This work concerns the wind-driven resuspension in a narrow marine channel and the risk of transport of contaminated bottom sediments in a wide brackish lagoon in the context of a planned anthropogenic infrastructure (with forced convection, by pumping sea water). It is based on the modelling and 3D numerical simulation of salinity, current distributions and bottom shear stress (BSS). The goal is to demonstrate that, even for narrow channels of a few tens of meter of width, a wind of 10-20 m/s is sufficient to create intensive currents, to resuspend muddy bottom sediments and transport polluted sediments downstream. Several model scenarios are considered for such wind speeds in two dominant and opposite wind directions, for a channel whose bottom sediments are mainly constituted of fine particles, typically 85% of mud and 15% of fine sand. It is known that finer sediments usually play an important role to transport contaminants (due to larger surface area of smaller particles). Our main results concern the bottom shear stress along such a long and narrow channel; namely the Rove channel which confluences the Etang de Berre lagoon, and for which a project of forced current circulation is planned by pumping sea water. Our numerical results show that the mobility threshold can be easily overpassed for the muddy sediments in the Rove channel. For a bottom roughness of 5 μm (coarse silt) and a wind speed of 20 m/s, BSS can reach 0.18 N/m 2 for the N-NW wind in the median part of the channel, and even 0.21 N/m 2 in one enlargement for the S-SE wind, while BSS cr is about 0.1 N/m 2. We conclude that these local winds can permit floc erosion and even surface erosion of fine sediments in the Rove channel. Concerning the resuspension of muddy sediments, our results are consistent with the experimental study presented by Carlin et al. (2016) for a windy shallow lagoon. They are also consistent with the conclusion of Mengual et al. (2017), from erodimetry experiment for estuarine sediments, that the sediment behaves like a pure mud if the percentage of the mud fraction is more than 70%, and that the critical BSS for mobility of such bed sediments is of the order of 0.1 N/m 2. Such a lower critical BSS when the mixture is muddier is opposite to trends most often published.