Evaluating satellite-derived bathymetry accuracy from Sentinel-2A high-resolution multispectral imageries for shallow water hydrographic mapping

Abstract

The wide-ranging evolution of remote sensing technology has offered a promising application to the hydrographic surveying, the Satellite-Derived Bathymetry (SDB). SDB is a contemporary acquisition technique which derives bathymetric data from multispectral satellite imagery for various hydrographic purposes. The introduction and availability of the chargeless Sentinel-2A multispectral satellite data in the region, has inspired the National Hydrographic Centre (NHC) and Universiti Teknologi Malaysia (UTM) to collaborate in assessing the level of accuracy provided by Sentinel-2A imageries as a data sources for shallow water hydrographic mapping. The assessment is concentrating on both empirical algorithm methods, Lyzenga and Stumpf models in two different seabed topology conditions. Both Lyzenga and Stumpf model produced almost identical results for the area with normal gradient seabed. Conversely, for the rough and irregular seabed surface, results from Lyzenga model was much better compared to Stumpf’s. Lyzenga model has delivered better results where a total of 166,747 (36.23%) of depths samples achieved the IHO minimum survey order standard compared to 103,418 (22.47%) by Stumpf model. Thus, this paper would deliberate the detail findings from the study primarily on the accuracy level of the Sentinel-2A data for SDB hydrographic mapping application over the tropical environmental setting in Malaysia