This study investigated the influence of non-landslide sampling strategies on landslide susceptibility assessment (LSA) performance and explored approaches to minimizing uncertainty in model selection. Five non-landslide sampling strategies were evaluated using the random forest (RF) model to generate landslide susceptibility maps (LSMs) for each scenario. To assess the impact of these strategies, this study employed a receiver operating characteristic (ROC) curve, a confusion matrix, and various statistical indicators. Additionally, the mean susceptibility indices derived from the gradient boosting decision tree (GBDT), support vector machine (SVM), and RF models were analyzed to evaluate their effectiveness in reducing the uncertainty during model selection. The GBDT, SVM, and RF were selected for their ability to handle complex, nonlinear relationships in the data, superior generalization capability, effective mitigation of overfitting risks, high predictive performance, and robustness. The findings revealed that selecting non-landslide samples from slope units without landslides enhances accuracy and averaging across models mitigated the uncertainty associated with landslide susceptibility models. Furthermore, this study demonstrated that the non-landslide sample selection method significantly improved prediction accuracy, particularly when samples were drawn from very-low-susceptibility zones identified by pre-classified machine learning models. These results highlight the importance of refining sample selection strategies and integrating multiple machine learning models to improve the reliability and accuracy of landslide susceptibility assessments. This approach provides valuable insights for future research and practical applications in risk mitigation and disaster management by offering a more precise depiction of low-susceptibility areas, thereby reducing the occurrence of false positives in landslide prediction.

A substantial proportion of long-distance oil and gas pipelines in China traverse active faults and high-risk areas characterised by intricate topographic and geological environments. These pipelines are susceptible to a range of safety concerns, exacerbated by the increasing frequency of strong earthquakes in recent years. To address this issue, a comprehensive risk investigation framework has been proposed for long-distance oil and gas pipelines following seismic events. This initiative aims to ensure the safety of pipeline transportation. In this paper, the elements of pipeline safety evaluation under the influence of coseismic hazards are first organized, followed by a construction of post-earthquake pipeline safety rapid assessment theoretical framework based on the seismic geological disaster risk evaluation system. Each method in the system is then introduced one by one. Unlike existing studies that predominantly focus on localized fault activity or static risk assessment, our framework introduces three key innovations: (1) a hierarchical integration of multi-source monitoring data (SCADA, UAV-AI, and numerous monitoring devices) into a unified GIS platform, overcoming the fragmentation of existing systems; (2) a dynamic four-step evaluation process (susceptibility → hazard → risk → safety) that incorporates both pre-earthquake geological conditions and post-earthquake real-time triggers (e.g., PGA, rainfall); (3) a novel risk matrix mechanism for pipeline safety, which dynamically updates risk levels based on field monitoring data rather than relying solely on probabilistic models. This study provides a novel theoretical framework for assessing the safety of pipelines after earthquakes, which can provide a timely basis for pipeline management decisions and reduce the potential damage to pipelines caused by earthquakes. It is important to note that this framework is still in a preliminary stage and needs to be continuously deepened and optimised.