Scholarlink

<p>Red-bed rocks typically exhibit low softening coefficients, which directly affect rock mass strength and stability. Secondary and lower-level highways located in this environment are widely distributed, and due to the lack of comprehensive slope support, their slopes are prone to geological disasters. Therefore, identifying and detecting potentially hazardous slopes along secondary and lower-grade highways in red-bed environments is a critical issue in geological disaster prevention and mitigation. This study focused on the construction project of the Pingma to Shaping section of Provincial Highway S210 in Guangxi, China, as the research area, investigating landslides, slope failures, and unstable highway slopes within the region. The SBAS-InSAR technique was employed to monitor surface deformation in the study area. Nine types of data were chosen as influencing factors for geological hazard susceptibility, and their information values were calculated based on their characteristics. The Logistic Regression and Random Forest (RF) models were selected to evaluate and predict the susceptibility of geological hazards. The results indicate that the annual average deformation rate of 90% of the points in the study area ranges between − 32&#xa0;mm/year and 28&#xa0;mm/year. The most effective model for the study area is the Logistic regression model. By integrating the results of both InSAR and machine learning methods, seven high-risk slopes along the S210 highway were identified, Field investigations confirmed that one of these slopes had recently experienced a landslide. Our study demonstrates that the integration of InSAR and machine learning methods enables complementary advantages in achieving accurate hazard assessment of highway slopes. This research provides a scientific reference for slope safety monitoring in road engineering and enhances the capability of geological safety assessment and emergency management for highway slopes.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

InSAR and machine learning-based red beds highway slope hazard assessment in Southern China

  • 期刊论文

  • 发表日期：

    2025年8月6日

查看全文
  • Jiayu Liang,
  • Zhibin Liu,
  • Fanxing Meng,
  • Chenghua Xu,
  • Guoyi Tang,
  • Yinjuan Sun

摘要

Red-bed rocks typically exhibit low softening coefficients, which directly affect rock mass strength and stability. Secondary and lower-level highways located in this environment are widely distributed, and due to the lack of comprehensive slope support, their slopes are prone to geological disasters. Therefore, identifying and detecting potentially hazardous slopes along secondary and lower-grade highways in red-bed environments is a critical issue in geological disaster prevention and mitigation. This study focused on the construction project of the Pingma to Shaping section of Provincial Highway S210 in Guangxi, China, as the research area, investigating landslides, slope failures, and unstable highway slopes within the region. The SBAS-InSAR technique was employed to monitor surface deformation in the study area. Nine types of data were chosen as influencing factors for geological hazard susceptibility, and their information values were calculated based on their characteristics. The Logistic Regression and Random Forest (RF) models were selected to evaluate and predict the susceptibility of geological hazards. The results indicate that the annual average deformation rate of 90% of the points in the study area ranges between − 32 mm/year and 28 mm/year. The most effective model for the study area is the Logistic regression model. By integrating the results of both InSAR and machine learning methods, seven high-risk slopes along the S210 highway were identified, Field investigations confirmed that one of these slopes had recently experienced a landslide. Our study demonstrates that the integration of InSAR and machine learning methods enables complementary advantages in achieving accurate hazard assessment of highway slopes. This research provides a scientific reference for slope safety monitoring in road engineering and enhances the capability of geological safety assessment and emergency management for highway slopes.