Susceptibility assessments of the Wenchuan earthquake-triggered landslides in Longnan using logistic regression

In this study a Wenchuan earthquake-induced landslide susceptibility assessment was carried out in the Longnan area in northwestern China using a GIS-based logistic regression model. This region has frequently been affected by landslides in the past, and was intensively affected by the 5.12 Wenchuan earthquake which received considerable international attention. The data used for this study consist of the landslides triggered by the Wenchuan earthquake and a landslide pre-disposing factor database. Information regarding the landslide causative factors came from additional data sources, such as a digital elevation model (DEM) with a 30 × 30 m² resolution, orthophotos, geological and land-use maps, precipitation records, and information on peak ground acceleration data from the 2008 earthquake. The statistical analysis of the relationship between the Wenchuan earthquake-triggered landslides and pre-disposing factors showed the great influence of lithological and topographical conditions on slope failures. The quality of susceptibility mapping was validated by splitting the study area into training and validation sections. The prediction capability analysis demonstrated that the landslide susceptibility map could be used for land planning as well as emergency planning by local authorities.     

Controlling parameter analyses and hazard mapping for earthquake-triggered landslides: an example from a square region in Beichuan County, Sichuan Province, China

On May 12, 2008, at 1428 hours (Beijing time), a catastrophic earthquake, with a magnitude of Ms 8.0, struck the Sichuan Province, China. About 200,000 landslides, as a secondary geological hazard associated with the earthquake, were triggered over a broad area. These landslides were of almost all types such as shallow, disrupted landslides, rock falls, deep-seated landslides, and rock avalanches. Some of these landslides damaged and destroyed large part of some towns, blocked roads, dammed rivers, and caused other serious damages. The purpose of this study is to detect correlations between landslide occurrence and the surface rupture plane, ground shaking conditions (measured by peak ground acceleration, PGA), lithology, slope gradient, slope aspect, topographic position, and distance from drainages by using two indices, landslide area percentage (LAP) and the landslide number density (LND), based on geographic information system (GIS) technology and statistical analysis method in a square region (study area) of Beichuan County, Sichuan Province, China. There were 5,096 landslides related with the earthquake which were delineated by visual interpretation and selected field checking throughout the study area. The total area (horizontal projection) of the 5,096 landslides is about 41.103 km². The LAP, which is defined as the percentage of the plane area affected by landslides, was 10.276 %, and the LND, means the number of landslides per square kilometers, was 12.74 landslides/km². Statistical analysis results show that both LAP and LND have a positive correlation with slope gradient and a negative correlation with distance from the surface rupture. However, the correlation between the occurrence of landslides with PGA, topographic position, and distance from drainages are uncertain, or has just a little positive correlation. The correlation between landslide and slope aspect also shows the effect of the directivity of the seismic wave. The Zbq formation had the most concentrated landslide activity with the LND value of 21.78 landslides/km _, ² and the ∈₁ q Gr. geological units had the highest LAP value. Furthermore, weight index (W _i) model is performed with a GIS platform to derive landslide hazard index map. The success rate of the model was 71.615 % and, thus, it was valid. In addition, comparison of five landslide controlling parameters’ influence on landslide occurrences was also carried out.     

Landslides triggered by slipping-fault-generated earthquake on a plateau: an example of the 14 April 2010, Ms 7.1, Yushu, China earthquake

On 14 April 2010 at 07:49 (Beijing time), a catastrophic earthquake with Ms 7.1 struck Yushu County, Qinghai Province, China. A total of 2,036 landslides were interpreted from aerial photographs and satellite images, verified by selected field checking. These landslides cover about a total area of 1.194 km². The characteristics and failure mechanisms of these landslides are presented in this paper. The spatial distribution of the landslides is evidently strongly controlled by the locations of the main co-seismic surface fault ruptures. The landslides commonly occurred close together. Most of the landslides are small; there were only 275 individual landslide (13.5 % of the total number) surface areas larger than 1,000 m². The landslides are of various types. They are mainly shallow, disrupted landslides, but also include rock falls, deep-seated landslides, liquefaction-induced landslides, and compound landslides. Four types of factors are identified as contributing to failure along with the strong ground shaking: natural excavation of the toes of slopes, which mean erosion of the base of the slope, surface water infiltration into slopes, co-seismic fault slipping at landslide sites, and delayed occurrence of landslides due to snow melt or rainfall infiltration at sites where slopes were weakened by the co-seismic ground shaking. To analyze the spatial distribution of the landslides, the landslide area percentage (LAP) and landslide number density (LND) were compared with peak ground acceleration (PGA), distance from co-seismic main surface fault ruptures, elevation, slope gradient, slope aspect, and lithology. The results show landslide occurrence is strongly controlled by proximity to the main surface fault ruptures, with most landslides occurring within 2.5 km of such ruptures. There is no evident correlation between landslide occurrences and PGA. Both LAP and LND have strongly positive correlations with slope gradient, and additionally, sites at elevations between 3,800 and 4,000 m are relatively susceptible to landslide occurrence; as are slopes with northeast, east, and southeast slope aspects. Q₄ al-pl, N, and T₃ kn ¹ have more concentrated landslide activity than others. This paper provides a detailed inventory map of landslides triggered by the 2010 Yushu earthquake for future seismic landslide hazard analysis and also provides a study case of characteristics, failure mechanisms, and spatial distribution of landslides triggered by slipping-fault generated earthquake on a plateau.     

Regional-scale landslide activity and landslide susceptibility zonation in the Nepal Himalaya

Landslide susceptibility zonation mapping is a fundamental procedure for geo-disaster management in tropical and sub-tropical regions. Recently, various landslide susceptibility zonation models have been introduced in Nepal with diverse approaches of assessment. However, validation is still a problem. Additionally, the role of various predisposing causative parameters for landslide activity is still not well understood in the Nepal Himalaya. To address these issues of susceptibility zonation and landslide activity, about 4,000 km² area of central Nepal was selected for regional-scale assessment of landslide activity and susceptibility zonation mapping. In total, 655 new landslides and 9,229 old landslides were identified with the study area with the help of satellite images, aerial photographs, field data and available reports. The old landslide inventory was “blind landslide database” and could not explain the particular rainfall event responsible for the particular landslide. But considering size of the landslide, blind landslide inventory was reclassified into two databases: short-duration high-intensity rainfall-induced landslide inventory and long-duration low-intensity rainfall-induced landslide inventory. These landslide inventory maps were considered as proxy maps of multiple rainfall event-based landslide inventories. Similarly, all 9,884 landslides were considered for the activity assessment of predisposing causative parameters. For the Nepal Himalaya, slope, slope aspect, geology and road construction activity (anthropogenic cause) were identified as most affective predisposing causative parameters for landslide activity. For susceptibility zonation, multivariate approach was considered and two proxy rainfall event-based landslide databases were used for the logistic regression modelling, while a relatively recent landslide database was used in validation. Two event-based susceptibility zonation maps were merged and rectified to prepare the final susceptibility zonation map and its prediction rate was found to be more than 82 %. From this work, it is concluded that rectification of susceptibility zonation map is very appropriate and reliable. The results of this research contribute to a significant improvement in landslide inventory preparation procedure, susceptibility zonation mapping approaches as well as role of various predisposing causative parameters for the landslide activity.     

Statistical analysis of landslides caused by the Mw 6.9 Yushu,China, earthquake of April 14, 2010

The April 14, 2010 Yushu, China, earthquake (Mw 6.9) triggered a great number of landslides. At least 2,036 co-seismic landslides, with a total coverage area of 1.194 km², were delineated by visual interpretation of aerial photographs and satellite images taken following the earthquake, and verified by field inspection. Based on the mapping results, a statistical analysis of the spatial distribution of these landslides is performed using the landslide area percentage (LAP), defined as the percentage of the area affected by the landslides, and landslide number density (LND), defined as the number of landslides per square kilometer. The purpose is to clarify how the landslides correlate the control factors, which are the elevation, slope angle, slope aspect, slope position, distance from drainages, lithology, distance from the surface rupture, and peak ground acceleration (PGA). The results show that both LAP and LND have strongly positive correlations with slope angle and negative correlations with distance from the surface rupture and distance from drainages. The highest LAP and LPD values are in places of elevations from 3,800 to 4,000 m. The slopes producing landslides are mostly facing toward NE, E, and SE. The geological units of Q₄ al-pl, N, and T₃ kn ¹ have the highest concentrations of co-seismic landslides. No apparent correlations are present between LAP and LND values and PGA. On both sides of the surface rupture, the landslide distributions are almost similar except a few exceptions, likely associated with the nature of the strike-slip seismogenic fault for this event. The bivariate statistical analysis shows that, in descending order, the earthquake-triggered landslide impact factors are distance from surface rupture > slope angle > distance from drainages > lithology > PGA. Besides, as the detailed co-seismic landslides inventories related to strike-slip earthquakes are still few compared with that of thrusting-fault earthquakes, this case study would shed new light on the subject. For instance, the landslide spatial distribution on both sides of the strike-slip seismogenic fault is rather different from that of thrusting-fault earthquakes. It reminds us to take different strategies of measures for prevention and mitigation of landslides induced by earthquakes with different mechanisms.     

Landslides induced by a low magnitude seismic sequence at continental interiors: a case study of the Santa Rosa Canyon,northeastern Mexico

Landslides caused by a low magnitude earthquake swarm (2.8?≤?M?≤?3.6) in 2012 were documented at the Santa Rosa Canyon in northeastern Mexico. Disrupted landslides from falls and slides, in both rocks and soils, were identified based on fieldwork and high-resolution satellite imagery along stream banks from natural cliffs and along the road cut in the epicentral area. Most of the landslides occurred on slopes greater than 40°, where geological features played a key role in triggering slope instabilities. The maximum distance limit for disrupted slides from the epicentral area was 7 km. The area affected by landslides during the early stage of the seismic sequence (July through August 2012) was 90 km². Landslide identification was limited in some areas by the resolution of the satellite imagery and dense cloud coverage. Both the epicentral distance and the area affected by landslides are above the global bounds reported in literature. The final landslide inventory is the first documented case of earthquake-induced landslides in northeastern Mexico.     

Spatial characteristics of landslides triggered by the 2015 M<Subscript>w</Subscript> 7.8 (Gorkha) and M<Subscript>w</Subscript> 7.3 (Dolakha) earthquakes in Nepal

High magnitude earthquakes trigger numerous landslides and their occurrences are mainly controlled by terrain parameters. We created an inventory of 15,551 landslides with a total area of 90.2 km² triggered by the 2015 M_w 7.8 (Gorkha) and M_w 7.3 (Dolakha) earthquakes in Nepal, through interpretation of very high resolution satellite images (e.g. WorldView, Pleiades, Cartosat-1 and 2, Resourcesat-2). Our spatial analysis of landslide occurrences with ground acceleration, slope, lithology and surface defomation indicated ubiquitous control of steep slope on landslides with ground acceleration as the trigger. Spatial distribution of landslides shows increasing frequency away from the Gorkha earthquake epicentre up to 130 km towards east, dropping sharply thereafter, which is an abnormal phenomenon of coseismic landslides. Landslides are laterally concentrated in three zones which matches well with the seismic rupture evolution of Gorkha earthquake, as reported through teleseismic measurements.     

Landslide susceptibility analysis based on ArcGIS and Artificial Neural Network for a large catchment in Three Gorges region,China

A landslide susceptibility map is very important and necessary to efficiently prevent and mitigate the losses brought by natural hazard for a large area. For the purpose of landslide susceptibility analysis for the whole Xiangxi catchment (3,209 km²), Artificial Neural Network (ANN) analysis was applied as the main method. The whole catchment was divided into two parts: the training area and the implementation area. The backwater area (559 km²) of Xiangxi catchment was used as the training area for the ANN method. In the training area the correlations between the landslide distribution and its causative factors, which includes lithology, slope angle, slope curvature and river network, have been analyzed based on the geological map and digital elevation model (DEM). The back-propagation training algorithm in ANN was selected to train the sample data from the training area, which were composed of input data (causative factors) and target output data (landslide occurrence), in order to find the correlations between them. Based on these correlations and input data in the implementation area (causative factors), the network output data were obtained for the implementation area. In the end, a map of landslide susceptibility, which was established by network output data, was presented for Xiangxi catchment. ArcGIS was applied to extract and quantify input information from a DEM for susceptibility analysis and also to present the result visually. As a result, a landslide susceptibility map, in which 70 % of all landslides are rightly classified in the training area (backwater area), was created for Xiangxi catchment.     

基于GIS与ANN模型的地震滑坡易发性区划

基于遥感数据、地理信息系统(GIS)技术和人工神经网络(ANN)模型,开展地震滑坡易发性区划研究.2010年4月14日玉树地震后,基于航片与卫星影像目视解译,并辅以野外调查的方法,在地震区圈定了2036处地震诱发滑坡.选择高程、坡度、坡向、斜坡曲率、坡位、与水系距离、地层岩性、与断裂距离、与公路距离、归一化植被指数(NDVI)、与同震地表破裂距离、地震动峰值加速度(PGA)共12个因子作为地震滑坡易发性评价因子.这些因子均是应用GIS技术与遥感影像处理技术,基于地形数据、地质数据、遥感数据得到.训练样本中的滑动样本有两组,一组是滑坡区整个单滑坡体的质心位置,另一组是滑坡滑源区滑前的坡体高程最高的位置.应用这12个影响因子,分别采用这两组评价样本,基于ANN模型建立地震滑坡易发性索引图,基于GIS工具建立地震滑坡易发性分级图.分别应用训练样本中滑坡分布的点数据去检验各自的结果正确率,正确率分别为81.53％与81.29％,表明ANN模型是一种高效科学的地震滑坡易发性区划模型.     

Landslides triggered by the 2016 Mj 7.3 Kumamoto,Japan, earthquake

The aim of this study is to establish a detailed and complete inventory of the landslides triggered by the Mj 7.3 (Mw 7.0) Kumamoto, Japan, earthquake sequence of 15 April 2016 (16 April in JST). Based on high-resolution (0.5–2 m) optical satellite images, we delineated 3,467 individual landslides triggered by the earthquake, occupying an area of about 6.9 km². Then they were validated by aerial photographs with very high-resolution (better than 0.5 m) and oblique field photos. Of them, 3,460 landslides are distributed in an elliptical area about 6000 km², with a NE-SW directed 120-km-long long axis and a 60-km-long NW-SE trending short axis. Most of the landslides are shallow, disrupted falls and slides, with a few flow-type slides and rock and soil avalanches. The analysis of correlation between the landslides and several control factors shows the areas of elevation 1000–1200 m, stratum of Q₃-Hvf, seismic intensity VIII and VIII+, and peak ground acceleration (PGA) 0.4–0.6 g register the highest landslide abundance. This study also discussed the relationship between the spatial pattern of the landslides and the seismotectonic structure featured by a strike-slip fault with a normal component and the volcanism in the study area.     

Landslides in the North of Lisbon Region (Portugal): Conditioning and triggering factors

Detailed geomorphological mapping carried out in 5 sample areas in the North of Lisbon Region allowed us to collect a set of geological and geomorphological data and to correlate them with the spatial occurrence of landslide. A total of 597 slope movements were identified in a total area of 61.7 km², which represents about 10 landslides per km².The main landslide conditioning factors are: lithology and geological structure, slope angle and slope morphology, land use, presence of old landslides, and human activity.The highest landslide density occurs in Cretaceous marls and marly limestones, but the largest movements are in Jurassic clays, marls and limestones.The landslide density is higher on slopes with gradients above 20 °, but the largest unstable area is found on slopes of 10 ° to 15 °, thus reflecting the presence of the biggest slope movements. There is a correlation between landslides and topographical concavities, a fact that can be interpreted as reflecting the significance of the hydrological regime in slope instability.Concerning land use, the highest density of landslides is found on slopes covered with shrub and undergrowth vegetation.About 26% of the total number of landslides are reactivation events. The presence of old landslides is particularly important in the occurrence of translational slides and complex and composite slope movements.20% of the landslide events were conditioned by anthropomorphic activity. Human's intervention manifests itself in ill-consolidated fills, cuts in potentially unstable slopes and, in a few cases, in the changing of river channels.Most slope movements in the study area exhibit a clear climatic signal. The analysis of rainfall distribution in periods of recognised slope instability allows the distinction of three situations: 1) moderate intensity rainfall episodes, responsible for minor slope movements on the bank of rivers and shallow translational slides, particularly in artificial trenches; 2) high intensity rainfall episodes, originating flash floods and most landslides triggered by bank erosion; 3) long-lasting rainfall periods, responsible for the rise of the groundwater table and triggering of landslides with deeper slip surfaces.     

GIS-based frequency ratio and index of entropy models for landslide susceptibility assessment in the Caspian forest,northern Iran

This study presents a landslide susceptibility assessment for the Caspian forest using frequency ratio and index of entropy models within geographical information system. First, the landslide locations were identified in the study area from interpretation of aerial photographs and multiple field surveys. 72 cases (70 %) out of 103 detected landslides were randomly selected for modeling, and the remaining 31 (30 %) cases were used for the model validation. The landslide-conditioning factors, including slope degree, slope aspect, altitude, lithology, rainfall, distance to faults, distance to streams, plan curvature, topographic wetness index, stream power index, sediment transport index, normalized difference vegetation index (NDVI), forest plant community, crown density, and timber volume, were extracted from the spatial database. Using these factors, landslide susceptibility and weights of each factor were analyzed by frequency ratio and index of entropy models. Results showed that the high and very high susceptibility classes cover nearly 50 % of the study area. For verification, the receiver operating characteristic (ROC) curves were drawn and the areas under the curve (AUC) calculated. The verification results revealed that the index of entropy model (AUC = 75.59 %) is slightly better in prediction than frequency ratio model (AUC = 72.68 %). The interpretation of the susceptibility map indicated that NDVI, altitude, and rainfall play major roles in landslide occurrence and distribution in the study area. The landslide susceptibility maps produced from this study could assist planners and engineers for reorganizing and planning of future road construction and timber harvesting operations.     

Landslide susceptibility assessment by Dempster–Shafer and Index of Entropy models,Sarkhoun basin,Southwestern Iran

Landslides are natural disasters often activated by interaction of different controlling environmental factors, especially in mountainous terrains. In this research, the landslide susceptibility map was developed for the Sarkhoun catchment using Index of Entropy (IoE) and Dempster–Shafer (DS) models. For this purpose, 344 landslides were mapped in GIS environment. 241 (70%) out of the landslides were selected for the modeling and the remaining (30%) were employed for validation of the models. Afterward, 10 landslide conditioning factor layers were prepared including land use, distance to drainage, slope gradient, altitude, lithology, distance to roads, distance to faults, slope aspect, Topography Wetness Index, and Stream Power Index. The relationship between the landslide conditioning factors and landslide inventory maps was determined using the IoE and DS models. In order to verify the models, the results were compared with validation landslide data not employed in training process of the models. Accordingly, Receiver Operating Characteristic (ROC) curves were applied, and Area Under the Curve (AUC) was calculated for the obtained susceptibility maps using the success (training data) and prediction (validation data) rate curves. The land use was found to be the most important factor in the study area. The AUC are 0.82, and 0.81 for success rates of the IoE, and DS models, respectively, while the prediction rates are 0.76 and 0.75. Therefore, the results of the IoE model are more accurate than the DS model. Furthermore, a satisfactory agreement is observed between the generated susceptibility maps by the models and true location of the landslides.     

Volume,gravitational potential energy reduction,and regional centroid position change in the wake of landslides triggered by the 14 April 2010 Yushu earthquake of China

In recent years, earthquake-triggered landslides have attracted much attention in the scientific community as a main form of seismic ground response. However, little work has been performed concerning the volume and gravitational potential energy reduction of earthquake-triggered landslides and their severe effect on landscape change. This paper presents a quantitative study on the volume, gravitational potential energy reduction, and change in landscape related to landslides triggered by the 14 April 2010 Yushu earthquake. At least 2,036 landslides were triggered by the earthquake. A total landslide scar area of 1.194 km² was delineated from the visual interpretation of aerial photographs and satellite images and was supported by selected field checking. In this paper, we focus on possible answers to the following five questions: (1) What is the total volume of the 2,036 landslides triggered by the earthquake, and what is the average landslide erosion thickness in the earthquake-stricken area? (2) What are the elevations of all landslide materials in relation to pre- and post-landsliding? (3) How much was the gravitational potential energy reduced due to the sliding of these landslide materials? (4) What is the average elevation change caused by these landslides in the study area? (5) What is the vertical change of the regional centroid position above sea level, as induced by these landslides? It is concluded that the total volume of the 2,036 landslides is 2.9399?×?10⁶ m³. The landslide erosion thickness throughout the study area is 2.02 mm. The materials of these landslides moved from an elevation of 4,145.243 to 4,104.697 m, resulting in a decreased distance of 40.546 m. The gravitational potential energy reduction related to the landslides triggered by the earthquake was 2.9213?×?10¹² J. The average regional elevation of the study area is 4,427.160 m, a value consistent with the assumption that the accumulated materials were remained in situ. This value changes from 4,427.160 to 4,427.158 m with all landslide materials moved out of the study area, resulting in a reduction in elevation of 2 mm. Based on the assumption that all landslide materials moved out of the study area, the elevations of the centroid of the study area’s crust changed from 2,222.45967 to 2,222.45867 m, which means the centroid value decreased by 1 mm. This value is 0.001 mm when assuming that the materials were remained in situ, which is almost negligible, compared with the situation of “all landslide materials moved out of the study area.”     

Application of frequency ratio,statistical index,and weights-of-evidence models and their comparison in landslide susceptibility mapping in Central Nepal Himalaya

The Mugling–Narayanghat road section falls within the Lesser Himalaya and Siwalik zones of Central Nepal Himalaya and is highly deformed by the presence of numerous faults and folds. Over the years, this road section and its surrounding area have experienced repeated landslide activities. For that reason, landslide susceptibility zonation is essential for roadside slope disaster management and for planning further development activities. The main goal of this study was to investigate the application of the frequency ratio (FR), statistical index (SI), and weights-of-evidence (WoE) approaches for landslide susceptibility mapping of this road section and its surrounding area. For this purpose, the input layers of the landslide conditioning factors were prepared in the first stage. A landslide inventory map was prepared using earlier reports, aerial photographs interpretation, and multiple field surveys. A total of 438 landslide locations were detected. Out these, 295 (67 %) landslides were randomly selected as training data for the modeling using FR, SI, and WoE models and the remaining 143 (33 %) were used for the validation purposes. The landslide conditioning factors considered for the study area are slope gradient, slope aspect, plan curvature, altitude, stream power index, topographic wetness index, lithology, land use, distance from faults, distance from rivers, and distance from highway. The results were validated using area under the curve (AUC) analysis. From the analysis, it is seen that the FR model with a success rate of 76.8 % and predictive accuracy of 75.4 % performs better than WoE (success rate, 75.6 %; predictive accuracy, 74.9 %) and SI (success rate, 75.5 %; predictive accuracy, 74.6 %) models. Overall, all the models showed almost similar results. The resultant susceptibility maps can be useful for general land use planning.     

GIS-based landslide susceptibility mapping with probabilistic likelihood ratio and spatial multi-criteria evaluation models (North of Tehran,Iran)

The aim of this study is to produce landslide susceptibility mapping by probabilistic likelihood ratio (PLR) and spatial multi-criteria evaluation (SMCE) models based on geographic information system (GIS) in the north of Tehran metropolitan, Iran. The landslide locations in the study area were identified by interpretation of aerial photographs, satellite images, and field surveys. In order to generate the necessary factors for the SMCE approach, remote sensing and GIS integrated techniques were applied in the study area. Conditioning factors such as slope degree, slope aspect, altitude, plan curvature, profile curvature, surface area ratio, topographic position index, topographic wetness index, stream power index, slope length, lithology, land use, normalized difference vegetation index, distance from faults, distance from rivers, distance from roads, and drainage density are used for landslide susceptibility mapping. Of 528 landslide locations, 70 % were used in landslide susceptibility mapping, and the remaining 30 % were used for validation of the maps. Using the above conditioning factors, landslide susceptibility was calculated using SMCE and PLR models, and the results were plotted in ILWIS-GIS. Finally, the two landslide susceptibility maps were validated using receiver operating characteristic curves and seed cell area index methods. The validation results showed that area under the curve for SMCE and PLR models is 76.16 and 80.98 %, respectively. The results obtained in this study also showed that the probabilistic likelihood ratio model performed slightly better than the spatial multi-criteria evaluation. These landslide susceptibility maps can be used for preliminary land use planning and hazard mitigation purpose.     

2008年汶川地震滑坡详细编目及其空间分布规律分析

最新研究成果表明, 2008年5月12日汶川MS 8.0级地震触发了超过197000处滑坡。首先,基于GIS与遥感技术构建了汶川地震滑坡的3类编目图,分别为单体滑坡面分布数据、滑坡中心点位置和滑坡后壁点位置。构建方法为基于地震前后高分辨率遥感影像的目视解译方法,区分单体滑坡并圈定其边界,对滑坡后壁进行识别与定点,并开展了部分滑坡的野外验证工作。这些滑坡分布在一个面积大约为110000km2的区域内,滑坡总面积约为1160km2。选择一个面积约为44031km2的区域作为研究区,区内滑坡数量为196007个,滑坡面积为1150.622km2,这是最详细完整的汶川地震滑坡编录成果,也是单次地震事件触发滑坡最多的记录。其次,开展研究区内的地震滑坡空间分布规律的研究。基于滑坡面与滑坡中心点分别构建滑坡空间分布面积密度图与点密度图,结果表明:滑坡多沿着映秀北川断裂分布,多发生在断裂的上盘。滑坡的高密度区位于映秀北川同震地表破裂的南西段(映秀镇与北川县之间)的上盘区域,这一区域恰对应着逆冲分量为主的断裂上盘,表明逆冲断裂对上盘区域发生滑坡的极强烈的控制作用,而该区域正是形变最大的区域,因此说明是地震滑坡发生的强烈控制作用。基于滑坡面密度(LAP)、滑坡中心点密度(LCND)与滑坡后壁点密度(LTND)这3个衡量指标,使用统计分析方法,评价了汶川地震滑坡与地震参数、地质参数、地形参数的关系。结果表明:LAP、LCND与LTND这3个衡量指标与坡度、地震烈度与PGA存在明显的正相关关系; 与距离震中、距离映秀北川同震地表破裂存在负相关关系; 斜坡曲率越接近0,滑坡越不易发生; LAP、LCND与LTND的高值高程区间为1200～3000m; 滑坡发生的优势坡向为E、SE、S方向; 滑坡发育的易发岩性为砂岩与粉砂岩(Z)、花岗岩; 滑坡与坡位的相关关系不太明显。统计结果还表明LCND与LTND两个衡量指标的差异对地震与地质因子不敏感,而对地形因子较敏感。最后将本文的统计结果与以往的汶川地震滑坡空间分布规律统计成果进行了一些对比,对比结果表明,对于某些因子,如高程、岩性、距离震中、距离映秀北川断裂的统计分析结果,采用不完整的滑坡分布数据或点数据,与采用较完整的滑坡分布面数据会有一定的差异,这种差异并未出现在针对坡度与坡向等因子的统计对比结果中。总之,作者认为一个完备、详细的地震滑坡分布面要素编目图是地震滑坡空间分布规律定量分析、危险性定量分析与滑坡控制的地震区地貌演化研究的重要基础,否则,与实际情况相比,得到统计结果会有一定的偏差,本文的研究成果与以往成果的对比结果证明了这一点。     

Landslide susceptibility mapping at Vaz Watershed (Iran) using an artificial neural network model: a comparison between multilayer perceptron (MLP) and radial basic function (RBF) algorithms

Landslide susceptibility and hazard assessments are the most important steps in landslide risk mapping. The main objective of this study was to investigate and compare the results of two artificial neural network (ANN) algorithms, i.e., multilayer perceptron (MLP) and radial basic function (RBF) for spatial prediction of landslide susceptibility in Vaz Watershed, Iran. At first, landslide locations were identified by aerial photographs and field surveys, and a total of 136 landside locations were constructed from various sources. Then the landslide inventory map was randomly split into a training dataset 70 % (95 landslide locations) for training the ANN model and the remaining 30 % (41 landslides locations) was used for validation purpose. Nine landslide conditioning factors such as slope, slope aspect, altitude, land use, lithology, distance from rivers, distance from roads, distance from faults, and rainfall were constructed in geographical information system. In this study, both MLP and RBF algorithms were used in artificial neural network model. The results showed that MLP with Broyden–Fletcher–Goldfarb–Shanno learning algorithm is more efficient than RBF in landslide susceptibility mapping for the study area. Finally the landslide susceptibility maps were validated using the validation data (i.e., 30 % landslide location data that was not used during the model construction) using area under the curve (AUC) method. The success rate curve showed that the area under the curve for RBF and MLP was 0.9085 (90.85 %) and 0.9193 (91.93 %) accuracy, respectively. Similarly, the validation result showed that the area under the curve for MLP and RBF models were 0.881 (88.1 %) and 0.8724 (87.24 %), respectively. The results of this study showed that landslide susceptibility mapping in the Vaz Watershed of Iran using the ANN approach is viable and can be used for land use planning.     

Application of fuzzy logic and analytical hierarchy process (AHP) to landslide susceptibility mapping at Haraz watershed, Iran

The main goal of this study is to produce landslide susceptibility maps of a landslide-prone area (Haraz) in Iran by using both fuzzy logic and analytical hierarchy process (AHP) models. At first, landslide locations were identified by aerial photographs and field surveys, and a total of 78 landslides were mapped from various sources. Then, the landslide inventory was randomly split into a training dataset 70?% (55 landslides) for training the models and the remaining 30?% (23 landslides) was used for validation purpose. Twelve data layers, as the landslide conditioning factors, are exploited to detect the most susceptible areas. These factors are slope degree, aspect, plan curvature, altitude, lithology, land use, distance from rivers, distance from roads, distance from faults, stream power index, slope length, and topographic wetness index. Subsequently, landslide susceptibility maps were produced using fuzzy logic and AHP models. For verification, receiver operating characteristics curve and area under the curve approaches were used. The verification results showed that the fuzzy logic model (89.7?%) performed better than AHP (81.1?%) model for the study area. The produced susceptibility maps can be used for general land use planning and hazard mitigation purpose.