The detection of seafloor signal photons in various topographies is challenging. Previous research has divided photons into clusters based solely on their density, which is closely related to the settings of the empirical parameters. Inappropriate parameters may mistakenly identify the water column noise photons as seafloor photons. To overcome these limitations, this study introduces a novel slope iterative adaptive filter (SIAF) method that innovatively integrates ICESat-2 ATL03 photon data with Sentinel-2-derived topographic slopes. Inspired by satellite-derived bathymetry, we extracted topographic slopes from multispectral images as auxiliary information to guide the photon extraction. The initial slope estimation was derived from the multispectral images, and the optimal slope direction was determined iteratively, using the detected signal photons in each step. The average and maximum overall accuracies of SIAF were 93.43% and 95.7%, respectively. The validation of the extraction results with sonar data indicated that the SIAF achieved an average root mean square error (RMSE) of 0.49 m. Crucially, the SIAF resolves critical shortcomings of prior techniques: (1) it avoids the isotropic assumption of density-based methods, (2) it mitigates AVEBM’s vulnerability to noise in steep-slope regions, and (3) it enables robust automation without manual parameter tuning. Consequently, SIAF proved to be an efficient approach for the automatic mapping of water depths in shallow-water zones.
Li et al. (Sun,) studied this question.