A Review of Advancements and Applications of Satellite-Derived Bathymetry
Abstract
Satellite-derived bathymetry (SDB) has revolutionized underwater mapping by using remote sensing techniques to estimate water depth with high spatial coverage and cost-effectiveness. This paper reviews the evolution of bathymetric methods from traditional ruler-based measurements to modern acoustic and satellite-based approaches. Different remote sensing technologies, including Light Detection and Ranging (LiDAR), multispectral, and altimetric bathymetry, are reviewed, emphasizing their advantages and limitations. The performance of SDB in various environments, including coastal zones, coral reefs, inland water bodies, and glacial regions, is analyzed, with a focus on how water transparency, spectral band selection, and machine learning techniques impact accuracy. Additionally, new deep learning models, including convolutional neural networks (CNNs) and U-Net architectures, are explored in terms of their potential to enhance bathymetric mapping. Although SDB provides significant advances in mapping capabilities, challenges, such as optical limitations in turbid waters and seasonal variability, require hybrid approaches that combine multiple sensing modalities. This study emphasizes the role of artificial intelligence in refining bathymetric estimates and highlights future research directions, including the combination of hyperspectral imagery and radar.