A Method for Extracting Tidal Flat Elevation by Integrating Land Submersion Probability Extremes and Geo-Informatic Tupu Analysis with Application to the Yancheng Coastal Zone

Abstract Elevation data constitutes a critical parameter for coastal dynamic monitoring. This study innovatively develops a method for extracting tidal flat elevation from 2D remote sensing imagery, integrating spectral-spatial pattern analysis (Geo-Informatic Tupu Analysis) to establish a transferable technical framework for spatiotemporal change monitoring. Using Landsat series imagery (1984–2024) as primary data with Sentinel-2 supplementation, we conducted 40-year continuous monitoring of Jiangsu Yancheng's coastal zone. Key findings reveal: (1) The land submersion probability extremes method effectively extracts mean high/low water lines (MHWL/MLWL), achieving ± 0.45m accuracy when combined with vegetation front boundary (VFB) as elevation benchmarks; (2) Dynamic analysis shows MHWL prograded seaward at 88.7 ± 1.18 m/yr while MLWL retreated landward at 41.83 ± 1.18 m/yr, forming convergent evolution patterns; (3) Spatial differentiation demonstrates accretion in southern sectors versus erosion in northern sectors, with net tidal flat loss at 9.2 km²/yr, exhibiting distinct "southern accretion-northern erosion" trends. The proposed "elevation extraction-terrain inversion-pattern analysis" framework enables precise historical reconstruction of coastal evolution, providing quantitative foundations for integrated coastal management.