Scholarlink

<p>Predictive frameworks for rainfall-induced landslides remain underdeveloped compared to static sensitivity evaluations. This study evaluates spatiotemporal landslide sensitivity in Xinning County, China, by integrating Certainty Factor (CF) modeling with a Random Forest algorithm. Maximum daily rainfall intensities for 5-, 10-, 30-, 50-, and 100-year return periods were simulated using Gumbel extreme value theory, CHIRPS data and Google Earth Engine. Among ten influence factors, rainfall was identified as the dominant external driver. The proportion of very high sensitivity (VHS) areas in the county ranged between 3.2% and 12.9%, with the lowest value recorded in 2023 and the peak observed under the 10-year return period. Decadal-scale expansion of VHS areas in GQ/MTQ/HLS towns is evident, with the extent of the 10-year return period doubling that of other return intervals. Spatially, high/very high sensitivity (HVHS) areas in central/northern lowlands align with median maximum daily rainfall, contrasting with low/very low sensitivity (LVLS) areas in southeastern/southwestern highlands linked to higher values. Proactive mitigation in GQ/MTQ/HLS Towns is urged before the 10-year return period.</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

Sensitivity evaluation and prediction of extreme rainfall-induced landslides in Xinning, Hunan, China, using Gumbel extreme value theory and random forest model

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月4日

查看全文
  • Yan Yu,
  • Zhengqing Wang,
  • Yulin Chen,
  • Tianhao Zhao,
  • Dinghao Zhang,
  • Lingjie Meng,
  • Yongfei Che,
  • Yan Zou,
  • Haiyang He,
  • Lijia Cao,
  • Shan Liu,
  • Kaituo Shi,
  • Qinglin Sui,
  • Hongsheng Liu

摘要

Predictive frameworks for rainfall-induced landslides remain underdeveloped compared to static sensitivity evaluations. This study evaluates spatiotemporal landslide sensitivity in Xinning County, China, by integrating Certainty Factor (CF) modeling with a Random Forest algorithm. Maximum daily rainfall intensities for 5-, 10-, 30-, 50-, and 100-year return periods were simulated using Gumbel extreme value theory, CHIRPS data and Google Earth Engine. Among ten influence factors, rainfall was identified as the dominant external driver. The proportion of very high sensitivity (VHS) areas in the county ranged between 3.2% and 12.9%, with the lowest value recorded in 2023 and the peak observed under the 10-year return period. Decadal-scale expansion of VHS areas in GQ/MTQ/HLS towns is evident, with the extent of the 10-year return period doubling that of other return intervals. Spatially, high/very high sensitivity (HVHS) areas in central/northern lowlands align with median maximum daily rainfall, contrasting with low/very low sensitivity (LVLS) areas in southeastern/southwestern highlands linked to higher values. Proactive mitigation in GQ/MTQ/HLS Towns is urged before the 10-year return period.