Scale-Dependent Associations of SAR-Derived Apparent Inundation and Hydrogeomorphic Setting with Salt-Marsh Vegetation Variability in the Minjiang Estuary, Southeastern China

Salt-marsh vegetation is jointly influenced by episodic hydrological exposure and the hydrogeomorphic setting in which inundation occurs, yet their roles may differ across temporal and spatial scales. Here, we examined these scale-dependent associations in the Minjiang Estuary, southeastern China, using 1127 fixed sampling sites monitored quarterly from 2018 to 2025. Sentinel-1 SAR, Sentinel-2 optical imagery, regional hydroclimatic data, and hydrogeomorphic covariates were integrated to quantify seasonal vegetation anomalies and multi-year spatial variability. Quarters with SAR-derived apparent inundation had EVI anomalies 0.0544 units lower than quarters without detected inundation (95% CI: −0.0624 to −0.0465; p < 0.001), equivalent to approximately 0.75 times the overall standard deviation of quarterly EVI anomalies (SD = 0.0726), and the negative association persisted across the major sensitivity analyses. By contrast, repeated inundation detections provided limited additional information beyond the occurrence signal under the current satellite-observation framework. At the spatial scale, adding elevation, slope, and distance to persistent water increased the spatial cross-validated XGBoost R2 from 0.143 to 0.360, whereas the available anthropogenic-context variables did not provide a stable detectable additional predictive gain beyond the hydrogeomorphic model. These results indicate that SAR-detected inundation primarily marks quarter-scale departures from normal seasonal vegetation condition, whereas persistent hydrogeomorphic setting better organizes multi-year spatial variability. Distinguishing these complementary dimensions provides a scale-explicit framework for monitoring vegetation dynamics and assessing ecosystem variability in morphologically active coastal wetlands.