基于GIS的长江三峡库区滑坡影响因子分析

Analysis of Landslide Causative Factors Using GIS in the Three Gorges Reservoir Area, China

利用GIS技术和统计方法,对三峡库区选定的研究区域(面积4539km2)滑坡空间分布和地形、地质等滑坡内部因子之间相关性进行统计计算。在建立地质、地形数据库等滑坡因子空间数据库和滑坡空间分布数据库(数据比例尺均为1∶10000)基础上,从地形数据库提取25m分辨率DEM,再派生出高程、高差、坡度、坡向、平面曲率、剖面曲率等地形影响因子;从地质数据库提取地层和岩性组合影响因子。将各个定性的因子按一定规则进行重分类、转换为25m分辨率的栅格数据格式,在GIS中进行地图代数运算、统计计算滑坡和各影响因子相关性。结果表明,滑坡分布和Q4、J1x,J1z、S岩性岩组;90m以下、90～135m和135～175m三个高程带;15～20m局部高差;10°～25°坡度;北、南和西北方向及-1～1曲率范围等影响因子相关性等级都大于1,为滑坡发生的主要影响因子类属。研究的结果是进行滑坡易发性评价的基础,可以指导库区滑坡灾害管理、土地利用等。

Geographical information system(GIS)are drawing more and more attention in the field of landslide studies because of the powerful spatial analysis and spatial database management. The author review statistical relationship between landslide occurrence and static and dynamic variables using powerful function of GIS. Slope instability problems exist over a wide range of scale terrain, climatic conditions and other triggering factors. The factors include static factors and dynamic factors. In the paper, dynamic factors such as rainfall, earthquake, and human activity haven't be used. To apply the statistical method, a spatial database that considered static factors were designed and constructed. These data are available either as a paper map or as a digital map. A digitized map of landslide location and boundaries were produced from the catalogue about regional landslides and (1∶10 000) distribution maps; topographical and geological data (1∶10 000-scale paper maps) are collected,digitized and manipulates into a spatial database using ArcGis GIS; Elevation data were obtained by converting the (1∶10 000) scale digitized topographical map(contour interval is 25 m) and a DEM (25 m resolution) was made. Using the DEM, the elevations, local relief, slope, aspect, profile curvature, plan curvature were calculated. Stratum, lithology formation were extracted from the geological database. In order to reveal the correlation between landslide area and these instability factors, all digital data were input to the GIS, all input variables were grouped into a few meaningful classes to integrate quantitative variable with qualitative variable, a vector-to raster conversion was undertaken to provide a raster data of factors and landslide areas with 25 m×25 m pixels. We used 7 landslide triggering factors to analyze the relation landslide spatial distribution and their contributing factors. All triggering factors have uncertainties and different factors have different relationship to the distribution of landslides. The influence of pertinent factors on the occurrence of landslides is Q 4、J 1x, J 1z,S lithology formation; 90 m, 90 m～135 m and 135 m～175 m elevations; 15 m～20 m local relief; 10°～25°slope; N,S and WN aspect and -1～1 curvature. The GIS was used to analyze the vast amount of data efficiently, and statistical programs were used to maintain specificity and accuracy. The results can be used to reduce associated hazards, and to plan land use and construction.