基于非线性回归的暴雨泥石流堵河预测
River blockage prediction of rainfall-induced debris flow based on nonlinear regression
刘人文 1周伟 1周亚萍 1任文铭 1聂浩文1
作者信息
- 1. 成都理工大学 地质灾害防治与地质环境保护国家重点实验室,四川 成都 610059
- 折叠
摘要
堵河型泥石流灾害在汶川震区较为常见,其链式灾害效应严重威胁到了区域内人民的生命财产安全.故准确预测泥石流堵河程度,并采取相应的防治措施是至关重要的.以四川省绵竹市清平镇于 2010 年 8 月13 日暴发的群发性泥石流为研究对象,通过分析因子权重,利用非线性回归建立了泥石流堵河程度的预测模型,并与已有模型进行了对比,论证了所建立模型的可用性,其后对不同降雨频率下的泥石流堵河程度进行预测.结果表明:流量比和流域面积是最佳的建模因子,所建立的模型绝对平均误差为0.049;与不同模型对比发现本文模型对于清平镇泥石流堵河预测具有很好的适用性;在极端降雨下,部分沟道堵河预测程度为1,会造成绵远河堵断.研究成果可为该地区泥石流灾害的防治和治理提供参考.
Abstract
River blockage resulting from debris flow disasters is common in the Wenchuan earthquake-stricken area,posing significant cascading effects that imperil the safety of life and property in the region.Therefore,accurately predicting the extent of river blockage induced by debris flows and implementing corresponding preventive measures are imperative.This study focused on the group debris flows that occurred in Qingping Town,Mianzhu City,Sichuan Province,on August 13,2010.Through the analysis of factor weights,we developed a prediction model for assessing the extent of river blockage induced by debris flows utilizing non-linear regression technique and we compared its performances with existing models.Upon demonstrating the efficacy of the new model,the extent of river blockage under different rainfall frequencies was predicated.The findings revealed that the flow ratio and catchment area were optimal predictive factors,and the absolute mean error of the established model was 0.049.Compared with other models,it was found that the established model had good applicability in predicating debris flow blockage in Qingping Town.Additionally,the predicted extent of river blockage reached 1 in some gullies under extreme rainfall conditions,it would definitely block the Mianyuan River.These findings can provide valuable insights for the prevention and mitigation of debris flow disasters in the region.
关键词
暴雨泥石流/堵河/非线性回归/预测模型/汶川震区Key words
rainfall-induced debris flow/river blockage/nonlinear regression/prediction model/Wenchuan earthquake area引用本文复制引用
基金项目
国家自然科学基金区域创新发展联合基金(U23A2044)
地质灾害防治与地质环境保护国家重点实验室自主课题(SKLGP2019Z015)
出版年
2024
国家科技期刊平台
NSTL
万方数据