Mass Movement Risk Assessment in the Loess Hilly Region of Northwest China Using a Weighted Information Theoretic Framework

Ground instability represents a major environmental hazard in the Loess Hilly region of Northwest China, threatening infrastructure and human safety. This study establishes an integrated information-theoretic framework for evaluating regional instability risk by coupling the information value model with analytic hierarchy process (AHP) weighting and subsequent hazard–exposure synthesis. Seven conditioning factors—geomorphic type, slope, aspect, lithology, distance to faults, river system, and NDVI—were analyzed to derive susceptibility, while rainfall, peak ground acceleration, and human engineering activity were incorporated as triggering elements of hazard. Exposure was quantified from population density and infrastructure exposure, and overall risk was defined as the product of hazard and exposure after normalization and calibration. Results indicate that hilly landforms, slopes of 10–20°, and NDVI values between 0.3 and 0.6 are the dominant controls on instability occurrence. Extreme-risk zones are concentrated in central Guyuan and northwest Shizuishan (0.16% of the study area), with high-risk zones covering 21.87%, moderate-risk zones covering 41.65%, and low-risk zones covering 6.32%. Model validation yields an AUC of 0.833 and a consistent increase in observed disaster-point density from low to extreme classes, confirming strong predictive reliability. These results demonstrate that the proposed calibrated framework provides a practical and transferable tool for ground-instability risk assessment and land-use planning in loess terrains.