Enhanced spatiotemporal fusion algorithm for long-term monitoring of intertidal zone topography

Global climate change and human activities exert dual pressures on intertidal zones, leading to significant alterations in their spatiotemporal distribution. Traditional topographic data collection methods in these regions are labor-intensive and resource-demanding, requiring considerable human and material inputs, and are often limited to localized areas. These challenges hinder large-scale, comprehensive monitoring efforts. In response, this study presents an improved Flexible Spatiotemporal Data Fusion model, tailored to account for the inherent spatiotemporal heterogeneity of intertidal zone topography. By integrating high-frequency modulation algorithms, we optimized the distribution of residuals, enhancing the overall accuracy and stability of the spatiotemporal model. We applied this enhanced algorithm to the central Jiangsu coastal tidal flats, where we successfully inverted the Digital Elevation Model (DEM) of the intertidal wetlands. The fused imagery produced by the algorithm demonstrated superior performance in quantitative assessments, as confirmed by the results of the All-round Performance Assessment. These results consistently outperformed alternative algorithms in terms of accuracy and resolution. Through interannual analysis of the generated DEM, we observed a clear trend: the area and volume of the intertidal zone first expanded and then contracted. This pattern is attributed to reduced sediment supply and localized human activities, which have altered the natural sedimentation and erosion dynamics. The improved algorithm not only provides a reliable method for simulating intertidal topography but also enables the monitoring of sedimentation and erosion in regions lacking direct observational data. This study addresses key challenges faced by satellite remote sensing technologies in accurately capturing intertidal zone topography, particularly in dynamic coastal environments, and offers a new approach for large-scale, long-term monitoring of these ecologically significant areas.