节点文献

基于灰色关联与粗糙依赖度的甘肃兰州市区泥石流危险性评价

Hazard assessment of debris flow in Lanzhou City of Gansu Province based on methods of grey relation and rough dependence

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 牛全福陆铭李月锋冯尊斌高文信陆小琳

【Author】 NIU Quanfu;LU Ming;LI Yuefeng;FENG Zunbin;GAO Wenxin;LU Xiaolin;College of Civil Engineering,Lanzhou University of Technology;Emergency Mapping Engineering Research Center of Gansu;

【机构】 兰州理工大学土木工程学院甘肃省应急测绘工程研究中心

【摘要】 危险性评价是泥石流灾害研究的主要问题之一,多因子综合评价是泥石流危险性评估的主要模式。由于泥石流灾害的不确定性和灾害系统的复杂性,因而非确定性分析方法成为其评估首选的研究方法。本研究以泥石流频发的兰州市为例,针对常规研究中选取的坡度等10个地形影响因子,首先应用复相关系数法去除地形因子间相关性,获得综合主要地形影响因子的地形易发性因子;然后,从地形易发性、地质岩性、地貌类型、降雨、植被覆盖度5个影响因子出发,应用灰色关联度、粗糙依赖度以及融合灰色关联度与粗糙依赖度的模糊综合方法获取泥石流危险性评价图。评价结果可知:(1)地形起伏度、坡向变率、地表粗糙度、坡长、坡度变率、植被覆盖度和降雨对该地区泥石流的发生具有一定的控制作用,与现场调查结果相吻合;(2)将三种评估结果与泥石流灾害进行叠加分析发现,研究区大多数泥石流点分布在0. 4~0. 6和0. 6~0. 8区间,且三种方法获得结果的趋势一致,均较客观地反映研究区泥石流危险性的空间分布;(3)由ROC曲线分析,灰色关联度和粗糙依赖度的评估结果精度比较接近,其ROC曲线下面积均接近0. 7;融合灰色关联和粗糙依赖法的模糊综合方法所得的ROC曲线下面积大于0. 7,其说明该模型评估的结果较好地反映泥石流危险性分布,具有较高的可信度。

【Abstract】 Hazard assessment is one of the key issues in disaster studies of debris flow. And comprehensive assessment with multiple factors is the main mode of them. Because of uncertainty and complexity,nondeterministic analysis has already become the preferred approach. This study take Lanzhou,a provincial capital with frequent debris flows,as an example. Firstly,the relationship among 10 topographic factors,such as slope,et al.,was taken away with the method of complex correlation coefficient and the topographic susceptibility factor was gained from them. Then, using five factors, for example, the topographic susceptibility,the geological lithology,landform,the precipitation and vegetation coverage and the hazard assessment maps were obtained with three methods of grey relational degree,rough dependence degree and fuzzy comprehensive method with grey relational degree and rough dependence degree. The conclusions are as follows:( 1) These debris flows in the study area are controlled by seven factors,including the relief,slope aspect,surface coarseness,slope length,slope of slope,vegetation cover and precipitation,which are coincided with field investigation.( 2) After overlaying and analyzing these assessment maps with the data of debris flows,it shows that most of the points of debris flows were taken place in the grades of 0. 4 ~ 0. 6 and0. 6 ~ 0. 8,these results from three methods are consistent and better show the special distribution of the debris flows in the study area.( 3) From ROC analysis,the accuracy from grey relational degree and rough dependence degree are rather similar and the area under ROC curve is close to 0. 7. However,the method of fuzzy comprehensive weight determination with grey relational degree and rough dependence degree shows high credibility with the area of bigger than 0. 7,which better reflect the debris flow distribution in the study area.

【关键词】 泥石流危险性不确定性分析因子精度
【Key words】 debris flowhazardnon-deterministic analysisfactoraccuracy
【基金】 国家自然科学基金:基于GIS的典型黄土区地质灾害致灾因子的优选与地表灾害过程模拟(41461084)
  • 【文献出处】 中国地质灾害与防治学报 ,The Chinese Journal of Geological Hazard and Control , 编辑部邮箱 ,2019年05期
  • 【分类号】P642.23
  • 【被引频次】8
  • 【下载频次】318
节点文献中: 

本文链接的文献网络图示:

本文的引文网络