Method For Landslide Susceptibility Mapping Research Articles

Naïve and Semi-Naïve Bayesian Classification of Landslide Susceptibility Applied to the Kulekhani River Basin in Nepal as a Test Case

Naïve Bayes classification is widely used for landslide susceptibility analysis, especially in the form of weights-of-evidence. However, when significant conditional dependence is present, the probabilities derived from weights-of-evidence are biased, resulting in an overestimation of landslide susceptibility. As a solution, this study presents a semi-naïve Bayesian method for landslide susceptibility mapping by combining logistic regression with weights-of-evidence. The utility of the method is tested by application to a case study in the Kulekhani River Basin in Central Nepal. The results show that the naïve Bayes approach with weights-of-evidence overpredicts the posterior probability of landslide occurrence by a factor of about two, while the semi-naïve Bayes approach, which uses logistic regression with weights-of-evidence, is unbiased and has more discriminatory power for landslide susceptibility mapping. In addition, the semi-naïve Bayes approach can statistically distinguish the main factors that promote landslides and allows us to estimate the model uncertainty by calculating the standard error of the predictions.

An innovative method for landslide susceptibility mapping supported by fractal theory, GeoDetector, and random forest: a case study in Sichuan Province, SW China

An innovative method for landslide susceptibility mapping supported by fractal theory, GeoDetector, and random forest: a case study in Sichuan Province, SW China

Application of Bagging, Boosting and Stacking Ensemble and EasyEnsemble Methods for Landslide Susceptibility Mapping in the Three Gorges Reservoir Area of China

Since the impoundment of the Three Gorges Reservoir area in 2003, the potential risks of geological disasters in the reservoir area have increased significantly, among which the hidden dangers of landslides are particularly prominent. To reduce casualties and damage, efficient and precise landslide susceptibility evaluation methods are important. Multiple ensemble models have been used to evaluate the susceptibility of the upper part of Badong County to landslides. In this study, EasyEnsemble technology was used to solve the imbalance between landslide and nonlandslide sample data. The extracted evaluation factors were input into three bagging, boosting, and stacking ensemble models for training, and landslide susceptibility mapping (LSM) was drawn. According to the importance analysis, the important factors affecting the occurrence of landslides are altitude, terrain surface texture (TST), distance to residences, distance to rivers and land use. The influences of different grid sizes on the susceptibility results were compared, and a larger grid was found to lead to the overfitting of the prediction results. Therefore, a 30 m grid was selected as the evaluation unit. The accuracy, area under the curve (AUC), recall rate, test set precision, and kappa coefficient of a multi-grained cascade forest (gcForest) model with the stacking method were 0.958, 0.991, 0.965, 0.946, and 0.91, respectively, which a significantly better than the values produced by the other models.

A robust discretization method of factor screening for landslide susceptibility mapping using convolution neural network, random forest, and logistic regression models

ABSTRACT The selection of discretization criteria and interval numbers of landslide-related environmental factors generally fails to quantitatively determine or filter, resulting in uncertainties and limitations in the performance of machine learning (ML) methods for landslide susceptibility mapping (LSM). The aim of this study is to propose a robust discretization criterion (RDC) to quantify and explore the uncertainty and subjectivity of different discretization methods. The RDC consists of two steps: raw classification dataset generation and optimal dataset extraction. To evaluate the robustness of the proposed RDC method, Lushan County of Sichuan Province in China was chosen as the study area to generate the LSM based on three datasets (optimal dataset, original dataset with continuous values, and statistical dataset) using three popular ML methods, namely, convolution neural network, random forest, and logistic regression. The results show that the areas under the receiver operating characteristic curve (AUCs) of the optimal dataset for the abovementioned ML models are 0.963, 0.961, and 0.930 which are higher than those of the original dataset (0.938, 0.947, and 0.900) and statistical dataset (0.948, 0.954, and 0.897). In conclusion, the RDC method can extract the more representative features from environmental factors and outperform the other conventional discretization methods.

An objective absence data sampling method for landslide susceptibility mapping

The accuracy and quality of the landslide susceptibility map depend on the available landslide locations and the sampling strategy for absence data (non-landslide locations). In this study, we propose an objective method to determine the critical value for sampling absence data based on Mahalanobis distances (MD). We demonstrate this method on landslide susceptibility mapping of three subdistricts (Upazilas) of the Rangamati district, Bangladesh, and compare the results with the landslide susceptibility map produced based on the slope-based absence data sampling method. Using the 15 landslide causal factors, including slope, aspect, and plan curvature, we first determine the critical value of 23.69 based on the Chi-square distribution with 14 degrees of freedom. This critical value was then used to determine the sampling space for 261 random absence data. In comparison, we chose another set of the absence data based on a slope threshold of < 3°. The landslide susceptibility maps were then generated using the random forest model. The Receiver Operating Characteristic (ROC) curves and the Kappa index were used for accuracy assessment, while the Seed Cell Area Index (SCAI) was used for consistency assessment. The landslide susceptibility map produced using our proposed method has relatively high model fitting (0.87), prediction (0.85), and Kappa values (0.77). Even though the landslide susceptibility map produced by the slope-based sampling also has relatively high accuracy, the SCAI values suggest lower consistency. Furthermore, slope-based sampling is highly subjective; therefore, we recommend using MD -based absence data sampling for landslide susceptibility mapping.

Stacking ensemble of machine learning methods for landslide susceptibility mapping in Zhangjiajie City, Hunan Province, China

The current study aims to apply and compare the performance of six machine learning algorithms, including three basic classifiers: random forest (RF), gradient boosting decision tree (GBDT), and extreme gradient boosting (XGB), as well as their hybrid classifiers, using the logistic regression (LR) method (RF + LR, GBDT + LR, and XGB + LR), to map the landslide susceptibility of Zhangjiajie City, Hunan Province, China. First, a landslide inventory map was created with 206 historical landslide points and 412 non-landslide points, which was randomly divided into two datasets for model training (80%) and model testing (20%). Second, a landslide factor database was initially established by selecting 15 landslide conditioning factors from the topography, hydrology, climate, geology, and artificial activities. Thereafter, the multicollinearity test and information gain ratio (IGR) technique were applied to rank the importance of the factors. Subsequently, we used a series of metrics (e.g., accuracy, precision, recall, f-measure, area under the ROC (receiver operating characteristic) curve (AUC), kappa index, mean absolute error (MAE), and root mean square error (RMSE)) to evaluate the accuracy and performance of the six models. Based on the AUC values derived from the models, the GBDT + LR model with the highest AUC value (0.8168) was identified as the most efficient model for mapping landslide susceptibility, followed by the XGB + LR, XGB, RF + LR, GBDT, and RF models, which achieved AUC values of 0.8124, 0.8118, 0.8060, 0.7927, and 0.7883, respectively. The results from this study suggest that the stacking ensemble machine learning method is promising for use in landslide susceptibility mapping in the Zhangjiajie area and is capable of targeting the areas prone to landslides.

Application and validation of AHP and FR methods for landslide susceptibility mapping around choke mountain, northwestern ethiopia

The present research was conducted around Choke Mountain, East Gojjam, Northwestern Ethiopia. This study aimed to apply and validate analytical hierarchy process (AHP) and frequency ratio (FR) Methods and produce landslide susceptibility maps of the area. At first, a landslide inventory map was produced with information on past and active landslide locations. To produce a susceptibility map, nine influencing factors including lithology, slope, elevation, distance to drainage, distance to road, profile curvature, plane curvature, NDVI, and aspect were determined and integrated into the ArcGIS platform. All these influencing factors were statistically analyzed to determine their relationship to past landslides. The landslide susceptibility maps generated in the GIS environment by the AHP and FR methods were classified into five susceptibility classes: very low, low, moderate, high, and very high. Accordingly, the AHP method showed that very high susceptibility zones and high susceptibility zones cover 9.8% and 19.3% of the area, respectively. Whereas, moderate, low, and very low susceptibility zones cover 26.1%, 28.0%, and 16.8% of the area, respectively. Similarly, the FR method showed that very high susceptibility zones and high susceptibility zones cover 12.2% and 22.5% of the area, respectively. Whereas, moderate, low, and very low susceptibility zones cover 27.6%, 24.2%, and 13.4% of the area, respectively. The final results were compared and validated using the area under the curve (AUC) method. According to the AUC graph, the AHP and FR methods had success rates of 85.8 and 82.0%, respectively, and prediction rates of 88.2 and 84.8%. According to the AUC validation results, the map obtained from the AHP method is more accurate than that obtained from the FR method. Thus, the validation of the landslide susceptibility map demonstrated that the methodologies used resulted in acceptable results. This type of map can be useful for damage and risk prevention management and land use planning.

Using Landslide Statistical Index Technique for Landslide Susceptibility Mapping: Case Study: Ban Khoang Commune, Lao Cai Province, Vietnam

Ban Khoang is a mountainous commune in Sa Pa district located in the central part of Lao Cai province, Vietnam. Landslides occur frequently in this area and seriously affect the local living conditions. To help the local authority in developing a landslide disaster action plan, the statistical index method for landslide susceptibility mapping is applied. As the result, the landslide susceptibility zonation (LSZ) map was created. The LSZ map indicates that areas of low, moderate, high and very high landslide susceptibility zones are, respectively, 20.3 km2, 12.4 km2, 15.4 km2, and 5.2 km2; most of the observed landslide areas that are well predicted belong to high or very high landslide susceptibility classes. In detail, 80% observed landslide areas and 78.57% number of observed landslides were well predicted, and the area (AUC) under the receiver operating characteristic (ROC) curve obtained 80.3%. Hence, the high and very high landslide susceptibility classes in the LSZ map can be considered highly believable, and the LSZ map will be reliable to use in the practice.

Spatial mapping of landslide susceptibility in Jerash governorate of Jordan using genetic algorithm-based wrapper feature selection and bagging-based ensemble model

This study employs five genetic algorithm (GA)-based machine learning (ML) models, namely the Decision Tree (DT), k-Nearest Neighbors (kNN), NaïveBayes (NB), Support Vector Machine (SVM), and Extreme Learning Machine (ELM), to build a novel ensemble algorithm that is founded on the Bagging method for landslide susceptibility mapping (LSM) in Jerash Governorate, north of Jordan. The GA-based wrapper feature selection (FS) was done based on the five individual models and in the initial stages of modeling, an inquiry for the best feature for each of the five models was made. Finally, five hybrid models, namely DT-GA, kNN-GA, NB-GA, SVM-GA, and ELM-GA were constructed and combined to create Bagging-based ensemble model. The FS outcomes uncovered that rainfall depth, distance to roads, the Stream Power Index, the Normalized Difference Vegetation Index, slope, geology, and aspect are the most influential determinants of landslides. After the significant variables were identified, they were selected as input predictors and entered into the models. GA-based Bagging ensemble model with the area under the receiver operating characteristic curve (AUROC) of 0.85 achieved the highest accuracy in the validation run, followed by SVM-GA (AUROC = 0.80), NB-GA (AUROC = 0.76), DT-GA (AUROC = 0.72), kNN-GA (AUROC = 0.70), and ELM-GA (AUROC = 0.48).

LAND-SUITE V1.0: a suite of tools for statistically based landslide susceptibility zonation

Abstract. In the past 50 years, a large variety of statistically based models and methods for landslide susceptibility mapping and zonation have been proposed in the literature. The methods, which are applicable to a large range of spatial scales, use a large variety of input thematic data, different model combinations, and several approaches to evaluate the models' performance. Despite the numerous applications available in the literature, a standard approach for susceptibility modeling and zonation is still missing. The literature search revealed that several software program and tools are available to evaluate regional slope stability using physically based analysis, but only a few use statistically based approaches. Among them, LAND-SE (LANDslide Susceptibility Evaluation) provides the possibility to perform and combine different statistical susceptibility models and to evaluate their performances and associated uncertainties. This paper describes the structure and the functionalities of LAND-SUITE, a suite of tools for statistically based landslide susceptibility modeling which integrates LAND-SE. LAND-SUITE completes and extends LAND-SE, adding functionalities to (i) facilitate input data preparation, (ii) perform preliminary and exploratory analysis of the available data, and (iii) test different combinations of variables and select the optimal thematic/explanatory set. LAND-SUITE provides a tool to assist the user during the data preparatory phase and to perform diversified statistically based landslide susceptibility applications.

Landslide Susceptibility Mapping along the Anninghe Fault Zone in China using SVM and ACO-PSO-SVM Models

Abstract In the present study, a hybrid machine learning model was designed by integrating ant colony optimization (ACO), particle swarm optimization (PSO), and support vector machine (SVM) algorithms. The model was used to map the landslide susceptibility of the Anninghe fault zone in Sichuan Province, China. Based on this, 12 conditioning factors associated with landslides were considered, namely, altitude, slope angle, cutting depth, slope aspect, relief amplitude, stream power index (SPI), gully density, lithology, rainfall, road density, distance to fault, and peak ground acceleration (PGA). The overall performance of the two resulting models was tested using the receiver operating characteristic (ROC), area under the ROC curve (AUC), Cohen’s kappa coefficient, and five statistical evaluation measures. The success rates of the ACO-PSO-SVM model and the SVM model were 0.898 and 0.814, respectively, while the prediction rates of the two models were 0.887 and 0.804, respectively. The results show that the ACO-PSO-SVM model yields better overall performance and accurate results than the SVM model. Therefore, in conclusion, the ACO-PSO-SVM model can be applied as a new promising method for landslide susceptibility mapping in subsequent studies. The results of this study will be useful for land-use planning, hazard prevention, and risk management.

GIS-Based AHP and FR Methods for Landslide Susceptibility Mapping in the Abay Gorge, Dejen–Renaissance Bridge, Central, Ethiopia

GIS-Based AHP and FR Methods for Landslide Susceptibility Mapping in the Abay Gorge, Dejen–Renaissance Bridge, Central, Ethiopia

A comparative study of different neural network models for landslide susceptibility mapping

Landslide susceptibility mapping (LSM) can be used to determine the spatial probability of landslide occurrence. There are many methods for LSM, including statistical methods, traditional machine learning methods and deep learning methods, etc. However, the difference comparison of these methods has been not perfect, especially the comparison of different neural network models for LSM and their application prospects were rarely studied. In this paper, the classical neural net-work multi-layer perceptron (MLP), convolutional neural network (CNN), gated recurrent unit (GRU) and multi-scale convolutional neural network (MSCNN) four models are selected for comparison. Taking Lanzhou city, Gansu Province, China as an example, eight landslide-related influencing factors and historical landslide and non-landslide locations were selected, and the training set and validation set were divided according to 7:3. Through training the four models, four landslide susceptibility maps were generated. The experimental results were verified and compared by the confusion matrix, Kappa coefficient, F1-score and other statistical indicators. The receiver operating characteristic (ROC) curve and Precision-Recall (PR) curve were plotted to evaluate the classification effect and generalization capability of four models. The results show that the constructed MSCNN is the optimal model, which has the best performance both in the training process and in the mapping results. MSCNN model has the highest value of Recall (99.93%), Kappa (0.96) and F1-score (0.98) in the confusion matrix. In addition, ROC curve and PR curve of MSCNN model maintain the maximum area under curve (AUC) on different data sets. In the comparison, MLP and GRU accept sequence features, while CNN and MSCNN accept neighborhood features. In general, the prediction model considering neighborhood features contains more information in the limited input data and is better than the prediction model considering sequence features in all evaluation indicators. Therefore, we think that the neighborhood features can better represent the landslide occurrence characteristics. In the future model design process for LSM, more attention should be paid to the neighborhood features of landslide influencing factors.

A comparative analysis of weight-based machine learning methods for landslide susceptibility mapping in Ha Giang area

ABSTRACT Landslide susceptibility maps (LSMs) are very crucial for planning policies in hazardous areas. However, the accuracy and reliability of LSMs depend on available data and the selection of suitable methods. This study is conducted to produce LSMs by combinations of machine learning methods and weighting techniques for Ha Giang province, Vietnam, where has limited data. In study area, we gather 11 landslide conditioning factors and establish a landslide inventory map. Computing the weights of classes (or factors) is very important to prepare data for machine learning methods to generate LSMs. We first use frequency ratio (FR) and analytic hierarchy process (AHP) techniques to generate the weights. Then, random forest (RF), support vector machine (SVM), logistic regression (LR), and AHP methods are combined with FR and AHP weights to yield accurate and reliable LSMs. Finally, the performance of these methods is evaluated by five statistical metrics, ROC and R-index. The empirical results have shown that RF is the best method in terms of R-index and the five metrics, i.e. TP rate (0.9661), FP rate (0.0), ACC (0.9835), MAE (0.0046), and RMSE (0.0350) for this study area. This study opens the perspective of weight-based machine learning methods for landslide susceptibility mapping.

Integrating Data Modality and Statistical Learning Methods for Earthquake-Induced Landslide Susceptibility Mapping

Earthquakes induce landslides worldwide every year that may cause massive fatalities and financial losses. Precise and timely landslide susceptibility mapping (LSM) is significant for landslide hazard assessment and mitigation in earthquake-affected areas. State-of-the-art LSM approaches connect causative factors from various sources without considering the fusion of different information at the data modal level. To exploit the complementary information of different modalities and boost LSM accuracy, this study presents a new LSM model that integrates data modality and machine learning methods. The presented method first groups causative factors into different modal types based on their intrinsic characteristics, followed by the calculation of the pairwise similarity of modal data. The similarities of different modalities are fused using nonlinear graph fusion to generate a unified graph, which is subsequently classified using different machine learning methods to produce final LSM. Experimental results suggest that the presented method achieves higher performance than existing LSM methods. This study provides a new solution for producing precise LSM from a fusion perspective that can be applied to minimize the potential landslide risk and for sustainable use of erosion-prone slopes.

Coupling Data- and Knowledge-Driven Methods for Landslide Susceptibility Mapping in Human-Modified Environments: A Case Study from Wanzhou County, Three Gorges Reservoir Area, China

Landslide susceptibility mapping (LSM) can provide valuable information for local governments in landslide prevention and mitigation. Despite significant improvements in the predictive performance of LSM, it remains a challenge to be carried out in areas with limited availability of data. For example, in the early stage of road construction, landslide inventory data can be particularly scarce, while there is a high need to have a susceptibility map. This study aims to set up a novel procedure for coupling the knowledge-driven and data-driven models for LSM in an area with limited landslide inventory data. In particular, we propose a two-step approach. The first step consists of applying four data-driven models (logistic regression, decision tree, support vector machines, and random forest (RF)) to derive a regional susceptibility map. In the second step, the application of a heuristic model (analytic hierarchy process, AHP) is proposed to calculate a local susceptibility map for the areas with incomplete landslide inventories. The final landslide susceptibility map is obtained by merging the most accurate regional map (RF) with the local map. We apply this novel procedure to a landslide-prone region with developed road construction (National Highway G69) in Wanzhou district, where landslide inventory is difficult to update due to timely recovery from landslide-induced road damage. Results show that the proposed methodology allows identifying new landslide-prone areas, and improving LSM predictive performance, as demonstrated by the fact that two new landslides developed along G69 were perfectly classified in the highly susceptible areas. The results show that implementing the landslide susceptibility assessment with different geographical settings and combining them into best-sensitivity partitions is more accurate than focusing on creating new models or hybrid models.

Comparative Study of Convolutional Neural Network and Conventional Machine Learning Methods for Landslide Susceptibility Mapping

Landslide susceptibility mapping (LSM) is a useful tool to estimate the probability of landslide occurrence, providing a scientific basis for natural hazards prevention, land use planning, and economic development in landslide-prone areas. To date, a large number of machine learning methods have been applied to LSM, and recently the advanced convolutional neural network (CNN) has been gradually adopted to enhance the prediction accuracy of LSM. The objective of this study is to introduce a CNN-based model in LSM and systematically compare its overall performance with the conventional machine learning models of random forest, logistic regression, and support vector machine. Herein, we selected Zhangzha Town in Sichuan Province, China, and Lantau Island in Hong Kong, China, as the study areas. Each landslide inventory and corresponding predisposing factors were stacked to form spatial datasets for LSM. The receiver operating characteristic analysis, area under the curve (AUC), and several statistical metrics, such as accuracy, root mean square error, Kappa coefficient, sensitivity, and specificity, were used to evaluate the performance of the models. Finally, the trained models were calculated, and the landslide susceptibility zones were mapped. Results suggest that both CNN and conventional machine learning-based models have a satisfactory performance. The CNN-based model exhibits an excellent prediction capability and achieves the highest performance but also significantly reduces the salt-of-pepper effect, which indicates its great potential for application to LSM.

Assessment on Recent Landslide Susceptibility Mapping Methods: A Review

Landslide is a destructive natural hazard that causes severe property loss and loss of lives. Numerous researchers have developed landslide susceptibility maps in order to forecast its occurrence, particularly in hill-site development. Various quantitative approaches are used in landslide susceptibility map production, which can be classified into three categories; statistical data mining, machine learning and deterministic approach. In this paper, we choose two regular models in each category, which are Weight of Evidence (WoE) and Frequency Ratio (FR), Artificial Neutral Networks (ANN) and Support Vector Machines (SVM), Shallow Landsliding Stability Model (SHALSTAB) and YonSei-Slope (YS-Slope). Discussion and assessment on these models are based on relevant literature.

An Assessment of Data Mining and Bivariate Statistical Methods for Landslide Susceptibility Mapping

Landslides are recognized as one of the environmental challenges that lead to land degradation, reduce fertility, and cause significant damage to the ecosystem. Therefore, proper identification of landslide-prone areas through modeling will be significantly helpful for land development managers and planners by providing them with appropriate management strategies to prevent land degradation. In this research, landslide susceptibility mapping was carried out for West Azerbaijan province in Iran using Frequency Ratio (FR), Shannon Entropy (SE), Random Forest (RF), and an ensemble of random forest and bagging (RF-BA) methods. Based on field surveys, local interviews, and review of similar studies, 12 factors influencing landslide occurrence, namely altitude, slope angle, slope aspect, distance from fault, distance from river, distance from road, drainage density, road density, rainfall, soil, land use, and lithology, were identified. In the field surveys, 110 landslides in the area were specified; 70% of the data (77 landslides) were randomly selected and utilized for modeling and the remaining 30% (33 landslides) for validation. The results of the ROC curve showed the accuracy of 0.92, 0.91, 0.89, and 0.88 with the RF-BA, RF, FR, and SE models, respectively.

Weight of Evidence Method for Landslide Susceptibility Mapping in Sigi Biromaru, Central Sulawesi

Sigi Biromaru is an area prone to landslides. This study aims to apply the statistical method of Weight of Evidence (WoE) in landslide susceptibility mapping using Geographic Information Systems (GIS). The 265 landslides that occurred 2009-2019 were randomly divided into two groups, 70% of the data were used as training dataset for susceptibility modelling and 30% of the data were used as test data for validation of the susceptibility model. Twenty-one parameters were tested for their influence on landslides. Based on the Area Under Curve (AUC), parameters that significant controlling the landslides are slope gradient, elevation, aspect, flow direction, peak ground acceleration, clay content (<0,002 mm), land cover, terrain ruggedness index (TRI), river density, soil type, lineament density, lithology, rainfall and stream power index (SPI) respectively. The validation results show that the AUC success rate is 0,811 using the training dataset and AUC prediction rate is 0,756 using the test dataset. These results indicate that the WoE method produces a good landslide susceptibility map in the Sigi Biromaru area.