Supporting Geospatial Dataset for Comparative and Explainable Deep Learning-Based Landslide Susceptibility Mapping in Hualong County, Qinghai Province, China

This supporting geospatial dataset accompanies the manuscript “Comparative and Explainable Deep Learning for Landslide Susceptibility Mapping in a Complex Plateau Transition Zone”. It contains 16 GeoTIFF raster files for Hualong Hui Autonomous County, Qinghai Province, China: 12 landslide conditioning-factor rasters (elevation, aspect, slope, plan curvature, NDVI, lithology, land use, distance to faults, distance to roads, distance to rivers, Topographic Wetness Index, and Stream Power Index) and four model-derived landslide susceptibility maps generated using Logistic Regression, Random Forest, a Convolutional Neural Network, and a CNN-LSTM model. The conditioning-factor rasters were converted, projected, clipped, resampled, and spatially aligned to a nominal spatial resolution of 30 m. The dataset supports inspection of the geospatial model inputs, map-level comparison of the four modelling approaches, and secondary spatial analysis of landslide susceptibility patterns in a complex plateau transition zone. Detailed file descriptions, methodological context, citation information, and reproducibility limitations are provided in the accompanying README.md and FILE_MANIFEST.csv.