Answers are grounded in this paper and cite their sources.
Traditional villages in China’s mountainous riparian zones are facing increasing pressure from urbanization and tourism-oriented redevelopment, which requires operational evidence for balancing heritage conservation, ecological protection, and rural revitalization. This study develops a combined, application-oriented workflow that links multi-scale GIS spatial analysis, GeoDetector-based factor detection, Sentinel-2 land cover segmentation, and AHP-based habitat suitability evaluation. The workflow was applied to 388 nationally recognized traditional villages in the Huizhou cultural sphere of southern Anhui Province, covering 15,761 km². Kernel density estimation and exploratory spatial clustering indicate a concentrated village distribution along the Xin’an River corridor. GeoDetector results identify distance to the river network as the strongest explanatory factor (q = 0.573), with nonlinear enhancement among hydrological, vegetation, and topographic variables, including Distance to River Network ∩ NDVI (q = 0.949) and NDVI ∩ Elevation (q = 0.960). The adapted ConvNeXt multi-attention segmentation model achieved mIoU = 84.3% and OA = 93.8%, outperforming the evaluated convolutional, Transformer-based, and MAE-based baseline models for five-class land cover extraction. Spatial cross-validation was used as an auxiliary consistency check between GIS-derived suitability patterns and deep learning-derived land cover information, rather than as a direct validation of either output alone. The High Suitability zone covers 3,940 km² (25.0% of the study area) and contains 68.8% of all villages, indicating strong over-representation of traditional villages in environmentally favorable valley landscapes. The results provide planning implications for zone-differentiated rural revitalization and heritage landscape management in southern Anhui.
No takes yet. Share an insight, caveat, or question.
Wang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: