High-accuracy shallow-water bathymetric method including reliability evaluation based on Sentinel-2 time-series images and ICESat-2 data

Abstract

Shallow-water bathymetric maps provide vital geographic information for various coastal and marine applications such as environmental management, engineering construction, oil and gas resource exploration, and ocean fisheries. Recently, satellite-derived bathymetry (SDB) has emerged as an alternative approach to shallow-water bathymetry, particularly in hard-to-reach areas. In this research, an innovative approach to bathymetry was introduced. This method provides a reliable approach for generating high-accuracy and high-reliability shallow water bathymetry results. By using Sentinel-2 time series imagery combined with ICESat-2 data, four bathymetry results at different time points are produced based on four traditional bathymetry methods. For the results at each location, a statistical method is applied to evaluate the bathymetry results, remove erroneous data, and generate high-confidence bathymetry results. The validation results indicated that the accuracy of the proposed bathymetric method achieved an R² range of 0.96 to 0.99 and an RMSE between 0.42 and 1.18 meters. When contrasted with traditional methods that utilize a single temporal image, a notable enhancement in bathymetric accuracy was observed.