OPTIMAL INTERPOLATION METHOD FOR GENERATING A DIGITAL BATHYMETRIC MODEL FOR SHALLOW WATERS: A CASE STUDY OVER MAURITIUS COAST

Bathymetry unveils the underwater topography of oceans, seas, rivers, and lakes. It is a fundamental data resource for various applications, like physical oceanography, marine geology, geophysics, and marine resources. The techniques to compute the seafloor depths are ship-borne echo sensors, empirical models of satellite-derived bathymetry, and aerial-space-borne laser altimetry. The digital bathymetric surfaces are generally generated from a distributed seafloor depths. Once these depth points are collected, the next step to generate a continuous surface is to select and implement interpolation. Numerous interpolation methods have advantages and disadvantages that can hamper the accuracy of the surface, which generally depends on the shape of the extent, distribution, and point density. To date, there is no recommended interpolation method when the study extent is circular with well-distributed points – the core objective of this research is oriented towards this. An attempt was made to generate a digital bathymetric surface for the Mauritius coast with ∼ 1.2 million depth points accrued from the NASA ICESat-2 geolocated photons and sounding depths from the marine charts. These points were used as input to interpolation methods like Inverse Distance Weighted, Natural Neighbour, and various forms of Ordinary Kriging. Our findings show that all the methods have generated visually similar surfaces, but the Inverse Distance Weighted interpolation has given the output with less quantified uncertainty

Bathymetry, Interpolation, LiDAR, Inverse Distance Weighted, Natural Neighbour, Kriging