An Empirical Analysis of the Impact of Scale in Bedforms Segmentation from Digital Bathymetric Models

Underwater dunes are the result of sediment transportation due to relatively strong or large-scale sustained flows. They can be associated with different hydrodynamical factors (i.e. grain size, flow velocity, tidal regime, etc.) and be present in different marine environments, such as deltas, estuaries, river channels, and shallow marine environments. Different approaches have been proposed in the literature to segment and characterize dunes aiming to reduce the ambiguity associated with the extraction of these structures from DBM allowing rigorous and repeatable extraction of these bedforms. Despite the success of these approaches, major challenges persist to further improve their performance. This manuscript focuses on analyzing the morphological classification of the seafloor that allows the identification of underwater dunes from a modelled seafloor surface.