Recurrent Overlap Between Tourism Suitability and Ecological Sensitivity in the Yellow River Basin

Large river basins require spatial evidence showing where tourism-support conditions and ecological constraints coincide. Using the Yellow River Basin, we built a reproducible workflow that combines Google Earth Engine preprocessing with local spatial analysis to map tourism development suitability (TDSI), ecological sensitivity (ESI), and conflict intensity on a 1 km equal-area grid in 2013, 2018, and 2023. Annual hotspot shares remained near one-fifth of valid cells, and 153,958 cells (19.80%) entered the hotspot set in at least two benchmark years. Spatially matched reference checks were mixed: official tourism cells showed a positive high-suitability contrast against 50 km block-matched cells (0.778 versus 0.563; empirical p=0.043), whereas OSM POIs, the named-attraction subset, and GBIF cells showed no clear geography-adjusted enrichment. Aggregation from 1 km to 2 km preserved broad structure, but weighting and ecological-constraint choices changed local membership materially. The full XGBoost–SHAP model identified precipitation variability, land-surface thermal change, protected-area proximity, relief, and road-network intensity as leading internal correlates; a strict two-covariate model achieved AUC = 0.680 and F1 = 0.538, limiting external mechanism claims. The products support uncertainty-aware regional screening. Site-specific approval and long-term persistence claims require finer regulatory, field, and temporal evidence.