论文部分内容阅读
本文通过统计学方法提取可能影响该地区滑坡的因子,并依据二元逻辑回归结果,利用GIS的空间分析和建模功能,对研究区滑坡进行建模,再由相关性等级分析方法进一步获取相关影响因子对滑坡的影响范围,最终得出该区域滑坡危险性评价图。而在不同分辨率尺度上,对滑坡产生影响的因子有所相同,当分辨率由1 000m~60m时,影响因子明显增多。以逻辑回归模型自身精度、滑坡实际发生比和坡度法为模型判定标准,发现120m分辨率下的滑坡发生概率模型能有效表达研究区滑坡发生状况。
In this paper, the factors that may affect the landslide in this area are extracted through statistical methods. Based on the binary logistic regression results, the spatial analysis and modeling functions of GIS are used to model the landslides in the study area, and then the correlations are further obtained Influence factors on landslide impact range, and finally come to the landslide risk assessment map. At different resolution scales, the factors that affect the landslide are the same. When the resolution is from 1 000 m to 60 m, the influencing factors are obviously increased. The self-precision of the logistic regression model, the ratio of actual occurrence of landslide and the gradient method were used as the criteria for the model. It was found that the landslide probability model with 120m resolution can effectively express the landslide occurrence in the study area.