逻辑回归和信息量模型在山区滑坡易发性评价中适宜性的分析

以万山区为例,在区域滑坡孕灾条件的基础上,筛选工程地质岩组、斜坡结构、平均坡度、地貌、距构造距离及距河流距离共6个易发条件因子,选取逻辑回归模型和信息量模型对山区滑坡进行易发性评价。结果显示逻辑回归模型中中高易发区面积占比分别为1578%和1970%,82%的地质灾害点落在该区域内;信息量模型中中高易发区面积占比为1241%、2519%,包含了区域88%的滑坡灾害点。最后通过实际发生的灾害点在各易发区的分布情况进行检验,逻辑回归模型中灾害点落在高易发区的比例远小于信息量模型,且高易发等级中灾害点实际发生的比值较小,说明针对山区区域滑坡地质灾害易发性评价结果预测上,信息量模型的评价结果更为客观准确。     

基于信息量、逻辑回归及其耦合模型的滑坡易发性评估研究:以青海沙塘川流域为例

滑坡易发性定量评估是预测滑坡发生空间概率的重要手段,基于统计分析原理的评估方法目前在国内外应用最为广泛,且不同评估方法的对比研究逐渐成为热点。以青海沙塘川流域黄土梁峁区为例,剖析了信息量模型和逻辑回归模型在滑坡易发性评估中的优越性和局限性,并探索提出基于二者的耦合模型。考虑坡度、坡向、起伏度、岩性、与干流距离、与支流距离和植被指数等7个影响因素,对比分析了基于信息量、逻辑回归及二者耦合模型的滑坡易发性评估的技术流程及结果。3种模型的成功率分别为:耦合模型成功率(78. 9%)>信息量模型成功率(71. 8%)>逻辑回归模型成功率(70. 8%)。在沙塘川流域黄土滑坡的易发性评估中,信息量和逻辑回归模型的表现基本相当,但信息量-逻辑回归耦合模型的成功率明显提升。该研究结果可为黄土高原区滑坡易发性定量评估提供借鉴。     

基于斜坡单元与信息量法结合的宝塔区黄土滑坡易发性评价

滑坡易发性评价对滑坡灾害的防治与管理具有重要意义。为了评价延安宝塔区黄土滑坡易发性,以斜坡为基本评价单元,选取斜坡坡度、坡高、坡向、坡形、斜坡结构类型、植被和人类工程活动7个指标作为评价因子,在Arc GIS平台下,利用信息量模型对研究区的黄土滑坡进行易发性分区评价。评价结果表明,宝塔区滑坡高易发区面积1092.39km~2,占全区面积的30.81%,主要分布于宝塔区的中部及北部地区,低易发区集中于宝塔区南部汾川河流域。以斜坡作为评价单元提高了与实际地形地貌的吻合度。应用信息量模型进行滑坡易发性评价具有较高的预测精度,已有滑坡点落在很高易发区和高易发区中的比例为95.7%,较真实地反映了客观实际。     

基于信息量与Logistic模型的矿区滑坡灾害易发性评价

针对矿区长期煤矿开采引起的滑坡灾害频发问题,快速高效地模拟和评价矿致滑坡灾害易发性是实现采矿地区科学防灾减灾的关键。基于此,本文应用信息量与Logistic回归模型结合多源高分辨率光学遥感数据等,选取相对高差、坡度、坡向、距断层距离、NDVI、距采空区距离6个滑坡影响因子来评价采煤矿区滑坡灾害易发性。结果表明:(1)信息量与Logistic回归模型耦合的综合预测准确率为96%,信息量模型滑坡预测准确率为95%,实验结果表明耦合模型的预测精度优于单一信息量评价模型,评价模型的合理性和预测精度皆符合检验要求;(2)研究结果也表明了采用信息量+Logistic回归模型耦合能较为客观准确、快速高效地评价地下采矿引起的滑坡灾害易发范围,评价结果可为类似地区高效快速划定滑坡灾害易发区间提供技术支撑。     

不同组合模型区域滑坡易发性及精度分析

为探索区域滑坡易发性评价模型的适用性和评价结果的合理性,以滑坡灾害高发的白龙江流域为研究区,首先选取坡度、地形起伏度、距断层距离、地层岩性、流域沟壑密度、植被指数等6项影响滑坡发生的孕灾因子作为易发性的评价指标,以研究区2 093处滑坡灾害点为样本数据,依据各指标条件下的信息量值、确定性系数值和证据权重值曲线突变规律,并结合滑坡面积及分级面积频率比曲线作为等级划分的临界值来确定因子分级状态;其次,基于指标因子状态分级和相关性分析结果,采用信息量法、确定性系数法、证据权法分别与逻辑回归组合的3种模型开展区域滑坡灾害易发性评价,并从模型结果、适用性和精度等方面采用多手段对3种组合模型进行比较和讨论。研究结果表明:在区域滑坡易发性评价方面,3组模型均表现较为理想,信息量和逻辑回归组合模型的预测精度为94.6%,其预测精度和准确性优于其他2种组合模型。笔者以白龙江流域中游及其岷江支流段为例,开展滑坡灾害易发性评价模型适用性、评价结果分析以及预测精度评价对比和研究等,成果可为该区地质灾害防灾减灾和国土空间用途管制规划决策提供参考。     

贵州省都匀市滑坡易发性评价研究

都匀市是贵州省城镇滑坡地质灾害多发频发区。文章以都匀市沙包堡镇为研究区,采用栅格单元提取高程、坡度、岩性、水系等9项致灾因子,分别使用都基于数学统计模型的定量分析方法（二元逻辑回归模型、信息量模型）和定性分析方法（层次分析模型）对都匀市研究区滑坡地质灾害易发性进行评价。结果表明:二元逻辑回归模型预测精度与预测效果均为最优,其ROC曲线下面积AUC值为0.873,易发性分区中高易发区和中易发区内预测发生滑坡面积比占95.41%,且最符合野外实地调查验证情况。评价方法与结果可为贵州城镇地区滑坡地质灾害评价和防治提供借鉴。     

基于贝叶斯优化机器学习超参数的滑坡易发性评价

利用机器学习模型进行滑坡易发性评价时,不同的超参数设置往往会导致评价结果的不同.采用贝叶斯算法对4种常见机器学习模型(逻辑回归LR、支持向量机SVM、人工神经网络ANN和随机森林RF)的超参数进行了优化,探索了该算法对滑坡易发性机器学习模型的优化效果.以湘中地区4县(安化县、新华县、桃江县和桃源县)滑坡易发性评价为例说...     

基于多模型的滑坡易发性评价以甘肃岷县地震滑坡为例

2013年7月22日,甘肃省岷县漳县交界处发生了MS6.6级地震(岷县地震),本文以这次地震烈度Ⅷ度区为研究区,根据地震前后遥感影像解译出来的2330个地震滑坡数据,以坡度、坡向、水系、岩性和断层为因子图层,分别应用模糊逻辑法,信息量模型及Shannon熵改进的信息量模型,对研究区的地震滑坡易发性进行评价。结果表明: 1滑坡的高易发性地区位于研究区的中间部分,以及水系0～50m这一缓冲区范围内,离水系越近滑坡易发性等级越高; 2应用ROC曲线对3个模型的易发性评价结果进行比较,信息量模型和Shannon熵改进的信息量模型的AUC值分别为0.8488, 0.8502; 模糊逻辑模型的AUC值为0.7640,表明前两个模型的表现较好,而模糊逻辑模型相对来说表现一般; 3通过对比3个模型中各等级易发性所占的面积比例和各等级易发性中滑坡数目占总数比例,表明Shannon熵改进后的模型更适用于灾害风险评价以及应急风险管理等实际应用。     

基于逻辑回归–信息量的川藏交通廊道滑坡易发性评价

川藏交通廊道位于青藏高原中东部,是世界上隆升和地貌演化最快的区域之一。在内外动力耦合作用下,区内滑坡灾害极其发育,严重制约着公路、铁路和水电工程的规划建设。在区域地质资料收集和整理的基础上,选取岩性、坡度、坡向、坡形、地形起伏度、地形粗糙度、断裂密度和河流距离8个因素为评价因子,结合传统信息量和逻辑回归模型的优势,采用逻辑回归–信息量模型对研究区滑坡进行易发性评价。通过对评价因子的多重共线性和显著性检验,得到评价因子不存在多重共线性且均对滑坡发生具有显著影响。采用ROC曲线对评价结果进行检验,其AUC值为0.81,表明评价模型能很好地预测滑坡的发生。易发性评价结果表明:研究区高易发区主要集中龙门山断裂带、金沙江断裂带、澜沧江断裂带、怒江断裂带、边坝–洛隆断裂带等大型活动断裂带控制区,以及区内坡度陡峭、地形起伏度大的大型河流深切河谷的两岸;中易发区在区内分布广泛,主要分布在岸坡较陡、地形起伏度中等的大型河流支流的两岸。研究结果有利于加深对川藏交通廊道滑坡发育分布的认识,也可为研究区的工程规划建设和防灾减灾提供科学依据。     

基于XGBoost模型的三峡库区燕山乡滑坡易发性评价与区划

滑坡易发性评价是精细化滑坡灾害风险评价的基础。为了提升滑坡易发性评价模型的精度和稳健性,以三峡库区万州区燕山乡为例,选取工程地质岩组、堆积层厚度等九个影响因子构建滑坡易发性评价指标体系,应用信息量模型定量分析滑坡发育与指标之间的关系。在此基础上,随机选取70%/30%的滑坡样本作为训练/验证数据集,应用极致梯度提升模型(extreme gradient boosting, XGBoost)开展易发性评价。随后从模型预测精度和模型稳定性两方面将其与决策树模型(decision tree, DT)和梯度提升树模型(gradient boosting decision tree, GBDT)进行对比。结果表明：研究区堆积层滑坡主要受长江水系、堆积层厚度和工程地质岩组影响。XGBoost模型具有最高的准确率(94.3%)和预测精度(97.3%)。在模型稳定性验证中,平均预测精度最高(97.3%),优于DT(91.3%)和GBDT(95.7%),模型标准差和变异系数均为0.01,低于其余两种模型。XGBoost在区域滑坡易发性评价与制图中得到了可靠的结果,为滑坡灾害空间预测提供了新的技术支撑。     

浙西梅雨滑坡易发性评价模型对比

我国目前滑坡易发性评价研究主要集中在西南地区,对东南部降雨引发特别是梅雨引发的滑坡研究较少.选取浙江省西北部梅雨控制区淳安县为研究区,通过遥感解译结合野外详细调查,共确定滑坡596处,并建立滑坡编录数据库.选取高程、坡向、坡度、曲率、工程岩组、断层、道路、建设用地、植被等9个滑坡影响因子,基于GIS栅格分析方法,采用人工神经网络(ANN)、logistic回归和信息量3种评价模型,分别对32种不同影响因子组合进行滑坡易发性对比评价,得到滑坡易发性指数图.应用评价曲线下面积AUC(area under curve)对评价结果进行检验,ANN、logistic回归和信息量3种模型的正确率分别是93.75%、89.76%和90.06%;采用淳安县2014年梅汛期发生的13处滑坡作为预测样本,3种模型预测率分别是94.75%、94.33%和77.21%.上述分析结果表明:ANN模型优于其他两者.以ANN模型评价结果指数图为基础进行易发性分区,采用滑坡强度指标进行分区结果检验,滑坡强度值由易发性低、较低、中和高依次递增,说明分区结果合理.研究成果可以为浙西降雨型滑坡特别是由梅雨引发滑坡的易发性评价提供参考.      

Quantitative landslide susceptibility mapping at Pemalang area,Indonesia

For quantitative landslide susceptibility mapping, this study applied and verified a frequency ratio, logistic regression, and artificial neural network models to Pemalang area, Indonesia, using a Geographic Information System (GIS). Landslide locations were identified in the study area from interpretation of aerial photographs, satellite imagery, and field surveys; a spatial database was constructed from topographic and geological maps. The factors that influence landslide occurrence, such as slope gradient, slope aspect, curvature of topography, and distance from stream, were calculated from the topographic database. Lithology was extracted and calculated from geologic database. Using these factors, landslide susceptibility indexes were calculated by frequency ratio, logistic regression, and artificial neural network models. Then the landslide susceptibility maps were verified and compared with known landslide locations. The logistic regression model (accuracy 87.36%) had higher prediction accuracy than the frequency ratio (85.60%) and artificial neural network (81.70%) models. The models can be used to reduce hazards associated with landslides and to land-use planning.     

基于集成学习与径向基神经网络耦合模型的三峡库区滑坡易发性评价

准确的滑坡易发性评价结果是滑坡风险评价的重要基础.为提升滑坡易发性评价精度,以三峡库区龙驹坝为例,选取坡度等10个因子构建滑坡易发性评价指标体系,应用频率比方法定量分析各指标与滑坡发育的关系.在此基础上,随机选取70%/30%的滑坡数据作为训练/测试样本,应用径向基神经网络和Adaboost集成学习耦合模型(RBNN-Adaboost),径向基神经网络和逻辑回归模型分别开展易发性评价.结果显示:水系距离、坡度等是滑坡发育的主控因素;RBNN-Adaboost耦合模型的预测精度最高(0.820),优于RBNN模型和LR模型的0.781和0.748.Adaboost集成算法能进一步提升模型的预测性能,所提出的耦合模型结合了两者的优点,具有更强的预测能力,是一种可靠的滑坡易发性评价模型.      

Landslide susceptibility mapping using GIS-based multi-criteria decision analysis,support vector machines,and logistic regression

Identification of landslides and production of landslide susceptibility maps are crucial steps that can help planners, local administrations, and decision makers in disaster planning. Accuracy of the landslide susceptibility maps is important for reducing the losses of life and property. Models used for landslide susceptibility mapping require a combination of various factors describing features of the terrain and meteorological conditions. Many algorithms have been developed and applied in the literature to increase the accuracy of landslide susceptibility maps. In recent years, geographic information system-based multi-criteria decision analyses (MCDA) and support vector regression (SVR) have been successfully applied in the production of landslide susceptibility maps. In this study, the MCDA and SVR methods were employed to assess the shallow landslide susceptibility of Trabzon province (NE Turkey) using lithology, slope, land cover, aspect, topographic wetness index, drainage density, slope length, elevation, and distance to road as input data. Performances of the methods were compared with that of widely used logistic regression model using ROC and success rate curves. Results showed that the MCDA and SVR outperformed the conventional logistic regression method in the mapping of shallow landslides. Therefore, multi-criteria decision method and support vector regression were employed to determine potential landslide zones in the study area.     

Susceptibility and risk assessment of earthquake-induced landslides based on Landslide Response Units in the Subao River basin,China

Generally, pixels are the basic unit for assessment of landslide susceptibility. However, even if the results facilitate the comparison, a pixel-based analysis does not clearly illustrate the distribution relationships. To eliminate this deficiency, the concept of the Landslide Response Unit (LRU) is proposed in this study, for which adjacent pixels that have similar properties are combined as a basic unit for susceptibility assessment. The Subao River basin, seriously impacted by the Wenchuan Earthquake, was selected as the study area, and three factors including slope gradient, slope aspect, and slope shape, which have a significant impact on landslides, were chosen to divide the basin into 25,984 LRUs. Then topographic, geologic, and distance factors were applied for the landslide susceptibility evaluation. The logistic regression method was used to establish the susceptibility assessing model by analyzing 2,000 susceptible LRUs and 2,000 un-susceptible LRUs. The model accuracy was defined in terms of the ROC curve value and the κ value, 0.531 and 0.84, respectively. The susceptibility of landslides was divided into low, moderate, high, and very high in Subao River basin, and 73% of historical landslides and all four new landslides are in the highly susceptible zone and very highly susceptible zones. Finally, the LRUs with houses, farmlands, and roads prone to sliding and burial hazard were assessed separately. On the basis of considering the potential movement directions of the LRUs, the result found that 1,001 and 835 LRUs probably would be destroyed by slope sliding and landslide burial, respectively.     

基于GIS与WOE-BP模型的滑坡易发性评价

区域滑坡易发性研究对地质灾害风险管理具有重要意义.以往研究中,将多元统计模型与机器学习方法相结合用于滑坡易发性评价的研究较少.以三峡库区万州区为例,首先选取9种指标因子（坡度、坡向、剖面曲率、地表纹理、地层岩性、斜坡结构、地质构造、水系分布及土地利用类型）作为滑坡易发性评价指标.基于证据权模型（weights of evidence,WOE）计算得到的对比度和滑坡面积比与分级面积比的相对大小,对各指标因子进行状态分级;再利用粒子群法优化的BP神经网络模型（PSO-BP）得到各指标因子权重.综合两种模型确定的状态分级权重和指标因子权重（WOE-BP）计算滑坡易发性指数（landslide susceptibility index,LSI）,基于GIS平台得到全区滑坡易发性分区图.结果表明:水系、地层岩性和地质构造是影响万州区滑坡发育的主要指标因子;WOE-BP模型的预测精度为80.8%,优于WOE模型的73.1%和BP神经网络模型的71.6%,可为定量计算指标因子权重和优化滑坡易发性评价提供有效途径.      

Assessing the spatial probability of landslides using GIS and informative value model in the Bamenda highlands

Landslides are a major natural hazard in the Bamenda highlands of Cameroon, and their occurrence in this region has most often been studied using qualitative methods. The aim of this research is to quantitatively assess the spatial probability of landslides using GIS and the informative value model. Landslide inventory was done through literature review, aerial photo-interpretation, participatory GIS and field survey. Six geo-environmental factors including slope, curvature, aspect, land use, lithology and geomorphology were used as landslide conditioning (static) factors. The susceptibility of the area to future landslide events was assessed by making a correlation between past landslides and geo-environmental factors using the informative value model. The landslide inventory involving 110 landslides was divided into two equal groups using random division criterion and was used to train and validate the model. The analysis showed that slope and land use are the most important causal factors of landslides in the area. The susceptibility index map predicted most landslides to occur around the steep slopes of the Bamenda escarpment that is being used for multiple anthropic activities. The training model had a success rate of 87%, and the validation model had a prediction rate of 90%. The prediction rate curve shows that 44, 32, 18 and 6% of future landslides will occur on 3, 8, 21 and 68% of the study area. The model correctly classified 89% of unstable areas and 81% of the stable areas with an accuracy rate of 0.90. This quantitative result complement other qualitative assessment results that show the Bamenda escarpment zone as a high-risk area. However, the area susceptible to landslide in this study goes beyond what earlier studies had indicated as houses and other infrastructure were found on old landslide sites whose scars have been eroded by human activities. This new input thus improves the quality of information placed at the disposal of civil protection units and land use managers during decision making.     

降雨型滑坡时间概率的逻辑回归拟合及连续概率滑坡危险性建模

提高降雨型滑坡危险性预警精度和空间辨识度具有重要意义.以江西宁都县1980—2001年156个降雨型滑坡为例,首先基于传统的EE-D（early effective rainfall-rainfall duration）阈值法计算不同降雨诱发滑坡的时间概率级别;然后以各级别临界降雨阈值曲线对应的时间概率为因变量,并以对应的前期有效降雨量（early effective rainfall,EE）和降雨历时（D）为自变量,采用逻辑回归拟合出上述因变量与自变量之间的非线性关系,得到降雨诱发滑坡的连续概率值;之后对比C5.0决策树和多层感知器的滑坡易发性预测性能;最后利用降雨诱发滑坡的连续概率值与易发性图相耦合以实现连续概率滑坡危险性预警.结果显示:（1）宁都降雨型滑坡连续概率值的逻辑回归方程为1/P=1+e4.062+0.747 4×D-0.079 44×EE,其拟合优度为0.983;（2）2002—2003年的20处用于连续概率阈值测试的降雨型滑坡大都落在连续概率值大于0.7的区域,只有4处落在小于0.7的区域;（3）C5.0决策树预测滑坡易发性的精度显著高于多层感知器;（4）近5年的4次降雨型滑坡的连续概率危险性值都在0.8以上,且高和极高预警区的面积较传统滑坡危险性分区更小.可见连续概率滑坡危险性预警法相较于传统危险性分区法具有更高的预警精度和空间辨识度,且通过叠加滑坡易发性图及其临界降雨阈值可开展实时滑坡危险性预警制图.