A virtual coastal band-driven optimization-based bathymetric approach (VOBA) for optically shallow water with multispectral imagery

Bathymetry mapping is important for detecting underwater topography. For optically shallow waters, the use of satellite-based high-spatial-resolution imagery (Such as Gaofen-2, Ziyuan-3, Worldview-2) and historical medium-spatial-resolution imagery (Landsat-5 thematic mapper (TM) and Landsat-7 enhanced thematic mapper plus (ETM+)) has great potential for detailed bathymetric information mapping and monitoring bathymetric temporal variations, respectively. However, some of the above-mentioned imagery provides only three visible bands (blue, green and red bands) for bathymetry mapping, such as Gaofen-2, Ziyuan-3, TM, ETM+. This limitation restricts the accuracy of optimization-based bathymetric approaches, which can estimate water depth without in situ bathymetric data. Therefore, a simple linear relationship was established using in situ reflectance data of optically shallow waters from four different sites in China to derive a virtual coastal band (~445 nm) of Landsat-8 from the adjacent blue band. Then, the virtual coastal band, along with the original three visible bands, was employed to estimate water depth using the optimization-based bathymetric approach. The new approach, called the virtual coastal band-driven optimization-based bathymetric approach (VOBA), was validated with in situ reflectance data, 3 simulated reflectance data sets, and 18 satellite-based images (6 high-spatial-resolution images of Gaofen-2, Ziyuan-3, and Worldview-2 for 4 study sites in China and 12 TM or ETM+ images for 12 different sites worldwide) that cover similar optically shallow waters. The comparison results indicate that VOBA can significantly improve the accuracy of the estimated water depth, especially for optically shallow water at depths greater than 5 m. Additionally, we showcased the potential of the VOBA for monitoring underwater topographic temporal changes by applying it to TM data (1987–2011) for two sites in China.