A Novel Classifier Based on Composite Hyper-cubes on Iterated Random Projections for Assessment of Landslide Susceptibility

In this paper, the main objective is to discover an application of a novel classifier based on Composite Hyper-cubes on Iterated Random Projections (CHIRP) for assessment of landslide susceptibility at the Uttarakhand Area (India). For this, 1295 historical landslides events and landslide affecting parameters were collected and used for creating training and testing datasets. Other benchmark models namely Logistic Regression (LR), RBF neural network (ANN-RBF), and Naïve Bayes (NB) were chosen for comparison. Analysis results indicate that the CHIRP is the best, followed by the LR, the ANN-RBF, and the NB, respectively. Overall, the CHIRP indicates as a promising and good alternative method that could be used to assess landslide susceptibility in other landslide prone areas.     

A novel hybrid integration model using support vector machines and random subspace for weather-triggered landslide susceptibility assessment in the Wuning area (China)

This study proposed a hybrid modeling approach using two methods, support vector machines and random subspace, to create a novel model named random subspace-based support vector machines (RSSVM) for assessing landslide susceptibility. The newly developed model was then tested in the Wuning area, China, to produce a landslide susceptibility map. With the purpose of achieving the objective of the study, a spatial dataset was initially constructed that includes a landslide inventory map consisting of 445 landslide regions. Then, various landslide-influencing factors were defined, including slope angle, aspect, altitude, topographic wetness index, stream power index, sediment transport index, soil, lithology, normalized difference vegetation index, land use, rainfall, distance to roads, distance to rivers, and distance to faults. Next, the result of the RSSVM model was validated using statistical index-based evaluations and the receiver operating characteristic curve approach. Then, to evaluate the performance of the suggested RSSVM model, a comparison analysis was performed to other existing approaches such as artificial neural network, Naïve Bayes (NB) and support vector machine (SVM). In general, the performance of the RSSVM model was better than the other models for spatial prediction of landslide susceptibility. The AUC results of the applied models are as follows: RSSVM (AUC = 0.857), followed by MLP (AUC = 0.823), SVM (AUC = 0.814) and NB (AUC = 0.783). The present study indicates that RSSVM can be used for landslide susceptibility evaluation, and the results are very useful for local governments and people living in the Wuning area.     

随机森林与GIS的泥石流易发性及可靠性

目前基于GIS的泥石流易发性（简称DFS）评价模型中,统计类型模型的因子须保证独立性,且权重受区间划分控制;线性机器学习难以处理非线性问题、而常用非线性模型调试效率低.鉴于随机森林（RF）能有效克服常用模型的诸多不足,且在DFS评价中的应用极少,首先展开基于RF的DFS评价,采用线性、RBF支持向量机、二次判别分析、RF等经贝叶斯优化的模型和26种泥石流影响因子;然后,分别以RF的相对权重排序和蒙特卡洛方法研究因子组合和建模样本变化下DFS评价的可靠性.结果表明:RF不易发和较易发区中有21个因子可指示泥石流孕育环境差异;RF的相对权重排序能有效确定易发模型的局部最优因子组合;随机样本划分导致的评价不确定性在中易发区最大,应通过提高建模样本比例和改善模型降低;RF的预测能力指标AUC为0.86、全局预测精度为0.79、F1分数为0.66、brier分数为0.14,以及它们的可靠度最优,可作为DFS定量评估的优先选择.      

Applying deep learning and benchmark machine learning algorithms for landslide susceptibility modelling in Rorachu river basin of Sikkim Himalaya,India

Landslide is considered as one of the most severe threats to human life and property in the hilly areas of the world. The number of landslides and the level of damage across the globe has been increasing over time. Therefore, landslide management is essential to maintain the natural and socio-economic dynamics of the hilly region. Rorachu river basin is one of the most landslide-prone areas of the Sikkim selected for the present study. The prime goal of the study is to prepare landslide susceptibility maps(LSMs) using computer-based advanced machine learning techniques and compare the performance of the models.To properly understand the existing spatial relation with the landslide, twenty factors, including triggering and causative factors, were selected. A deep learning algorithm viz. convolutional neural network model(CNN) and three popular machine learning techniques, i.e., random forest model(RF), artificial neural network model(ANN), and bagging model, were employed to prepare the LSMs. Two separate datasets including training and validation were designed by randomly taken landslide and nonlandslide points. A ratio of 70:30 was considered for the selection of both training and validation points.Multicollinearity was assessed by tolerance and variance inflation factor, and the role of individual conditioning factors was estimated using information gain ratio. The result reveals that there is no severe multicollinearity among the landslide conditioning factors, and the triggering factor rainfall appeared as the leading cause of the landslide. Based on the final prediction values of each model, LSM was constructed and successfully portioned into five distinct classes, like very low, low, moderate, high, and very high susceptibility. The susceptibility class-wise distribution of landslides shows that more than 90% of the landslide area falls under higher landslide susceptibility grades. The precision of models was examined using the area under the curve(AUC) of the receiver operating characteristics(ROC) curve and statistical methods like root mean square error(RMSE) and mean absolute error(MAE). In both datasets(training and validation), the CNN model achieved the maximum AUC value of 0.903 and 0.939, respectively. The lowest value of RMSE and MAE also reveals the better performance of the CNN model. So, it can be concluded that all the models have performed well, but the CNN model has outperformed the other models in terms of precision.     

Landslide susceptibility mapping using machine learning algorithms and comparison of their performance at Abha Basin,Asir Region,Saudi Arabia

The current study aimed at evaluating the capabilities of seven advanced machine learning techniques(MLTs),including,Support Vector Machine(SVM),Random Forest(RF),Multivariate Adaptive Regression Spline(MARS),Artificial Neural Network(ANN),Quadratic Discriminant Analysis(QDA),Linear Discriminant Analysis(LDA),and Naive Bayes(NB),for landslide susceptibility modeling and comparison of their performances.Coupling machine learning algorithms with spatial data types for landslide susceptibility mapping is a vitally important issue.This study was carried out using GIS and R open source software at Abha Basin,Asir Region,Saudi Arabia.First,a total of 243 landslide locations were identified at Abha Basin to prepare the landslide inventory map using different data sources.All the landslide areas were randomly separated into two groups with a ratio of 70%for training and 30%for validating purposes.Twelve landslide-variables were generated for landslide susceptibility modeling,which include altitude,lithology,distance to faults,normalized difference vegetation index(NDVI),landuse/landcover(LULC),distance to roads,slope angle,distance to streams,profile curvature,plan curvature,slope length(LS),and slope-aspect.The area under curve(AUC-ROC)approach has been applied to evaluate,validate,and compare the MLTs performance.The results indicated that AUC values for seven MLTs range from 89.0%for QDA to 95.1%for RF.Our findings showed that the RF(AUC=95.1%)and LDA(AUC=941.7%)have produced the best performances in comparison to other MLTs.The outcome of this study and the landslide susceptibility maps would be useful for environmental protection.     

Applying different scenarios for landslide spatial modeling using computational intelligence methods

Landslides every year impose extensive damages to human beings in various parts of the world; therefore, identifying prone areas to landslides for preventive measures is essential. The main purpose of this research is applying different scenarios for landslide susceptibility mapping by means of combination of bivariate statistical (frequency ratio) and computational intelligence methods (random forest and support vector machine) in landslide polygon and point formats. For this purpose, in the first step, a total of 294 landslide locations were determined from various sources such as aerial photographs, satellite images, and field surveys. Landslide inventory was randomly split into a testing dataset 70% (206 landslide locations) for training the different scenarios, and the remaining 30% (88 landslides locations) was used for validation purposes. To providing landslide susceptibility maps, 13 conditioning factors including altitude, slope angle, plan curvature, slope aspect, topographic wetness index, lithology, land use/land cover, distance from rivers, drainage density, distance from fault, distance from roads, convergence index, and annual rainfall are used. Tolerance and the variance inflation factor indices were used for considering multi-collinearity of conditioning factors. Results indicated that the smallest tolerance and highest variance inflation factor were 0.31 and 3.20, respectively. Subsequently, spatial relationship between classes of each landslide conditioning factor and landslides was obtained by frequency ratio (FR) model. Also, importance of the mentioned factors was obtained by random forest (RF) as a machine learning technique. The results showed that according to mean decrease accuracy, factors of altitude, aspect, drainage density, and distance from rivers had the greatest effect on the occurrence of landslide in the study area. Finally, the landslide susceptibility maps were produced by ten scenarios according to different ensembles. The receiver operating characteristics, including the area under the curve (AUC), were used to assess the accuracy of the models. Results of validation of scenarios showed that AUC was varying from 0.668 to 0.749. Also, FR and seed cell area index indicators show a high correlation between the susceptibility classes with the landslide pixels and field observations in all scenarios except scenarios 10_RF and 10_SVM. The results of this study can be used for landslides management and mitigation and development activities such as construction of settlements and infrastructure in the future.     

Multi-geohazards susceptibility mapping based on machine learning—a case study in Jiuzhaigou,China

Jiuzhaigou, located in the transitional area between the Qinghai–Tibet Plateau and the Sichuan Basin, is highly prone to geological hazards (e.g., rock fall, landslide, and debris flow). High-performance-based hazard prediction models, therefore, are urgently required to prevent related hazards and manage potential emergencies. Current researches mainly focus on susceptibility of single hazard but ignore that different types of geological hazards might occur simultaneously under a complex environment. Here, we firstly built a multi-geohazard inventory from 2000 to 2015 based on a geographical information system and used satellite data in Google earth and then chose twelve conditioning factors and three machine learning methods—random forest, support vector machine, and extreme gradient boosting (XGBoost)—to generate rock fall, landslide, and debris flow susceptibility maps. The results show that debris flow models presented the best prediction capabilities [area under the receiver operating characteristic curve (AUC 0.95)], followed by rock fall (AUC 0.94) and landslide (AUC 0.85). Additionally, XGBoost outperformed the other two methods with the highest AUC of 0.93. All three methods with AUC values larger than 0.84 suggest that these models have fairly good performance to assess geological hazards susceptibility. Finally, evolution index was constructed based on a joint probability of these three hazard models to predict the evolution tendency of 35 unstable slopes in Jiuzhaigou. The results show that these unstable slopes are likely to evolve into debris flows with a probability of 46%, followed by landslides (43%) and rock falls (29%). Higher susceptibility areas for geohazards were mainly located in the southeast and middle of Jiuzhaigou, implying geohazards prevention and mitigation measures should be taken there in near future.

     

Landslide susceptibility mapping using a modified decision tree classifier in the Xanthi Perfection,Greece

The objective of this study was to validate the outcomes of a modified decision tree classifier by comparing the produced landslide susceptibility map and the actual landslide occurrence, in an area of intensive landslide manifestation, in Xanthi Perfection, Greece. The values that concerned eight landslide conditioning factors for 163 landslides and 163 non-landslide locations were extracted by using advanced spatial GIS functions. Lithological units, elevation, slope angle, slope aspect, distance from tectonic features, distance from hydrographic network, distance from geological boundaries and distance from road network were among the eight landslide conditioning factors that were included in the landslide database used in the training phase. In the present study, landslide and non-landslide locations were randomly divided into two subsets: 80 % of the data (260 instances) were used for training and 20 % of the data (66 instances) for validating the developed classifier. The outcome of the decision tree classifier was a set of rules that expressed the relationship between landslide conditioning factors and the actual landslide occurrence. The landslide susceptibility belief values were obtained by applying a statistical method, the certainty factor method, and by measuring the belief in each rule that the decision tree classifier produced, transforming the discrete type of result into a continuous value that enabled the generation of a landslide susceptibility belief map. In total, four landslide susceptibility maps were produced using the certainty factor method, the Iterative Dichotomizer version 3 algorithm, the J48 algorithm and the modified Iterative Dichotomizer version 3 model in order to evaluate the performance of the developed classifier. The validation results showed that area under the ROC curves for the models varied from 0.7936 to 0.8397 for success rate curve and 0.7766 to 0.8035 for prediction rate curves, respectively. The success rate and prediction curves showed that the modified Iterative Dichotomizer version 3 model had a slightly higher performance with 0.8397 and 0.8035, respectively. From the outcomes of the study, it was induced that the developed modified decision tree classifier could be efficiently used for landslide susceptibility analysis and in general might be used for classification and estimation purposes in spatial predictive models.     

Landslide susceptibility mapping using random forest,boosted regression tree,classification and regression tree,and general linear models and comparison of their performance at Wadi Tayyah Basin,Asir Region,Saudi Arabia

The purpose of the current study is to produce landslide susceptibility maps using different data mining models. Four modeling techniques, namely random forest (RF), boosted regression tree (BRT), classification and regression tree (CART), and general linear (GLM) are used, and their results are compared for landslides susceptibility mapping at the Wadi Tayyah Basin, Asir Region, Saudi Arabia. Landslide locations were identified and mapped from the interpretation of different data types, including high-resolution satellite images, topographic maps, historical records, and extensive field surveys. In total, 125 landslide locations were mapped using ArcGIS 10.2, and the locations were divided into two groups; training (70 %) and validating (25 %), respectively. Eleven layers of landslide-conditioning factors were prepared, including slope aspect, altitude, distance from faults, lithology, plan curvature, profile curvature, rainfall, distance from streams, distance from roads, slope angle, and land use. The relationships between the landslide-conditioning factors and the landslide inventory map were calculated using the mentioned 32 models (RF, BRT, CART, and generalized additive (GAM)). The models’ results were compared with landslide locations, which were not used during the models’ training. The receiver operating characteristics (ROC), including the area under the curve (AUC), was used to assess the accuracy of the models. The success (training data) and prediction (validation data) rate curves were calculated. The results showed that the AUC for success rates are 0.783 (78.3 %), 0.958 (95.8 %), 0.816 (81.6 %), and 0.821 (82.1 %) for RF, BRT, CART, and GLM models, respectively. The prediction rates are 0.812 (81.2 %), 0.856 (85.6 %), 0.862 (86.2 %), and 0.769 (76.9 %) for RF, BRT, CART, and GLM models, respectively. Subsequently, landslide susceptibility maps were divided into four classes, including low, moderate, high, and very high susceptibility. The results revealed that the RF, BRT, CART, and GLM models produced reasonable accuracy in landslide susceptibility mapping. The outcome maps would be useful for general planned development activities in the future, such as choosing new urban areas and infrastructural activities, as well as for environmental protection.     

基于自组织特征映射网络-随机森林模型的滑坡易发性评价

为深入探讨评价单元和非滑坡样本选取对滑坡易发性预测的影响,构建了一种基于自组织特征映射网络-随机森林模型的滑坡易发性评价模型。该模型针对栅格单元和斜坡单元在滑坡易发性评价中的不足,结合栅格单元和斜坡单元的相互关系,提出了滑坡易发性指数的优化计算方法。在此基础上,基于随机森林Tree Bagger分类器构建滑坡易发性评价模型,通过对比分析自组织特征映射网络和随机方法选取非滑坡样本对评价结果的影响,探讨自组织特征映射网络、随机森林和自组织特征映射网络-随机森林三种评价模型的有效性;将评价模型应用于大余县滑坡易发性评价。结果显示,随机森林模型和自组织特征映射网络-随机森林模型的预测精度较高,分别达到91.19%和94.94%,成功率曲线的AUC值分别为0.822和0.849,表明自组织特征映射网络-随机森林模型具有更高的预测率和成功率, 自组织特征映射网络聚类的预测精度虽然有限,但作为非滑坡样本的选择方法,能够有效提高随机森林模型的评价精度。     

How do machine learning techniques help in increasing accuracy of landslide susceptibility maps?

Landslides are abundant in mountainous regions.They are responsible for substantial damages and losses in those areas.The A1 Highway,which is an important road in Algeria,was sometimes constructed in mountainous and/or semi-mountainous areas.Previous studies of landslide susceptibility mapping conducted near this road using statistical and expert methods have yielded ordinary results.In this research,we are interested in how do machine learning techniques help in increasing accuracy of landslide susceptibility maps in the vicinity of the A1 Highway corridor.To do this,an important section at Ain Bouziane(NE,Algeria) is chosen as a case study to evaluate the landslide susceptibility using three different machine learning methods,namely,random forest(RF),support vector machine(SVM),and boosted regression tree(BRT).First,an inventory map and nine input factors were prepared for landslide susceptibility mapping(LSM) analyses.The three models were constructed to find the most susceptible areas to this phenomenon.The results were assessed by calculating the receiver operating characteristic(ROC) curve,the standard error(Std.error),and the confidence interval(CI) at 95%.The RF model reached the highest predictive accuracy(AUC=97.2%) comparatively to the other models.The outcomes of this research proved that the obtained machine learning models had the ability to predict future landslide locations in this important road section.In addition,their application gives an improvement of the accuracy of LSMs near the road corridor.The machine learning models may become an important prediction tool that will identify landslide alleviation actions.     

Landslide Susceptibility Mapping Using GIS-based Fuzzy Logic and the Analytical Hierarchical Processes Approach: A Case Study in Constantine (North-East Algeria)

The main purpose of this study was to compare and evaluate the performance of two multicriteria models for landslide susceptibility assessment in Constantine, north-east of Algeria. The landslide susceptibility maps were produced using the analytic hierarchy process (AHP) and Fuzzy AHP (FAHP) via twelve landslides conditioning factors, including the slope gradient, lithology, land cover, distance from drainage network, distance from the roads, distance from faults, topographic wetness index, stream power index, slope curvature, Normalized Difference Vegetation Index, slope aspect and elevation. In this study, the mentioned models were used to derive the weighting value of the conditioning factors. For the validation process of these models, the receiver operating characteristic analysis, and the area under the curve (AUC) were applied by comparing the obtained results to The landslide inventory map which prepared using the archives of scientific publications, reports of local authorities, and field survey as well as analyzing satellite imagery. According to the AUC values, the FAHP model had the highest value (0.908) followed by the AHP model (0.777). As a result, the FAHP model is more consistent and accurate than the AHP in this case study. The outcome of this paper may be useful for landslide susceptibility assessment and land use management.

     

Comparison of machine-learning techniques for landslide susceptibility mapping using two-level random sampling (2LRS) in Alakir catchment area,Antalya, Turkey

The aim of this study is to make a comparison of the performances of two machine-learning algorithms that support vector machine (SVM) and random forest (RF) for landslide susceptibility mapping. The study makes use of a sampling strategy called two-level random sampling (2LRS). During landslide susceptibility mapping, training and testing samples must be collected from different landslide seed cells, which are then put through a fully independent sampling using the 2LRS algorithm. This approach requires fewer samples for the improvement of the computation time of both machine-learning classifications. The proposed approach was tested in the Alakir catchment area (Western Antalya, Turkey) which features numerous active deep-seated rotational landslides. In order to compare the performance of the machine-learning algorithms, three random sets were generated for SVM and three random sets generated for 10, 100, 1000 and 10,000-tree size RF. A total of 15 models were generated for comparison, and their spatial performances were performed by the area under the receiver-operating characteristic curves, which ranged between 0.82 and 0.87. The highest and lowest performances were recorded from two models in SVM and two models from the 1000-tree and 10,000-tree sized RF, respectively. These results were confirmed the landslide happened just after producing the susceptibility maps in the field.     

Landslide susceptibility mapping by binary logistic regression, analytical hierarchy process, and statistical index models and assessment of their performances

The current research presents a detailed landslide susceptibility mapping study by binary logistic regression, analytical hierarchy process, and statistical index models and an assessment of their performances. The study area covers the north of Tehran metropolitan, Iran. When conducting the study, in the first stage, a landslide inventory map with a total of 528 landslide locations was compiled from various sources such as aerial photographs, satellite images, and field surveys. Then, the landslide inventory was randomly split into a testing dataset 70 % (370 landslide locations) for training the models, and the remaining 30 % (158 landslides locations) was used for validation purpose. Twelve landslide conditioning factors such as slope degree, slope aspect, altitude, plan curvature, normalized difference vegetation index, land use, lithology, distance from rivers, distance from roads, distance from faults, stream power index, and slope-length were considered during the present study. Subsequently, landslide susceptibility maps were produced using binary logistic regression (BLR), analytical hierarchy process (AHP), and statistical index (SI) models in ArcGIS. The validation dataset, which was not used in the modeling process, was considered to validate the landslide susceptibility maps using the receiver operating characteristic curves and frequency ratio plot. The validation results showed that the area under the curve (AUC) for three mentioned models vary from 0.7570 to 0.8520 $ ({\text{AUC}}_{\text{AHP}} = 75.70\;\% ,\;{\text{AUC}}_{\text{SI}} = 80.37\;\% ,\;{\text{and}}\;{\text{AUC}}_{\text{BLR}} = 85.20\;\% ) $ ( AUC AHP = 75.70 % , AUC SI = 80.37 % , and AUC BLR = 85.20 % ) . Also, plot of the frequency ratio for the four landslide susceptibility classes of the three landslide susceptibility models was validated our results. Hence, it is concluded that the binary logistic regression model employed in this study showed reasonably good accuracy in predicting the landslide susceptibility of study area. Meanwhile, the results obtained in this study also showed that the statistical index model can be used as a simple tool in the assessment of landslide susceptibility when a sufficient number of data are obtained.     

Bagging based Support Vector Machines for spatial prediction of landslides

A hybrid Bagging based Support Vector Machines (BSVM) method, which is a combination of Bagging Ensemble and Support Vector Machine (SVM) classifier, was proposed for the spatial prediction of landslides at the district of Mu Cang Chai, Viet Nam. In the present study, 248 past landslides and fifteen geo-environmental factors (curvature, elevation, distance to rivers, slope, aspect, river density, plan curvature, distance to faults, profile curvature, fault density, lithology, distance to roads, rainfall, land use, and road density) were considered for the model construction. Different evaluation criteria were applied to validate the proposed hybrid model such as statistical index-based methods and area under the receiver operating characteristic curve (AUC). The single SVM and the Naïve Bayes Trees (NBT) models were selected for comparison. Based on the AUC values, the proposed hybrid model BSVM (0.812) outperformed the SVM (0.804) and NBT (0.8) models. Thus, the BSVM is a promising and better method for landslide prediction.     

滑坡易发性预测不确定性:环境因子不同属性区间划分和不同数据驱动模型的影响

对于滑坡易发性预测建模,连续型环境因子在频率比分析时的属性区间划分数量（attribute interval numbers,AIN）和不同易发性预测模型是两个重要不确定性因素.为研究这两个因素对建模的影响规律,以江西省上犹县为例,考虑5种连续型环境因子AIN划分（4、8、12、16及20）和5种数据驱动模型（层次分析法（analytic hierarchy process,AHP）、逻辑回归（logistic regression,LR）、BP神经网络（back-propagation neural network,BPNN）、支持向量机（support vector machine,SVM）和随机森林（random forest,RF））,总计25种不同工况下的滑坡易发性预测研究.再开展滑坡易发性指数的不确定性（包括精度评价和统计规律等）分析.结果表明:（1）对于同一模型,随着AIN值从4增加至8再到20时,易发性预测精度先逐渐提升,然后缓慢提升直至稳定;（2）对于同一AIN值,RF模型预测精度最高,其后依次为SVM、BPNN、LR和AHP模型;（3）在25种组合工况下,AIN=20和RF模型的预测精度最高,AIN=4和AHP模型精度最低,但在AIN=8和RF模型组合下的易发性建模效率较高且精度也较高;（4）更大的AIN值和更先进的机器学习模型预测出的滑坡易发性指数的不确定性相对较低,更符合实际的滑坡概率分布特征.在环境因子属性区间划分为8和RF模型工况下高效准确地构建滑坡易发性预测模型.      

Mapping landslide susceptibility with frequency ratio,statistical index,and weights of evidence models: a case study in northern Iran

The main objective of this study is to investigate potential application of frequency ratio (FR), weights of evidence (WoE), and statistical index (SI) models for landslide susceptibility mapping in a part of Mazandaran Province, Iran. First, a landslide inventory map was constructed from various sources. The landslide inventory map was then randomly divided in a ratio of 70/30 for training and validation of the models, respectively. Second, 13 landslide conditioning factors including slope degree, slope aspect, altitude, plan curvature, stream power index, topographic wetness index, sediment transport index, topographic roughness index, lithology, distance from streams, faults, roads, and land use type were prepared, and the relationships between these factors and the landslide inventory map were extracted by using the mentioned models. Subsequently, the multi-class weighted factors were used to generate landslide susceptibility maps. Finally, the susceptibility maps were verified and compared using several methods including receiver operating characteristic curve with the areas under the curve (AUC), landslide density, and spatially agreed area analyses. The success rate curve showed that the AUC for FR, WoE, and SI models was 81.51, 79.43, and 81.27, respectively. The prediction rate curve demonstrated that the AUC achieved by the three models was 80.44, 77.94, and 79.55, respectively. Although the sensitivity analysis using the FR model revealed that the modeling process was sensitive to input factors, the accuracy results suggest that the three models used in this study can be effective approaches for landslide susceptibility mapping in Mazandaran Province, and the resultant susceptibility maps are trustworthy for hazard mitigation strategies.     

The assessment of landslide susceptibility mapping using random forest and decision tree methods in the Three Gorges Reservoir area,China

Landslide susceptibility mapping is an indispensable prerequisite for landslide prevention and reduction. At present, research into landslide susceptibility mapping has begun to combine machine learning with remote sensing and geographic information system (GIS) techniques. The random forest model is a new integrated classification method, but its application to landslide susceptibility mapping remains limited. Landslides represent a serious threat to the lives and property of people living in the Zigui–Badong area in the Three Gorges region of China, as well as to the operation of the Three Gorges Reservoir. However, the geological structure of this region is complex, involving steep mountains and deep valleys. The purpose of the current study is to produce a landslide susceptibility map of the Zigui–Badong area using a random forest model, multisource data, GIS, and remote sensing data. In total, 300 pre-existing landslide locations were obtained from a landslide inventory map. These landslides were identified using visual interpretation of high-resolution remote sensing images, topographic and geologic data, and extensive field surveys. The occurrence of landslides is closely related to a series of environmental parameters. Topographic, geologic, Landsat-8 image, raining data, and seismic data were used as the primary data sources to extract the geo-environmental factors influencing landslides. Thirty-four layers of causative factors were prepared as predictor variables, which can mainly be categorized as topographic, geological, hydrological, land cover, and environmental trigger parameters. The random forest method is an ensemble classification technique that extends diversity among the classification trees by resampling the data with replacement and randomly changing the predictive variable sets during the different tree induction processes. A random forest model was adopted to calculate the quantitative relationships between the landslide-conditioning factors and the landslide inventory map and then generate a landslide susceptibility map. The analytical results were compared with known landslide locations in terms of area under the receiver operating characteristic curve. The random forest model has an area ratio of 86.10%. In contrast to the random forest (whole factors, WF), random forest (12 major factors, 12F), decision tree (WF), decision tree (12F), the final result shows that random forest (12F) has a higher prediction accuracy. Meanwhile, the random forest models have higher prediction accuracy than the decision tree model. Subsequently, the landslide susceptibility map was classified into five classes (very low, low, moderate, high, and very high). The results demonstrate that the random forest model achieved a reasonable accuracy in landslide susceptibility mapping. The landslide hazard zone information will be useful for general development planning and landslide risk management.     

Landslide susceptibility mapping using analytical hierarchy process (AHP) in Tehri reservoir rim region,Uttarakhand

A comprehensive use of analytical hierarchy process (AHP) method in landslide susceptibility mapping (LSM) has been presented for rim region of Tehri reservoir. Using remote sensing data, various landslide causative factors responsible for inducing instability in the area were derived. Ancillary data such as geological map, soil map, and topographic map were also considered along with remote sensing data. Exhaustive field checks were performed to define the credibility of the random landslide conditioning factors considered in this study. Apart from universally acceptable inherent causative factors used in the susceptibility mapping, others such as impact of reservoir impoundment on terrain, topographic wetness index and stream power index were found to be important causative factors in rim region of the Tehri reservoir. The AHP method was used to acquire weights of factors and their classes respectively. Weights achieved from AHP method matched with the existing field conditions. Acceptable consistency ratio (CR) value was achieved for each AHP matrix. Weights of each factor were integrated with weighted sum technique and a landslide susceptibility index map was generated. Jenk’s natural break classifier was used to classify LSI map into very low, low, moderate, high and very high landslide susceptible classes. Validation of the susceptibility map was performed using cumulative percentage/success rate curve technique. Area under curve value of the success rate curve was converted to percentage validation accuracy and a reasonable 78.7% validation accuracy was achieved.     

不同环境因子联接和预测模型的滑坡易发性建模不确定性

拟深入探讨滑坡与其环境因子间的非线性联接计算以及不同数据驱动模型等因素,对滑坡易发性预测建模不确定性的影响规律.以江西省瑞金市为例共获取370处滑坡和10种环境因子,通过概率统计（probability statistics,PS）、频率比（frequency ratio,FR）、信息量（information value,Ⅳ）、熵指数（index of entropy,IOE）和证据权（weight of evidence,WOE）等5种联接方法分别耦合逻辑回归（logistic regression,LR）、BP神经网络（BP neural networks,BPNN）、支持向量机（support vector machines,SVM）和随机森林（random forest,RF）模型共构建出20种耦合模型,同时构建无联接方法直接将原始数据作为输入变量的4种单独LR、BPNN、SVM和RF模型,预测出总计24种工况下的滑坡易发性;最后分别使用ROC曲线、均值、标准差和差异显著性等指标分析上述24种工况下易发性结果的不确定性.结果表明:（1）基于WOE的耦合模型预测滑坡易发性的平均精度最高且不确定性较低,基于PS的耦合模型预测精度最低且不确定性最高,基于FR、Ⅳ和IOE的耦合模型介于两者之间;（2）单独数据驱动模型易发性预测精度略低于耦合模型,且未能计算出环境因子各子区间对滑坡发育的影响规律,但其建模效率高于耦合模型;（3）RF模型预测精度最高且不确定性较低,其次分别为SVM、BPNN和LR模型.总之WOE是更优秀的联接法且RF模型预测性能最优,WOE-RF模型预测的滑坡易发性不确定性较低且更符合实际滑坡概率分布特征.