Abstract: | Water depth is an important marine factor, and satellite derived bathymetry is an important means to obtain shallow water depth. At present, the research and application of satellite derived bathymetry are mainly based on foreign satellite image, while the research and application of domestic satellite image are few. This study, based on the image of GF-6 and Nanshan Port of Sanya as the research area, single-band regression model, double-band ratio model and multi-band regression model were established respectively to study the satellite derived bathymetry capability of multi-spectral image. The satellite derived bathymetry results of GF-6 remote sensing image and Sentinel-2 remote sensing image were compared experimentally. Results indicate that the multi-band model shows the best effect, the dual-band model the second, and the single-band model the last. For the single-band model, the satellite derived bathymetry ability order from the best to the worst is: green band, blue band, red band, and near infrared band. The satellite derived bathymetry accuracy of GF-6 remote sensing data is roughly the same as that of Sentinel-2 remote sensing image, which indicates that GF-6 has the ability to replace foreign remote sensing image for satellite derived bathymetry and has the potential of large-scale application. In this paper, the research and results of the bathymetric inversion ability of the remote sensing image of GF-6 satellite will provide a useful reference for the research and application of the bathymetric inversion of domestic GaoFen satellite data. |