三峡库区万州区滑坡灾害易发性评价研究

发布时间：2018-09-07 11:53

【摘要】：滑坡灾害易发性研究在滑坡灾害风险管理与城市规划等方面具有非常重要的现实意义。以往的研究中,鲜有对指标因子状态划分作有关深入分析和讨论的。鉴于此,以滑坡灾害频发的三峡库区万州区为研究对象:首先,选取影响滑坡发生的7个致灾因子(地层岩性、地质构造、水系分布、坡度、坡向、坡体结构及土地利用)作为滑坡易发性的评价指标,依据各指标条件下滑坡累计发生频率曲线斜率的变化,并结合滑坡面积比和分级面积比曲线对指标因子的状态进行分级;其次,根据全区655个历史滑坡数据,分别运用信息量模型和逻辑回归模型建立各自的滑坡易发性评价体系;再则,采用快速聚类法(K-means cluster)对以上2种方法所得到的易发性结果进行分级,并基于GIS平台,得到全区滑坡易发性区划图;最后,从模型结果、精度、适用条件等方面对2个模型进行讨论和比较,研究结果表明:信息量模型和逻辑回归模型的预测精度分别为73.0%和54.9%,前者预测能力要优于后者。
[Abstract]:The study of landslide hazard vulnerability is of great practical significance in landslide risk management and urban planning. In previous studies, there are few in-depth analyses and discussions on the classification of index factors. In view of this, taking Wanzhou District of the three Gorges Reservoir area, where landslide disasters occur frequently, as the research object: firstly, 7 disaster factors (stratigraphic lithology, geological structure, water distribution, slope, slope direction) affecting landslide occurrence are selected. Slope structure and land use are regarded as the evaluation indexes of landslide susceptibility. According to the slope change of landslide cumulative frequency curve under the condition of each index, the state of the index factor is classified according to the landslide area ratio and graded area ratio curve. Secondly, according to the 655 historical landslide data in the whole area, the information quantity model and the logical regression model are used to establish their own landslide vulnerability evaluation system. Rapid clustering method (K-means cluster) was used to classify the results of the two methods, and based on the GIS platform, the susceptibility map of landslide was obtained. The results show that the prediction accuracy of the information quantity model and the logical regression model are 73.0% and 54.9% respectively, and the former is superior to the latter.
【作者单位】：中国地质大学工程学院;中国地质大学地质调查研究院;
【基金】：国家自然科学基金资助项目(41572292、41572289、41301589)~~
【分类号】：P642.22

【参考文献】