基于分形的滑坡易发程度区划方法

Fractal-based landslide susceptibility mapping

本文发展了一种基于分形统计的滑坡易发程度评价方法,该方法仅使用已有的滑坡数据,首先通过分形统计获得滑坡分布的分形丛集关系,再通过GIS的空间操作与分析生成滑坡易发程度区划图。提出一种对滑坡易发程度区划图的可信度和预测效果进行评价的方法。本文介绍了这些方法及其在浙江地区应用的结果。

In this study,we developed a new landslide susceptibility analysis technique using historical landslide inventories and fractal statistics on a GIS platform.We also presented an examining approach to the effectiveness of a landslide susceptibility map.As an illustration of this method application,we measure the spatial fractal clustering distribution of 3285 landslides occurred from 1990 to 2003 in Zhejiang region,a susceptibility map for the region is produced by applying the fractal clustering relation using GIS-supported spatial analysis,and then the 459 landslides,which occurred from 2004 to 2007,are used to examine the effectiveness of the susceptibility map.The verification results show that 384 landslides（accounting for 83.66%of total test samples）fall in the high and very high susceptibility zones which account for 15.23%of total area and 28 landslides（accounting for 6.18%of total test samples）is within the low and very low susceptibility zones which account for 68.11%of total area.