摘要
滑坡预测在防灾减灾工作中具有重要意义,它包括空间、时间预测两个方面。基于统计模型进行区域评价与空间预测是滑坡灾害研究的重要方向,但是预测结果往往依赖样本数量和空间分布等。本文以马来西亚金马伦高原为研究区,选择高程、坡度、坡向、地表曲率、构造、土地覆盖、地貌类型、道路和排水系统作为评价因子,探讨运用地理信息系统(GIS)和遥感(RS)获取与管理滑坡灾害信息,以及热带雨林地区湿热环境下滑坡空间预测的方法。支持向量机(SVM)和逻辑(Logistic)回归模型分别应用于滑坡空间预测,结果显示平均预测精度分别为95.9%和86.2%,SVM法具有较高的描述精度,值得推荐;同时,基于SVM模型的滑坡空间预测受样本影响较小,预测结果相对比较稳定,这对于滑坡灾害区域评价与预测的快速实现具有实际意义。
Landslide prediction is very important in disaster prevention and reduction procedures, including spatial and temporal landslide prediction, and it is one of practical research fields to evaluate and predict landslide hazards using statistic analysis model, but the prediction result depends mostly on sample numbers and spatial distribution. The aim of this study is to analyze and compare the landslide prediction using different models in Cameron highland, Malaysia, and nine evaluation factors are selected, i.e. elevation, topographic slope, topographic aspect, topographic curvature, distance from lineament, land use and land cover, geomorphic characteristics, distance from road and drainage. Support vector machine （SVM） and logistic regression model are applied to landslide spatial prediction and mapping, and the results show that average prediction accuracy using logistic regression model is about 86. 2% , but 95.9% using SVM model, at the same time, the prediction result based on SVM model is more changeless, less influenced by sample numbers. So the SVM model is commended for actual applications, and it is more efficient and accurate for landslide hazard evaluation and spatial prediction.
出处
《遥感学报》
EI
CSCD
北大核心
2007年第6期852-859,共8页
NATIONAL REMOTE SENSING BULLETIN
基金
国家自然科学基金项目(编号:40671130)
中国-马来西亚政府间科技合作项目(Application of Malaysia Airborne Remote Sensing Project)资助
