Susceptibility assessment for rainfall-induced landslides using a revised logistic regression method

Natural Hazards - Landslide susceptibility is the likelihood of a landslide occurring in an area. The logistic regression (LR) method is one of the most popular methods for landslide susceptibility...     

基于有效降雨强度和逻辑回归的降雨型 滑坡预测模型

以湖南省张家界市桑植县为研究区,在全面分析近30年降雨及滑坡数据的基础上,对滑坡及滑坡数量与降雨因子的关系开展了统计分析研究。首先确定了区域最佳有效降雨衰减系数,同时分别按滑坡规模、坡度、厚度大小统计了降雨与历史滑坡信息,得出有效降雨强度（I）与持续时间（D）散点图,由此确定各不同概率下诱发滑坡的区域有效降雨强度阈值,并进行了滑坡灾害危险性等级划分。进而,利用部分样本数据进行逻辑回归分析,得到了该研究区的滑坡发生概率预测方程,并给出了降雨强度临界值定量表达式,最后选用实际降雨诱发滑坡事件与未诱发滑坡事件进行对比验证。结果表明,文章所建立的滑坡预测模型准确性较高,预测情况与实际情况比较吻合。     

Assessment of susceptibility to landslides through geographic information systems and the logistic regression model

Natural Hazards - The increase in the frequency of natural disasters in recent years and its consequent social, economic and environmental impacts make it possible to prioritize areas of risk as an...     

基于确定性系数方法的龙门山地区地震滑坡影响因子敏感性分析——以汶川地震与芦山地震诱发滑坡为例

在我国西南地区,沿龙门山断裂带分别发生了2008年汶川M_S8.0级地震以及2013年芦山M_S7.0级地震,这两次地震均造成了严重的地表破裂,并诱发了大量的滑坡和崩塌等次生地质灾害。文章选择了位于青藏高原向四川盆地过渡的区域——以龙门山断裂带为中心的30°~34°N,102°~106°E区域作为研究区,借助GIS工具,基于确定性系数（CF）方法,选取了地震、地质构造、自然环境和人类活动4大类因子,包括烈度、震中距、岩性、断裂、高程、坡度、坡向、河流、降雨、公路共10个因子（子集）,对汶川和芦山地震诱发的次生滑坡灾害进行影响因子敏感性分析,基于z值确定该区域内地震滑坡的关键因子类以及基于CF值确定各类集（子集）下各特定因子的具体地震滑坡敏感性。研究结果显示：地震因子具有最高的z值,是龙门山地区地震滑坡产生的关键影响因子,表明地震活动的强弱直接关系到斜坡的稳定性和次生滑坡灾害的面积分布。而对比烈度子集中的具体CF值表明：当烈度小于Ⅷ度时,烈度对次生滑坡发生的影响极低,区域内的主要影响因子则由地震因子转变为震中距、自然因子等其他类别的因子;其次以坡度、高程、与河流的距离为主的自然因子类别以及与断层的距离在地震滑坡过程中也有较高权重,而人类活动对研究区内坡体的稳定性也有着不可忽略的作用。本研究结果可作为该地区后续区域地震滑坡相关研究和发展规划的基础科学依据。

     

A subaqueous hazard map for earthquake-triggered landslides in Lake Zurich,Switzerland

The awareness of geohazards in the subaqueous environment has steadily increased in the past years and there is an increased need to assess these hazards in a quantitative sense. Prime examples are subaqueous landslides, which can be triggered by a number of processes including earthquakes or human activities, and which may impact offshore and onshore infrastructure and communities. In the literature, a plenitude of subaqueous landslide events are related to historical earthquakes, including cases from lakes in Switzerland. Here, we present an approach for a basin-wide earthquake-triggered subaquatic landslide hazard assessment for Lake Zurich, which is surrounded by a densely populated shoreline. Our analysis is based on high-resolution sediment-mechanical and geophysical input data. Slope stabilities are calculated with a grid-based limit equilibrium model on an infinite slope, which uses Monte Carlo sampled input data from a sediment-mechanical stratigraphy of the lateral slopes. Combined with probabilistic ground-shaking forecasts from a recent national seismic hazard analysis, subaquatic earthquake-triggered landslide hazard maps are constructed for different mean return periods, ranging from 475 to 9975 years. Our results provide a first quantitative landslide hazard estimation for the lateral slopes in Lake Zurich. Furthermore, a back-analysis of a case-study site indicates that pseudostatic accelerations in the range between 0.04 and 0.08 g were needed to trigger a well-investigated subaqueous landslide, dated to ~2210 cal. years B.P.     

基于逻辑回归模型的汶川地震滑坡危险性评价与检验

以2008年5月12日汶川地震区为研究区,基于高分辨率航片与卫星影像开展地震滑坡目视解译,制作了汶川地震滑坡编录图.选择坡度、坡向、高程、与水系距离、与公路距离、与映秀-北川断裂距离、地震烈度、岩性共8个影响因子开展地震滑坡危险性评价工作.滑坡样本采用前期48007处滑坡编录点数据,不滑样本为在基于证据权重模型的滑坡危险性评价结果的低危险区与极低危险区随机选择的48000个点.基于这8个影响因子与逻辑回归模型,建立了汶川地震滑坡危险性索引图.采用这48007个滑坡样本点与汶川地震滑坡最新编录的增加滑坡,分别进行模型的成功率与预测率检验.结果表明,模型成功率为81.739％,预测率达到86.278％.     

Susceptibility mapping of landslides in Beichuan County using cluster and MLC methods

Cluster analysis and maximum likelihood classification (MLC) are exploited to map the post-earthquake landslide susceptibility in Beichuan County that was affected by the Ms 8.0 Wenchuan earthquake. The methodology is applicable even if there is short of training data. Six effective factors are chosen for mapping the susceptibility, including land use, seismic intensity, average annual rainfall, relative relief, slop gradient and lithology. Four clusters are grouped from sampling grid cells by k-means clustering approach. MLC classifies all the cells in the study area into the four clusters according to their statistical characteristics. Four susceptibility classes (extreme low, low, moderate and high) are assigned to these clusters applying expert experience and hazard density. The final map gives a reasonable assessment of post-earthquake landslide susceptibility in Beichuan County. Comparing with the pre-earthquake susceptibility map made in Beichuan County geological disaster survey project, the result t using cluster and MLC classification has a better agreement with the dot density value of post-earthquake landslides in Beichuan County. The susceptibility map can be used to identify safety spots within the high danger area, which are suitable for habitations and facilities. It is also found that more landslides are densely concentrated at the boundary between high and moderate regions, and between high and extreme low regions.     

Susceptibility analysis of shallow landslides source areas using physically based models

Rainfall-induced shallow landslides of the flow-type involve different soils, and they often cause huge social and economical disasters, posing threat to life and livelihood all over the world. Due to the frequent large extension of the rainfall events, these landslides can be triggered over large areas (up to tens of square kilometres), and their source areas can be analysed with the aid of distributed, physically based models. Despite the high potential, such models show some limitations related to the adopted simplifying assumptions, the quantity and quality of required data, as well as the use of a quantitative interpretation of the results. A relevant example is provided in this paper referring to catastrophic phenomena involving volcaniclastic soils that frequently occur in southern Italy. Particularly, three physically based models (SHALSTAB, TRIGRS and TRIGRS-unsaturated) are used for the analysis of the source areas of huge rainfall-induced shallow landslides occurred in May 1998 inside an area of about 60 km². The application is based on an extensive data set of topographical, geomorphological and hydrogeological features of the affected area, as well as on both stratigraphical settings and mechanical properties of the involved soils. The results obtained from the three models are compared by introducing two indexes aimed at quantifying the “success” and the “error” provided by each model in simulating observed source areas. Advantages and limitations of the adopted models are then discussed for their use in forecasting the rainfall-induced source areas of shallow landslides over large areas.     

Landslide susceptibility mapping of the Sera River Basin using logistic regression model

Of the natural hazards in Turkey, landslides are the second most devastating in terms of socio-economic losses, with the majority of landslides occurring in the Eastern Black Sea Region. The aim of this study is to use a statistical approach to carry out a landslide susceptibility assessment in one area at great risk from landslides: the Sera River Basin located in the Eastern Black Sea Region. This paper applies a multivariate statistical approach in the form of a logistics regression model to explore the probability distribution of future landslides in the region. The model attempts to find the best fitting function to describe the relationship between the dependent variable, here the presence or absence of landslides in a region and a set of independent parameters contributing to the occurrence of landslides. The dependent variable (0 for the absence of landslides and 1 for the presence of landslides) was generated using landslide data retrieved from an existing database and expert opinion. The database has information on a few landslides in the region, but is not extensive or complete, and thus unlike those normally used for research. Slope, angle, relief, the natural drainage network (including distance to rivers and the watershed index) and lithology were used as independent parameters in this study. The effect of each parameter was assessed using the corresponding coefficient in the logistic regression function. The results showed that the natural drainage network plays a significant role in determining landslide occurrence and distribution. Landslide susceptibility was evaluated using a predicted map of probability. Zones with high and medium susceptibility to landslides make up 38.8 % of the study area and are located mostly south of the Sera River Basin and along streams.     

Landslide hazard mapping using logistic regression model in Mackenzie Valley,Canada

A logistic regression model is developed within the framework of a Geographic Information System (GIS) to map landslide hazards in a mountainous environment. A case study is conducted in the mountainous southern Mackenzie Valley, Northwest Territories, Canada. To determine the factors influencing landslides, data layers of geology, surface materials, land cover, and topography were analyzed by logistic regression analysis, and the results are used for landslide hazard mapping. In this study, bedrock, surface materials, slope, and difference between surface aspect and dip direction of the sedimentary rock were found to be the most important factors affecting landslide occurrence. The influence on landslides by interactions among geologic and geomorphic conditions is also analyzed, and used to develop a logistic regression model for landslide hazard mapping. The comparison of the results from the model including the interaction terms and the model not including the interaction terms indicate that interactions among the variables were found to be significant for predicting future landslide probability and locating high hazard areas. The results from this study demonstrate that the use of a logistic regression model within a GIS framework is useful and suitable for landslide hazard mapping in large mountainous geographic areas such as the southern Mackenzie Valley.     

Run-out analysis of flow-like landslides triggered by the Ms 8.0 2008 Wenchuan earthquake using smoothed particle hydrodynamics

Flow-like landslides have caused significant damage and casualties worldwide. However, studying such phenomena with traditional simulation methods is made difficult by their complex fluidization characteristics. In this paper, we use smoothed-particle hydrodynamics (SPH) for the run-out analysis of flow-like landslides. Compared with conventional methods, the proposed SPH modeling technique is the combination of a Bingham flow model and Navier?CStokes equations in the framework of computational fluid dynamics. At first, two benchmark problems of dam break and granular flow are simulated and verified to evaluate the accuracy of the SPH model. Then, run-out analyses are performed for flow-like landslides triggered by the Ms 8.0 Wenchuan earthquake that occurred on 12 May 2008 in Sichuan Province, China. Run-out analyses of the Tangjiashan, Wangjiayan, and Donghekou landslides are conducted by the application of SPH models to real flow-like landslides. All simulations show good agreement with characteristics of flow-like landslides observed in the field. We have found that numerical modeling can capture the fundamental dynamic behavior of these flow-like landslides and produce preliminary results for hazard assessment and site selection for reconstruction in earthquake-prone areas.     

Landslide susceptibility analysis in the Hoa Binh province of Vietnam using statistical index and logistic regression

The purpose of this study is to evaluate and compare the results of applying the statistical index and the logistic regression methods for estimating landslide susceptibility in the Hoa Binh province of Vietnam. In order to do this, first, a landslide inventory map was constructed mainly based on investigated landslide locations from three projects conducted over the last 10 years. In addition, some recent landslide locations were identified from SPOT satellite images, fieldwork, and literature. Secondly, ten influencing factors for landslide occurrence were utilized. The slope gradient map, the slope curvature map, and the slope aspect map were derived from a digital elevation model (DEM) with resolution 20 × 20 m. The DEM was generated from topographic maps at a scale of 1:25,000. The lithology map and the distance to faults map were extracted from Geological and Mineral Resources maps. The soil type and the land use maps were extracted from National Pedology maps and National Land Use Status maps, respectively. Distance to rivers and distance to roads were computed based on river and road networks from topographic maps. In addition, a rainfall map was included in the models. Actual landslide locations were used to verify and to compare the results of landslide susceptibility maps. The accuracy of the results was evaluated by ROC analysis. The area under the curve (AUC) for the statistical index model was 0.946 and for the logistic regression model, 0.950, indicating an almost equal predicting capacity.     

Landslide susceptibility zonation of the Chamoli region, Garhwal Himalayas, using logistic regression model

A remote sensing and Geographic Information System-based study has been carried out for landslide susceptibility zonation in the Chamoli region, part of Garhwal Himalayas. Logistic regression has been applied to correlate the presence of landslides with independent physical factors including slope, aspect, relative relief, land use/cover, lithology, lineament, and drainage density. Coefficients of the categories of each factor have been obtained and used to assess the landslide probability value to ultimately categorize the area into various landslide susceptibility zones; very low, low, moderate, high, and very high. The results show that 71.13% of observed landslides fall in 21.96% of predicted very high and high susceptibility zone, which in fact should be the case. Furthermore, lineament first buffer category (0–500 m) and the east and south aspects are the most influential in causing landslides in the region.     

Correction to: Landslide susceptibility mapping of the Sera River Basin using logistic regression model

Natural Hazards - The article “Landslide susceptibility mapping of the Sera River Basin using logistic regression model,” written by Nussaïbah B. Raja, Ihsan Çiçek, Necla...     

A comparative study on the landslide susceptibility mapping using logistic regression and statistical index models

The logistic regression and statistical index models are applied and verified for landslide susceptibility mapping in Daguan County, Yunnan Province, China, by means of the geographic information system (GIS). A detailed landslide inventory map was prepared by literatures, aerial photographs, and supported by field works. Fifteen landslide-conditioning factors were considered: slope angle, slope aspect, curvature, plan curvature, profile curvature, altitude, STI, SPI, and TWI were derived from digital elevation model; NDVI was extracted from Landsat ETM7; rainfall was obtained from local rainfall data; distance to faults, distance to roads, and distance to rivers were created from a 1:25,000 scale topographic map; the lithology was extracted from geological map. Using these factors, the landslide susceptibility maps were prepared by LR and SI models. The accuracy of the results was verified by using existing landslide locations. The statistical index model had a predictive rate of 81.02%, which is more accurate prediction in comparison with logistic regression model (80.29%). The models can be used to land-use planning in the study area.     

Susceptibility regional zonation of earthquake-induced landslides in Campania,Southern Italy

In this paper, we present a GIS-based method for regional zoning of seismic-induced landslide susceptibility and show its application to the territory of the Campania region, Southern Italy. The method employs only three factors that we believe are most significant in the susceptibility assessment: the type of outcropping rock/soil, the slope angle, and the MCS intensity. Each of the three parameters is quantified in terms of relative weight expressed as indices, and the resulting Seismic Landslide Susceptibility index of an area is given by the average of the indices of the first two factors multiplied by the index of the third factor. The result of this susceptibility zonation applied to Campania shows a good agreement between the distribution of the historical earthquake-triggered landslides and the highly susceptible zones.     

逻辑回归模型在边坡稳定性分析中的应用

采用逻辑回归模型和确定性系数CF建立概率模型,使用收集到的22个边坡实例,将重度、内聚力、坡角、坡高、内摩擦角和孔隙压力比六个影响因子作为自变量,边坡状态作为因变量,应用SPSS取不同的滑坡概率P与不滑坡概率1-P的比值的自然对数Z进行回归分析。将获得的滑坡概率拟合模型反馈到样本参数上,对比其误差,选取Z=±10时的函数模型对另外8组边坡进行验证,证明预测模型的合理性。