四川西部是我国地质灾害高易发区。为了对滑坡孕灾环境因子的敏感性进行定量分析,以普格县境内分布的滑坡为例,基于变维分形理论、ArcGIS平台,借助累积和变换次数和分维值研究区内小、中、大型滑坡孕灾环境因子的敏感性,并引入分维值方差这一概念来度量分维值的离散程度。主要成果与结论为:①坡度、断裂和公路等因子对滑坡发育的影响较大,河流和起伏度等因子对滑坡发育的影响较小;②中型滑坡的孕灾环境因子分维值方差最大,小型滑坡次之,大型滑坡最小;③分维值为滑坡对孕灾环境因子的敏感性提供了定量的分析,而分维值方差为滑坡对各个孕灾环境因子敏感性程度的离散性进行了度量;④变维分形理论为滑坡灾害危险性评价中的孕灾环境因子权重提供了分形学依据。

The west of Sichuan Proince is highly prone to geological disasters.In an attempt to quantify the sensitivityof landslide-brewing environmental factors (including fault, petrofabric, river, slope gradient, land relief, road, and elevation),the discrete degree of fractal dimension values of the abovementioned factors are measured by calculating the variance of fractal dimension values in association with ArcGIS. Landslide in Puge County of Sichuan Province are taken as research objective. The main results are concluded as follows:(1)Slope gradient, fault, and road have large impact on the development of landslide, while river and land relief has small influence on landslide development; (2) The variance of fractal dimension value of medium-sized landslide is the largest, followed by that of small landslide and large-scale landslide; (3) The sensitivity of landslide to hazard-brewing factors can be quantified by fractal dimension values, and the dicrete degree of the sensitivity can be measured by variance of fractal dimension values; (4)The variable dimension fractal theory provide a basis for obtaining the weights of hazard-brewing environmental factors in landslide risk assessment.