改进逻辑回归方法在滑坡敏感性评价中的应用研究  

IMPROVING AUTOLOGISTIC REGRESSION METHOD IN THE RESEARCH ON THE APPLICATION OF LANDSLIDE SENSITIVITY EVALUATION

作　　者：夏添[1] 常鸣[1] 

机构地区：[1]成都理工大学地质灾害防治与地质环境保护国家重点实验室,四川成都610059

出　　处：《物探化探计算技术》2013年第2期185-188,119-120,共4页Computing Techniques For Geophysical and Geochemical Exploration

基　　金：国家科技支撑计划(2011BAK12B01)

摘　　要：以四川省北川县县城及周边为研究区,在研究区内进行了滑坡的遥感解译和野外调查完成北川滑坡分布图。然后选择了坡度,坡向,高程,曲率,水系影响,断层影响共六个因子,采用了一种改进的逻辑回归模型(AutoLogistic)模型,在ArcGIS平台上进行了滑坡敏感性的评价。完成了滑坡敏感性分级图的制图并对评价的结果和现实滑坡的发生进行了统计分析,结果表明敏感性分级比较合理。This article chooses Beichuan county and the surrounding area for research area with finishing landslides distribution map by remote sensing interpretation and field investigation.Firstly,six evaluation factors were selected,including slop,aspect,elevation,curvature,river system influence are,fault influence,for landslide sensitivity evaluation by an improving autologistic regression model in ArcGIS platform.This paper also completes the landslide susceptibility and Statistical analysis of possible appearance of the current landslides.The results shows the sensitivity of the susceptibility is reasonable.

关 键 词：Autologistic 回归模型 遥感解译 滑坡敏感性 

分 类 号：X141[环境科学与工程—环境科学]