Bathymetry monitoring of shallow coastal environment using remote Sensing data

Abstract

The Caspian Sea (CS), the largest lake in the world, is an exceptional water body that is well-known for its rapid sea level (SL) change. This study aims to monitor the bathymetry of southeastern CS and Gorgan Bay (GB) under rapid sea level change. For this purpose, the water depth was extracted from the visible bands (blue, green, and red) of Landsat images using the Random Forest (RF) algorithm. The bathymetry survey results were validated through hydrography (field) and altimetry data. The results of bathymetry showed that R2 and RMSE values were 0.996 and 0.106 m, respectively in 2016. Moreover, the predicted depth for the period between 1992 and 2016 was consistent with the altimetry data. Due to rapid sea level fall, shallow coastal features like GB need to be monitored continuously. Therefore, the GB water depth was predicted by the RF algorithm for 2001 and 2023. The results showed that the GB bathymetry map corresponds closely to the altimetry data. The study also emphasizes the need for remote sensing data to prepare bathymetry maps which are hardly possible through field surveys.