Here we present a high-resolution landslide susceptibility model for Guiyang County, China, developed to support sustainable disaster risk management. Our approach couples optimized positive and negative training samples with an ensemble of machine-learning algorithms to maximize predictive fidelity. We compiled a georeferenced inventory of 146 landslides by integrating historical records with systematic field validation. Sample optimization was central to our methodology: landslide presence points were refined via buffer-based dilution, and four classifiers—SVM, LDA, RF, and ET—were trained with identical covariate sets to ensure comparability. Three strategies for selecting pseudo-absences—buffering, low-slope filtering, and coupling with the IOE—were benchmarked. The Slope-IOE-O model, which synergizes low-gradient screening with entropy-weighted sampling, yielded the highest predictive capacity (AUC = 0.965). SHAP-based interpretability revealed that slope, monthly maximum rainfall, surface roughness, and elevation collectively dominate susceptibility, with pronounced non-linearities and interactions. Slope contribution peaks at 20–30°, monthly maximum rainfall exhibits a critical threshold near 225 mm, and the synergy between high roughness and road density amplifies landslide risk. Spatially, susceptibility follows a pronounced north–south gradient, with high-hazard corridors aligned along northern and southern mountain belts and the urban core of southern Guiyang County. By integrating rigorously curated training data with robust machine-learning workflows, this study provides a transferable framework for proactive landslide risk assessment, offering scientific support for sustainable land-use planning and resilient development in mountainous regions.

In karst mountainous areas where high-dimensional features coexist with extremely limited sample sizes, accurate landslide susceptibility mapping remains challenging. To address this issue, we propose an ensemble framework termed the Triple-Source Probabilistic Fusion Bidirectional Long Short-Term Memory network (TSPF-BiLSTM). The approach was tested in Sangzhi County, Hunan Province, by integrating three base learners—Random Forest (RF), LightGBM, and AdaBoost. Their raw outputs were first calibrated using five-fold Platt scaling to generate posterior probabilities on a unified scale. A bidirectional LSTM was then employed to perform deep nonlinear fusion of these cross-model probability features. Using a total of 618 landslide and 618 non-landslide samples (split into training and testing sets), the TSPF-BiLSTM model achieved a mean AUC of 0.9525 (±0.0115) under ten-fold cross-validation, outperforming not only the individual base learners but also standalone deep learning models (CNN and Transformer). The frequency ratio in the very high susceptibility zone reached 3.97, significantly exceeding all benchmark models and confirming its superior capability in high-risk area identification. Multi-model importance analysis identified NDVI, elevation, and annual rainfall as the dominant regional landslide predisposing factors. Within the specific ranges of NDVI 0–0.686, elevation 155–462 m, and annual rainfall 1273.6–1301 mm, landslide frequency ratios consistently exceeded 1.96. The proposed framework, with its probability-level fusion and embedded regularization mechanisms, effectively mitigated overfitting despite the small sample size, providing a robust technical solution for geological hazard risk identification and prevention in the data-scarce karst terrain of the Wuling Mountains.

Against the background of global warming, snow cover, as an extremely sensitive and active component of the cryosphere, plays an indispensable role in regulating regional water circulation, energy balance mechanisms and the climate system. To explore the dynamic variation characteristics and driving mechanisms of snow depth in southeastern Xizang, this study took Chamdo City as the research area. Based on multi-source datasets including snow depth, meteorology, vegetation, topography, and population density from 2000 to 2020, methods such as the coefficient of variation, Theil–Sen trend analysis, Mann–Kendall test, Hurst index, and geographical detector were adopted to systematically analyze the spatiotemporal patterns of snow depth variations and their influencing factors. The results indicate that, temporally, the overall snow depth in Chamdo City showed a fluctuating increasing trend over the past 20 years, with an annual growth rate of 0.03 cm. It exhibited distinct characteristics across three stages: snow depth increased at a rate of 0.12 cm·a⁻¹ from 2000 to 2005, decreased at 0.05 cm·a⁻¹ during 2005–2015, and rose rapidly from 2015 to 2020 at a growth rate of 0.52 cm·a⁻¹. Spatially, the distribution of snow depth varied significantly. The extremely shallow snow cover area (≤2 cm) accounted for 51.71% of the total area, primarily concentrated in low-altitude regions with intensive human activities. In contrast, the relatively deep (6–10 cm) and extremely deep (>10 cm) snow cover areas together constituted 14.34% of the total, mainly distributed in high-altitude regions with sparse populations. Hurst index analysis revealed that 61.71% of the study area exhibited persistent changes in snow depth, with a trend toward deepening snow cover in the future. The results from the geographical detector show that air temperature (X9, q = 0.90) was the core driving factor dominating the static spatial differentiation of multi-year average snow depth. Furthermore, the interactions between slope (X4) and air temperature (X9), vegetation type (X6) and air temperature (X9), and population density (X5) and aspect (X8) all demonstrated bivariate enhancement effects, with explanatory power significantly higher than that of individual factors. This study provides a scientific reference for water resource management, snowmelt runoff prediction and snow disaster prevention in Chamdo City.