to cohesive sediments. Its development has followed three different steps: i) development of the hydrodynamic model; ii) development of the hydro-sedimentary model without considering wave-mud interaction; iii) and incorporation of waves to the hydro-sedimentary model, and has been validated via in-situ measurements and remote sensing data. The influence of the external forcings has been analysed for the dry and the rainy season. Despite the assumptions of the model, it reproduces qualitatively SPM trends and patterns on the subtidal part of the mud bank and the orders of magnitude are comparable to those provided by field and remote sensing data. This PhD provides evidences of the interest of these approaches and the combination of them to obtain better results in such complicated study area. The results obtained may help to further advance the state of knowledge on mud-bank processes and its migration.