共查询到19条相似文献，搜索用时 218 毫秒
1.
自然灾害的预测预报被认为是主动减灾防灾研究中较为经济有效的方式,其中,滑坡空间预测是滑坡灾害研究的基础工作。以汶川地震重灾区北川县为研究区,选取坡度、高程、岩石类型、地震烈度、水系、道路等6个重要滑坡影响因素作为评价因子,全面分析了地震滑坡分布与各影响因子之间的统计相关性,分别采用多元回归模型与神经网络模型计算滑坡灾害敏感性指数,并进行分级和制图。结果表明,极高和高敏感区主要分布在曲山、陈家坝等乡镇,主要沿着龙门山断裂带周边地区的河流和道路呈带状分布。其中,回归模型的预测精度为73.7%,神经网络模型的预测精度为81.28%,在本区域内,神经网络模型在滑坡灾害空间预测方面更具优势。 相似文献
2.
《云南地理环境研究》2019,(5)
采用芦山地震震后滑坡灾害地面排查及无人机遥感影像解译,利用敏感性分析方法进行"4.20"震区滑坡发育因子敏感性分析,结合地震Arias强度分布开展区域性地震次生滑坡灾害危险性评价,得到结论如下:1.基于因子敏感性分析结合地震Arias强度分布方法可解决大震后大区域尺度滑坡灾害危险性评价问题,结果与震后已发滑坡灾害点吻合,准确度和精度均得到提高;2."4.20"震后滑坡灾害发生随坡度增加呈指数上升,与岩性组合有关,最敏感的组合依次为硬岩夹软岩、硬岩夹硬岩岩层及硬岩夹极软岩层,侏罗纪及白垩纪地层易发,易于在降雨变差率中等的区域发生;3."4.20"震后滑坡高危区表现为与龙门山断裂带平行的5个高危带,在其内的乡镇、县城及与之相交的公路成为重点危害对象;4.宝兴县五龙乡、芦山县鱼泉乡和宝盛乡存在较大滑坡风险,要进一步排查,以防震后雨季发生滑坡。宝兴县城虽然极度危险区呈零星分布,但由于地处县城,灾害一旦发生危害极大,需要进行治理;5.在沿线主要道路设置雨季监测员,对来往车辆进行人工预警;6.加强极度危险区群策群防工作。 相似文献
3.
该文以提升滑坡危险性评价精度为核心目标,对深度神经网络在滑坡危险性评价中的可行性和适用性进行研究,以期充分发挥深度神经网络强大的非线性学习和拟合能力,取得更加合理的滑坡危险性评价结果。选取滑坡灾害多发的深圳市作为实例,基于深圳市815条历史滑坡数据,开展了深度神经网络建模训练;通过与广义线性模型及分类与回归树模型训练效果的对比,对深度神经网络的建模效果进行了评价,深度神经网络、广义线性模型和分类与回归树模型的AUC值依次是0.908、0.861和0.857。将训练所得的模型应用于深圳市全区,对3种模型输出的滑坡危险性评价成果的合理性和可靠性进行了对比分析,结果表明:深度神经网络建模精度良好,优于常见的广义线性模型和分类与回归树模型,输出的滑坡危险性评价成果具有合理性,适用于滑坡危险性评价工作。 相似文献
4.
基于GIS的滑坡危险性逻辑回归评价研究 总被引：7，自引：0，他引：7
该文针对地质灾害研究中的核心问题——灾害危险性评价,以万州滑坡地质灾害为例,将滑坡风险评价中的各种因子归一化处理后转换成相同分辨率的定量数据,根据特定模型进行运算,得到风险评价图,利用逻辑回归分析法,进行滑坡地质灾害危险性评价。以解决过去地质灾害危险性评价中效率低、精度差、费时费力等问题,实现滑坡地质灾害的信息化、科学化。 相似文献
5.
《山地学报》2017,(2)
新疆果子沟区域是我国与中亚地区社会、经济、文化交流的重要通道以及我国石油、天然气等能源资源安全大通道,沿途滑坡灾害频发,威胁人类安全、影响社会、经济发展,但该区域针对滑坡灾害的研究程度较低,需借助滑坡灾害易发性分析与风险等级评估结果指导防灾减灾。本文基于GF-1号卫星影像数据进行滑坡解译,选取地层岩性、断层密度、坡度、坡向、地表高程、植被指数等6个评估因子,探讨运用GIS、RS技术及统计分析模型进行滑坡灾害易发性分析与建模。基于频率比法分析各因子敏感性,利用二元Logistic回归模型进行灾害易发性分析,将研究区滑坡灾害风险分为极低、低、中、高、极高五个等级。将模型计算结果与历史滑坡信息进行比较,并借助ROC Curve检验模型准确性,AUC为0.844,表明模型预测结果具有较高准确性,因此建立的分析模型可以满足新疆果子沟区域滑坡灾害分析与评估应用,研究成果可为研究区重大线型工程保护、边坡加固提供辅助决策支持。 相似文献
6.
基于数字高程模型(DEM)计算得到的坡度、坡向等地形属性是滑坡危险性评价模型的重要输入数据, DEM误差会导致地形属性计算结果不确定性, 进而影响滑坡危险性评价模型的结果。本文选择基于专家知识的滑坡危险性评价模型和逻辑斯第回归模型, 采用蒙特卡洛模拟方法, 研究DEM误差所导致的滑坡危险性评价模型结果不确定性。研究区位于长江中上游的重庆开县, 采用5 m分辨率的DEM, 以序贯高斯模拟方法模拟了不同大小(误差标准差为1 m、7.5 m、15 m)和空间自相关性(变程为0 m、30 m、60 m、120 m)的12 类DEM误差场参与滑坡危险性评价。每次模拟包括100 个实现, 通过对每次模拟分别计算滑坡危险性评价结果的标准差图层和分类一致性百分比图层, 用以评价结果不确定性。评价结果表明, 在不同的DEM精度下, 两个滑坡危险性评价模型所得结果的总体不确定性随空间自相关程度的变化趋势并不相同。当DEM空间自相关性程度不同时, 基于专家知识的滑坡危险性评价模型的评价结果总体不确定随着DEM误差增加而呈现不同的变化趋势, 而逻辑斯第回归模型的评价结果总体不确定性随着DEM误差大小增加而单调增加。从评价结果总体不确定性角度而言, 总体上逻辑斯第回归模型比基于专家知识的滑坡危险性评价模型更加依赖于DEM数据质量。 相似文献
7.
选择"4·20"芦山地震灾区为研究区,在野外实地考察的基础上,结合高精度遥感影像解译分析,针对地震灾区滑坡、崩塌和泥石流等次生山地灾害的形成条件展开系统分析;在ArcGIS 9.3软件支持下,采用信息量模型,选择坡向、坡度、地层岩性、断裂带、河流冲刷作用、地震烈度和降水量7个影响因子作为芦山地震灾区次生山地灾害易发性评价的指标参数,将其划分为高易发区、中易发区和低易发区,该评价结果与实地考察结果基本吻合。基于GIS的信息量模型能够很好地为芦山地震灾区次生山地灾害易发性区划研究提供指导,其结果可以用来解决次生山地灾害易发性评价中效率低、精度差、费时费力等问题,从而实现次生山地灾害易发性评价的信息化和科学化。 相似文献
8.
GIS支持下三峡库区秭归县滑坡灾害空间预测 总被引：3，自引：1，他引：2
基于GIS空间分析和统计模型相结合进行区域评价与空间预测是滑坡灾害研究的重要方向之一。以三峡库区秭归县为研究区,选择坡度、坡向、边坡结构、工程岩组、排水系统、土地利用和公路开挖作为评价因子。为提高模型的预测精度、可信度和推广能力,利用窗口采样规则降低训练样本之间的空间相关性。建立Logistic回归模型,对滑坡灾害与评价因子进行定量相关性分析。计算研究区滑坡灾害易发性指数,对其进行聚类分析,绘制滑坡易发性分区图,其中高、中易发区占整个研究区面积的38.9%,主要分布在人类工程活动频繁和靠近排水系统的区域。经过验证,该模型的预测精度达到77.57%。 相似文献
9.
泥石流的发生受控于多种因子,而各因子属性段对泥石流的发生起到不同的作用.以甘肃南部武都地区为研究区,考虑到地震扰动影响并结合当地的地质环境条件,选取了高程、坡度、岩性、土地利用类型、滑坡点密度、地震因子、地质构造缓冲区及归一化植被指数(NDVI)等8个评价因子进行危险性评价.而后基于信息量模型和敏感性分析法对各影响因子属性类进行计算分析,得到各属性类的量化值.最后通过叠置分析,得到泥石流灾害危险性分区.结果表明,在中度、高度及极高度危险区内,信息量模型中分布的泥石流面积占研究区泥石流总面积的百分比为22.5％,24.7％,37.2％,而敏感性模型中占到的比例分别为25.5％,28.1％和36.4％.敏感性模型中中度危险区以上包含的泥石流比例占到90％,大于信息量模型的84.4％,因此,采用敏感性模型得到的危险性分区结果具有更高的精确性,能够为实际应用提供参考依据. 相似文献
10.
基于三种不同模型的区域滑坡灾害敏感性评价及结果检验研究 总被引：1，自引：0，他引：1
选取相对高差、坡度、坡向、水系、距断层距离、植被覆盖、地层岩性和道路等影响因子,采用信息量法、Logistic回归和人工神经网络3种模型进行滑坡灾害的敏感性评价,并对评价结果进行检验。结果表明：① 评价分类结果的准确性会关系到社会经济成本。经过采用Cohen’s Kappa系数法、Sridevi Jadi精度评估方法和ROC曲线3种方法对评价结果进行比较分析,结果显示人工神经网络模型具有更好的评价精度。② 宁强县滑坡地域分布上,呈现一带三区。其中高、中和低敏感区分别占全县总面积的39.96%,37.7%和22.33%。 相似文献
11.
12.
13.
Prediction of landslide susceptibility using rare events logistic regression: A case-study in the Flemish Ardennes (Belgium) 总被引：9，自引：8，他引：9
M. Van Den Eeckhaut T. Vanwalleghem J. Poesen G. Govers G. Verstraeten L. Vandekerckhove 《Geomorphology》2006,76(3-4):392-410
In this article a statistical multivariate method, i.e., rare events logistic regression, is evaluated for the creation of a landslide susceptibility map in a 200 km2 study area of the Flemish Ardennes (Belgium). The methodology is based on the hypothesis that future landslides will have the same causal factors as the landslides initiated in the past. The information on the past landslides comes from a landslide inventory map obtained by detailed field surveys and by the analysis of LIDAR (Light Detection and Ranging)-derived hillshade maps. Information on the causal factors (e.g., slope gradient, aspect, lithology, and soil drainage) was extracted from digital elevation models derived from LIDAR and from topographical, lithological and soil maps. In landslide-affected areas, however, we did not use the present-day hillslope gradient. In order to reflect the hillslope condition prior to landsliding, the pre-landslide hillslope was reconstructed and its gradient was used in the analysis. Because of their limited spatial occurrence, the landslides in the study area can be regarded as “rare events”. Rare events logistic regression differs from ordinary logistic regression because it takes into account the low proportion of 1s (landslides) to 0s (no landslides) in the study area by incorporating three correction measures: the endogenous stratified sampling of the dataset, the prior correction of the intercept and the correction of the probabilities to include the estimation uncertainty. For the study area, significant model results were obtained, with pre-landslide hillslope gradient and three different clayey lithologies being important predictor variables. Receiver Operating Characteristic (ROC) curves and the Kappa index were used to validate the model. Both show a good agreement between the observed and predicted values of the validation dataset. Based on a qualified judgement, the created landslide susceptibility map was classified into four classes, i.e., very high, high, moderate and low susceptibility. If interpreted correctly, this classified susceptibility map is an important tool for the delineation of zones where prevention measures are needed and human interference should be limited in order to avoid property damage due to landslides. 相似文献
14.
Spatial patterns of old, deep-seated landslides: A case-study in the northern Ethiopian highlands 总被引：1，自引：0，他引：1
M. Van Den Eeckhaut J. Moeyersons J. Nyssen Amanuel Abraha J. Poesen Mitiku Haile J. Deckers 《Geomorphology》2009,105(3-4):239-252
During the last decade, slope failures were reported in a 500 km2 study area in the Geba–Werei catchment, northern Ethiopia, a region where landslides were not considered an important hazard before. Field observations, however, revealed that many of the failures were actually reactivations of old deep-seated landslides after land use changes. Therefore, this study was conducted (1) to explore the importance of environmental factors controlling landslide occurrence and (2) to estimate future landslide susceptibility. A landslide inventory map of the study area derived from aerial photograph interpretation and field checks shows the location of 57 landslides and six zones with multiple landslides, mainly complex slides and debris flows. In total 14.8% of the area is affected by an old landslide. For the landslide susceptibility modelling, weights of evidence (WofE), was applied and five different models were produced. After comparison of the models and spatial validation using Receiver Operating Characteristic curves and Kappa values, a model combining data on elevation, hillslope gradient, aspect, geology and distance to faults was selected. This model confirmed our hypothesis that deep-seated landslides are located on hillslopes with a moderate slope gradient (i.e. 5°–13°). The depletion areas are expected on and along the border of plateaus where weathered basalts rich in smectite clays are found, and the landslide debris is expected to accumulate on the Amba Aradam sandstone and upper Antalo limestone. As future landslides are believed to occur on inherently unstable hillslopes similar to those where deep-seated landslides occurred, the classified landslide susceptibility map allows delineating zones where human interventions decreasing slope stability might cause slope failures. The results obtained demonstrate that the applied methodology could be used in similar areas where information on the location of landslides is essential for present-day hazard analysis. 相似文献
15.
GIS and ANN model for landslide susceptibility mapping 总被引：1，自引：0，他引：1
XU Zeng-wang 《地理学报》2001,11(3):374-381
Landslide hazard is as the probability of occurrence of a potentially damaging landslide phenomenon within specified period of time and within a given area. The susceptibility map provides the relative spatial probability of landslides occurrence. A study is presented of the application of GIS and artificial neural network model to landslide susceptibility mapping, with particular reference to landslides on natural terrain in this paper. The method has been applied to Lantau Island, the largest outlying island within the territory of Hong Kong. A three-level neural network model was constructed and trained by the back-propagate algorithm in the geographical database of the study area. The data in the database includes digital elevation modal and its derivatives, landslides distribution and their attributes, superficial geological maps, vegetation cover, the raingauges distribution and their 14 years 5-minute observation. Based on field inspection and analysis of correlation between terrain variables and landslides frequency, lithology, vegetation cover, slope gradient, slope aspect, slope curvature, elevation, the characteristic value, the rainstorms corresponding to the landslide, and distance to drainage line are considered to be related to landslide susceptibility in this study. The artificial neural network is then coupled with the ArcView3.2 GIS software to produce the landslide susceptibility map, which classifies the susceptibility into three levels: low, moderate, and high. The results from this study indicate that GIS coupled with artificial neural network model is a flexible and powerful approach to identify the spatial probability of hazards. 相似文献
16.
GIS and ANN model for landslide susceptibility mapping 总被引：4，自引：0，他引：4
XU Zeng-wang 《地理学报(英文版)》2001,11(3):374-381
1 IntroductionThe population growth and the expansion of settlements and life-lines over hazardous areas exert increasingly great impact of natural disasters both in the developed and developing countries. In many countries, the economic losses and casualties due to landslides are greater than commonly recognized and generate a yearly loss of property larger than that from any other natural disasters, including earthquakes, floods and windstorms. Landslides in mountainous terrain often occur a… 相似文献
17.
以阿坝藏族羌族自治州地质灾害频发的理县为研究区,从地形地貌、地质环境、水文条件和人类工程活动等方面选取11个影响因子,通过皮尔森相关系数研究各因子之间的相关性,从而构建滑坡易发性评价指标体系。利用信息量模型计算各影响因子的信息量值,从信息量模型得出的极低和低易发性分区中选取非滑坡样本,在此基础上将样本数据代入随机森林和径向基函数神经网络2种机器学习模型开展滑坡易发性评价,并通过接收灵敏度（Receiver Operating Characteristic,ROC）曲线进行精度验证。结果显示：随机森林模型预测出的高易发区单位面积内分布的滑坡点数量更为集中,在仅占6.666%的区域分布了74.026%的灾害点,评价结果优于径向基函数神经网络模型。ROC曲线中两模型AUC（Area Under Curve）值分别为0.893、0.874,说明随机森林模型具有更高的可靠性,比径向基函数神经网络在该区域地质灾害易发性评价中更具优势。 相似文献
18.
Modeling typhoon- and earthquake-induced landslides in a mountainous watershed using logistic regression 总被引：4，自引：2，他引：4
Landslides can be caused by storms and earthquakes. Most logistic regression models proposed in recent years have been targeted at rainfall-induced landslides. In areas such as Taiwan, where landslides can be triggered by typhoons (tropical cyclones) and earthquakes, a rainfall-induced model is insufficient because it provides only a partial explanation of landslide occurrence and overlooks the potential effect of earthquakes on typhoon-triggered landslides. This study used landslides triggered by a major earthquake and a typhoon prior to the earthquake to develop an earthquake-induced model and a typhoon-induced model. The models were then validated by using landslides triggered by three typhoons after the earthquake. According to the results, typhoon-triggered landslides tended to be near stream channels and earthquake-triggered landslides were more likely to be near ridge lines. Moreover, a major earthquake could still affect the locations of typhoon-triggered landslides 6 years after the earthquake. This study therefore demonstrates that an earthquake-induced model both sheds light on the environmental factors for triggering landslides, and augments a rainfall-induced model in its predictive capability in areas such as Taiwan. 相似文献
19.
强震区泥石流启动机制 总被引：2，自引：0，他引：2
地震区的泥石流物源主要来源于滑坡、崩塌等松散体,具有结构疏松,密实度低,堆积时间短等特点,与非地震环境中的滑坡、崩塌堆积体的结构有所不同,堆积体的物理力学性质发生了改变,堆积体转换为泥石流所需的外界条件也相应的改变。以汶川地震区都江堰市龙池镇典型泥石流灾害为例,分析了地震滑坡、崩塌松散体的堆积形态和堆积体的应力环境。从静力学和动力学角度分析堆积体在强降雨条件下的起动特征,探讨了降雨作用形成的地表径流水深与堆积体失稳时的应力极限状态的关系。分析得出沟道岸坡滑坡堆积体发生侵蚀时的地表径流力为F=(τ1f-f1sinα)/cos(α-26.65),并建立径流水深与地表径流力的关系:H=F/4ρsgJ。分析在动量守恒条件下,堆积体单位时间内的侵蚀体积dV=dM/γs模型。为了进一步探讨在实际现场的应用,以汶川地震区都江堰市的水打沟泥石流为例,分析发生泥石流时的地表径流水深为0.011 m,其结论与实际调查结果基本一致。 相似文献