Scholarlink

<p>Against accelerating global climate change, clustered landslide hazards triggered by extreme precipitation have become increasingly frequent, posing significant challenges to regional hazard prevention and mitigation. Taking the 25 July 2013 extreme rainfall–induced landslide cluster event in China’s Western Qinling Mountains as a case study, this research proposes a critical rainfall threshold determination method based on landslide density-probability analysis, departing from traditional intensity-duration (I-D) and effective rainfall-duration (E-D) curve approaches. Based on multi-temporal rainfall spatial analysis, Pearson correlation analysis was employed to investigate the spatiotemporal relationships between landslides and precipitation processes. Two characteristic rainfall parameters were identified: antecedent effective rainfall (3-day antecedent effective rainfall with <i>α</i> = 0.8) and current triggering rainfall (5-h antecedent triggering rainfall), which exhibited Pearson correlation coefficients of 0.486 and 0.646 with landslide kernel density, respectively—significantly higher than those of non-characteristic parameters. Nonlinear models ExpDec1 and Gauss were then developed (<i>R</i><sup>2</sup> = 0.987 and 0.921, respectively). Using probability distribution functions, four-tiered critical rainfall thresholds were established by calibrating against cumulative probability of landslide density. For example, the extremely high alert level corresponds to ≥ 97.8&#xa0;mm antecedent effective rainfall and ≥ 38.4&#xa0;mm triggering rainfall. Building upon this, a comprehensive meteorological alert matrix was developed by integrating antecedent effective and current triggering rainfall threshold ranges, quantifying the interactions between multi-temporal-scale precipitation. Results demonstrate that this methodology effectively addresses limitations of conventional threshold models, including dependence on historical stochastic events and subjectivity in parameter selection. The probabilistic alert classification aligns closely with the physical mechanisms of extreme precipitation-induced landslide clustering, thereby providing a technical framework and conducting empirical demonstration for dynamic regional geological hazard forecasting and risk management. It is expected to improve the traditional fixed-time rainfall warning model.</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

Critical rainfall thresholds for landslides based on extreme rainfall–induced clustered landslides and characteristic rainfall parameter analysis: a case study in Western Qinling Mountains, China

  • 期刊论文

  • 发表日期：

    2026年2月16日

查看全文
  • Shuai Liu,
  • Tao Wang,
  • Gulian Xing,
  • Shuai Zhang,
  • Jiamei Liu,
  • Yongjun Zhang,
  • Hailong Chen

摘要

Against accelerating global climate change, clustered landslide hazards triggered by extreme precipitation have become increasingly frequent, posing significant challenges to regional hazard prevention and mitigation. Taking the 25 July 2013 extreme rainfall–induced landslide cluster event in China’s Western Qinling Mountains as a case study, this research proposes a critical rainfall threshold determination method based on landslide density-probability analysis, departing from traditional intensity-duration (I-D) and effective rainfall-duration (E-D) curve approaches. Based on multi-temporal rainfall spatial analysis, Pearson correlation analysis was employed to investigate the spatiotemporal relationships between landslides and precipitation processes. Two characteristic rainfall parameters were identified: antecedent effective rainfall (3-day antecedent effective rainfall with α = 0.8) and current triggering rainfall (5-h antecedent triggering rainfall), which exhibited Pearson correlation coefficients of 0.486 and 0.646 with landslide kernel density, respectively—significantly higher than those of non-characteristic parameters. Nonlinear models ExpDec1 and Gauss were then developed (R2 = 0.987 and 0.921, respectively). Using probability distribution functions, four-tiered critical rainfall thresholds were established by calibrating against cumulative probability of landslide density. For example, the extremely high alert level corresponds to ≥ 97.8 mm antecedent effective rainfall and ≥ 38.4 mm triggering rainfall. Building upon this, a comprehensive meteorological alert matrix was developed by integrating antecedent effective and current triggering rainfall threshold ranges, quantifying the interactions between multi-temporal-scale precipitation. Results demonstrate that this methodology effectively addresses limitations of conventional threshold models, including dependence on historical stochastic events and subjectivity in parameter selection. The probabilistic alert classification aligns closely with the physical mechanisms of extreme precipitation-induced landslide clustering, thereby providing a technical framework and conducting empirical demonstration for dynamic regional geological hazard forecasting and risk management. It is expected to improve the traditional fixed-time rainfall warning model.