Gas station sites pose potential risks of soil and groundwater contamination, which not only threatens public health and property but may also damage the assets and reputation of businesses and government entities. Given the complex nature of soil and groundwater contamination at gas station sites, this study utilizes field data from basic and environmental information, maintenance information for tank and pipeline monitoring, and environmental monitoring to develop machine learning models for predicting potential contamination risks and evaluating high-impact risk factors. The research employs three machine learning models: XGBoost, LightGBM, and Random Forest (RF). To compare the performance of these models in predicting soil and groundwater contamination, multiple performance metrics were utilized, including Receiver Operating Characteristic (ROC) curves, Precision-Recall graphs, and Confusion Matrix (CM). The Confusion Matrix analysis revealed the following results: accuracy of 85.1-87.4 %, precision of 86.6-88.3 %, recall of 83.0-87.2 %, and F1 score of 84.8-87.8 %. Performance ranking across all metrics consistently showed: XGBoost > LightGBM > RF. The area under the ROC curve and precision-recall curve for the three models were 0.95 (XGBoost), 0.94 (LightGBM), and 0.93 (RF), respectively. While all three machine learning approaches demonstrated satisfactory predictive capabilities, the XGBoost model exhibited optimal performance across all evaluation metrics. This research demonstrates that properly trained machine learning models can serve as effective tools for environmental risk assessment and management. These findings have significant implications for decision-makers in environmental protection, enabling more accurate prediction and control of contamination risks, thereby enhancing the preservation of ecological systems, public health, and property security.

This study investigates the application of machine learning (ML) algorithms for seismic damage classification of bridges supported by helical pile foundations in cohesive soils. While ML techniques have shown strong potential in seismic risk modeling, most prior research has focused on regression tasks or damage classification of overall bridge systems. The unique seismic behavior of foundation elements, particularly helical piles, remains unexplored. In this study, numerical data derived from finite element simulations are used to classify damage states for three key metrics: piers' drift, piles' ductility factor, and piles' settlement ratio. Several ML algorithms, including CatBoost, LightGBM, Random Forest, and traditional classifiers, are evaluated under original, oversampled, and undersampled datasets. Results show that CatBoost and LightGBM outperform other methods in accuracy and robustness, particularly under imbalanced data conditions. Oversampling improves classification for specific targets but introduces overfitting risks in others, while undersampling generally degrades model performance. This work addresses a significant gap in bridge risk assessment by combining advanced ML methods with a specialized foundation type, contributing to improved post-earthquake damage evaluation and infrastructure resilience.

青藏高原作为中低纬度地区最大的高山冻土区，多年冻土和季节冻土广泛分布。高精度的地表冻融监测结果对研究该区域的水热交换、碳氮循环和土壤冻融侵蚀非常重要。本文基于4个青藏高原典型地区的土壤温湿度观测网数据，开展利用LightGBM算法和随机森林算法进行土壤冻融循环监测的研究。在构建土壤冻融监测模型的过程中，发现土壤湿度是影响冻融判别的一个关键因子。使用AMSR2亮温数据和ERA5-Land土壤湿度数据，基于两种机器学习算法判别地表冻融状态，将结果与传统冻融判别式算法进行对比分析。结果表明：相比冻融判别式算法，LightGBM算法在白天和夜间的总体判对率提高了12.09%;14.45%，随机森林算法在白天和夜间的总体判对率提高了13.23%和14.96%。近80%的错分样本分布在-4.0℃～4.0℃之间，说明2个机器学习算法能够识别出稳定的土壤冻结状态和融化状态。另外，LightGBM算法和随机森林算法得到的日冻融转换天数的平均RMSE降低了112.82和117.00；冻结天数的平均RMSE降低了47.87和53.96；融化天数的平均RMSE降低了37.10和39.80。同时，基于随机森林算...