首页|GIS支持下基于归一化信息量模型的地质灾害易发性评价

GIS支持下基于归一化信息量模型的地质灾害易发性评价

扫码查看
以沅陵县为研究区,对地质灾害的发育和分布特征进行定量化分析,基于控制因素与诱发因素,选取高程、坡向、坡度、剖面曲率、岩组、距断层的距离、距道路的距离、距河流的距离、年平均降雨量等 9 个因素作为评价指标,并利用线性函数转换的Min-Max归一化算法,将信息量模型(Information quantity model,I模型)得到的数据进行归一化处理;然后通过ROC曲线、灾积比对模型精度进行检验.结果表明:(1)信息量模型和归一化信息量模型的AUC值分别为0.730 和0.784,归一化处理后的AUC值提高了 5.4%;(2)极低易发区至极高易发区的面积分布差异显著,归一化处理后的模型在高易发和极高易发区面积上减少了 428.51 km2.归一化处理后的信息量模型具有更高的精准度,研究成果将为该地区地质灾害防治工作提供一种有效的方法.
Geological Hazard Susceptibility Assessment Based on Normalized Information Model Supported By GIS
Taking Yuanling county as the study area,the development and distribution characteristics of geological hazards are quantitatively analyzed.Based on the controlling factors and triggering factors,nine factors such as elevation,aspect,slope,profile curvature,rock group,distance from faults,distance from roads,distance from rivers,and average annual rainfall are selected as evaluation indexes.Min-Max normalization algorithm based on linear function transformation is used to normalize the data obtained from the information quantity model.Finally,the model accuracy is examined by ROC curve and disaster product ratio.Results show first the AUC values of the information quantity model and the normalized information quantity model are 0.730 and 0.784,respectively,and the AUC value is increased by 5.4%after the normalization process.Second,the area distribution from very low susceptibility area to very high susceptibility area is significantly different,the normalized information model reduces the area of high and very high susceptibility zones by 428.51 km2.Last,the results show that the normalized information quantity model has a higher accuracy,and the research results provides a more effective method for the prevention and control of geologic hazards in this area.

Yuanling county of Huaihua prefectureinformation quantity modelGISMin-Max normalized analysisgeological hazardsusceptibility assessment

张云、资锋、曹运江、成湘伟、韩用顺、唐龙

展开 >

湖南科技大学 地球科学与空间信息工程学院,湖南 湘潭 411201

湖南省自然资源事务中心,湖南 长沙 410007

怀化沅陵 信息量模型 GIS Min-Max归一化分析 地质灾害 易发性评价

2024

矿业工程研究
湖南科技大学

矿业工程研究

影响因子:0.409
ISSN:1674-5876
年,卷(期):2024.39(2)