地理信息系统/遥感技术支持下三峡库区青干河流域滑坡危险性评价

LANDSLIDE HAZARD ASSESSMENT IN QINGGAN RIVER OF THE THREE GORGES RESERVOIR BASED ON GEOGRAPHICAL INFORMATION SYSTEMS AND REMOTE SENSING

以统计模型为基础、地理信息系统作为工具的滑坡灾害评价模式已经得到普遍认可和使用,数字高程模型(DEM)、遥感影像、区域地质调查资料已经成为区域滑坡评价研究的因子数据源。选择三峡库区青干河流域顺向坡滑坡多发地段为研究区,在滑坡编目数据库基础上,通过:(1)数字高程模型获取高程、坡度、地形聚水能力因子;(2)遥感影像获取植被指数;(3)区域地质调查资料、数字高程模型计算斜坡类型定量因子TOBIA指数及获取岩石地层单元因子。采用二分类变量逻辑回归评价方法对上述6种因子建立滑坡危险性评价模型,开展地理信息系统/遥感技术支持下顺向坡滑坡危险性评价研究。研究结果表明,根据模型概率值分布和已知滑坡发育关系,可以将研究区划分为高危险区、中等危险区、低危险区3个等级,高危险区包含70%已知滑坡,中等危险区包含14%已知滑坡,评价结果和实际滑坡发育情况吻合,合理地反映区内滑坡灾害发育的总体特征。

Landslide hazard assessment is necessary for disaster management and planning in dam projects.This paper presents a geographical information systems/remote sensing-aided procedure for landslide hazard mapping at a regional scale around the Qinggan River of the Three Gorges,China,where most of landslides are active in dip slopes.A landslide inventory is carried out based on field investigations and aerial photo interpretation,while another data set of environmental factors is constructed,such as geological and topographic thematic maps,lithology and vegetation maps.The factors contributing to landslide occurrence such as altitude,slope and flow accumulation,are derived from Digital Elevation Models,and rock strata is also extracted from geological map,vegetation index from satellite images of Landsat TM bands 2/3/4.Quantitative geometric alignment relationships between strikes of slopes and strata are established by TOBIA index,and a method of generalized likelihood ratio is then utilized to analyze the relationships between landslide occurrence and environmental factors,such as lithology,slope angle,aspect,flow accumulation and vegetation,etc.Based on the database,the casual factors,which make possible contribution to landslide occurrence,are combined into a binary logistic regression model,and then the landslide probabilities are calculated by cell to cell.The results from the logistical regression model coincides well with the previous landslide occurrence.