基于普适型降雨监测设备的江西省滑坡灾害降雨阈值分析
Analysis of Rainfall Thresholds for Landslides in Jiangxi Province Based on Universal Rainfall Monitoring Instruments
杨连伟 1黄传胜 2李华 1李鹏 2欧阳昊明2
作者信息
- 1. 江西省地质调查勘查院地质环境监测所,330006,南昌
- 2. 东华理工大学土木与建筑工程学院,330013,南昌
- 折叠
摘要
降雨是诱发山体滑坡的重要因素之一.以江西省 416 起降雨型滑坡为研究对象,充分考虑前期有效降雨对滑坡的触发作用,基于群专结合监测预警点雨量计数据构建了累计降雨量-降雨历时阈值模型和降雨强度-降雨历时阈值模型,采用概率分级方法建立了不同的预警等级,并验证了模型可靠性.结果表明,江西省降雨型滑坡主要由 10 d内降雨引发,前期降雨对滑坡的发生有着显著影响;可以按 5%、20%、50%、80%滑坡发生概率确定降雨阈值曲线,采用 4 个预警等级对滑坡灾害进行监测预警;降雨强度-降雨历时阈值模型预测精度更高,更适用于江西省降雨型滑坡监测预警.研究成果能够为政府部门在滑坡预警方面提供指导,有效降低滑坡的危害性.
Abstract
Rainfall events are oftentimes the important triggering cause for landslides.In this paper,the rainfall information of 416 rainfall-induced landslides in Jiangxi province was analyzed,consid-ering the role of previous effective rainfall.Based on these datasets,accumulated precipitation-rain-fall duration(E-D)thresholds and mean rainfall intensity-rainfall duration(I-D)thresholds for different probabilities of occurrence,corresponding to various early warning levels,were calculated using a Frequentist method,and the model reliability was verified.The following results are ob-tained:previous rainfall has a significant impact on the occurrence of landslide,and the rainfall-in-duced landslides in Jiangxi province are mainly triggered by rainfall within 10 days.Four early warn-ing levels can be used to monitor landslide disasters,and the rainfall threshold curves can be deter-mined according to the occurrence probabilities of 5%,20%,50%and 80%,respectively.The mean rainfall intensity-rainfall duration(I-D)threshold model has higher prediction accuracy and is more suitable for rainfall-induced landslide early warning in Jiangxi province.The research can provide guidance for government departments in landslide early warning and effectively reduce the harm of landslides.
关键词
滑坡/累计降雨量/降雨历时/监测预警/阈值模型Key words
landslide/accumulated precipitation/rainfall duration/monitoring warning/threshold model引用本文复制引用
基金项目
江西省自然科学基金(20202BABL214047)
东华理工大学博士科研启动基金(DH-BK2019240)
东华理工大学博士科研启动基金(DHBK2019243)
出版年
2024
国家科技期刊平台
NSTL
万方数据