The dynamical balance on the Amazon shelf and its implication on the properties of the Amazon River plume is not fully understood and poorly represented in global‐ and basin‐scale ocean models. In this study, the sensitivity of the Amazon shelf dynamics to tidal forcing is explored with a set of high‐resolution numerical simulations (1/36°) with and without the tide. A comparison of the simulations with sea surface salinity in situ measurements at 5°N (a location where the plume seasonally detaches from the coast and retroflects toward the east) revealed that the explicit resolution of the tide significantly improves the representation of the offshore spread of the river plume. This study further highlights the finding that tidal currents affect the properties of the whole Amazon plume. This sensitivity is explained by a near total collapse of the northwestward alongshore mean flow located near the river mouth, once the tidal forcing is included. This weakening of the ambient flow reduces (i) the dilution ratio between the ambient salty shelf waters and the riverine freshwaters and (ii) the constraint on the cross‐shore extension of the low‐salinity bulge. With tides, the plume is fresher near the river mouth (by up to 5 units), more extended in the cross‐shore direction, and more easily exported offshore by the North Brazil Current at the shelf break.

Almost every year, Bangladesh experiences disasters such as tropical cyclones, storm surges, coastal erosion or floods. Tropical cyclones originate from the North Indian Ocean and often cause devastating flood inundations in Bangladesh. Storm surges of the Bay of Bengal (BOB) are larger compared to other regions of the world for similar cyclones due to amplification by the shallow water depth, huge continental shelf and convergent coastlines. This scenario of inundation for such storm surges in the future when the sea level rise (SLR) occurs due to global warming will be different from the present. The densely populated coastal region of Bangladesh is likely to become more vulnerable in the future due to SLR. Disaster risks can be reduced if storm surges can be predicted well ahead. To assess the possible changes of inundation in the future, a widely used coastal model, Delft3D, has been applied for this BOB region. The model has been validated for the storm surge of three recent devastating cyclones, namely, Sidr, Aila and Roanu in the southern coast of Bangladesh. The validated model has been run to produce inundation maps and statistics for cyclonic storm surges such as Sidr, Aila and Roanu and probable SLR. Three possible SLR boundary conditions are chosen from the business-as-usual climate scenario representative concentration pathway 8.5 with values of 0.5 m (lower limit), 1 m (upper limit) and 1.5 m (extreme case, considering the subsidence and a rapid collapse of the Antarctic ice sheet). It is found that a category 4 cyclone such as cyclone Sidr would inundate 2.6%, 3.67% and 5.84% of the area of the country if the SLR is 0.5, 1 and 1.5 m, which will affect the livelihood of nearly 4.1, 7.0 and 9.1 million people of Bangladesh, respectively. It will also inundate up to 21.0%, 42.1% and 65.1% of the Sundarbans mangrove forest, which will undoubtedly affect the ecology of this unique ecosystem.