Debris Flow Hazard-Causing Factors Dataset for the Yunnan Section of the Nujiang River Basin, China

This dataset contains hazard-causing factors for debris flow prediction using deep learning models. The data were collected from the Yunnan Section of the Nujiang River Basin, China, a region characterized by complex geological conditions and high debris flow susceptibility. The dataset encompasses topographical, hydrological, geological, meteorological, and vegetation-related factors derived from multi-source data integration, including SRTM DEM (30m resolution), meteorological station observations, and field investigations. The dataset comprises 159 debris flow-prone gullies with 15 influencing factors systematically selected through multicollinearity diagnosis. Four variables (X10, X14, X17, X19) were excluded from the original 19 candidate factors due to multicollinearity (Variance Inflation Factor > 5.0) to ensure model stability and parameter precision. The variable numbering is preserved for consistency with the research methodology. Variable Code Description Unit Data Type No Sample ID - Integer X1 Maximum 24-hour Rainfall mm Float X2 Annual Average Rainfall 10³mm Float X3 Annual Average Temperature °C Float X4 Annual Average Wind Speed m/s Float X5 Annual Average Vapor Pressure hPa Float X6 Annual Average Relative Humidity % Float X7 Potential Source Energy 10⁹ J/m² Float X8 Catchment Area km² Float X9 Average Gradient of Main Channel - Float X11 Length of Main Channel km Float X12 Sinuosity of Main Channel - Float X13 Drainage Density km/km² Float X15 Vegetation Cover Rate % Float X16 Average Slope ° Float X18 Lithological Impact - Float