Comparison of Intertidal Terrain Extraction Methods Based on ICESat‐2 and Tidal Data

The intertidal zone is a transitional area between land and sea, characterized by both marine and terrestrial features, with rich resources in mudflats. Accurately mapping the intertidal topography and understanding its dynamic characteristics are of great significance. In this study, Sentinel-2 imagery was used, combined with tidal data and ICESat‐2 data, respectively. Four methods–the waterline method, the inundation frequency method, random forest regression, and the long short-term memory (LSTM) model–were applied to extract intertidal topography in the large radial sand ridges along the Jiangsu coast. When validated with ICESat‐2 data and unmanned aerial vehicle (UAV) data, the root mean square error (RMSE) values of all four methods combined with ICESat‐2 data were lower than those combined with tidal data. Using ICESat‐2 data for validation, the waterline method combined with ICESat‐2 data achieved the lowest RMSE of 0.218 m. When validated with UAV data, the inundation frequency method combined with ICESat‐2 data yielded the lowest RMSE of 0.864 m. From 2020 to 2024, the intertidal zone in this region was initially dominated by erosion, followed by deposition, ultimately reaching a dynamic equilibrium. This study achieved two objectives: (1) under identical area conditions using the same image data and two elevation data, four different methods were validated to compare their topographic extraction performance and identify the optimal approach; and (2) the optimal method was applied to generate multi-temporal topographic results of a local intertidal zone along the Jiangsu coast, analyzing terrain change in the region.