Model For Landslide Susceptibility Mapping Research Articles

A comparative study of the bivariate, multivariate and machine-learning-based statistical models for landslide susceptibility mapping in a seismic-prone region in China

Statistical landslide susceptibility mapping (LSM) models have been most widely used in literatures. However, limitations and uncertainties remain in these methods. The main goal of the current study was to test and compare the efficiency of a bivariate model (the weight of evidence (WoE)), a multivariate model (logistic regression (LR)) and a machine-learning algorithm (the support vector machine (SVM)) in LSM. Lushan County of China was chosen because of its mountainous terrain and high risky of devastating seismic activities. An inventory of 867 landslides was utilized in this study, 70% of which were used to train these models, and the rest 30% were used to validate their accuracies. Ten factors of aspect, elevation, slope, curvature, peak ground acceleration (PGA), distance to the river (DtoR), lithology, topographic wetness index (TWI), stream power index (SPI) and percentage of tree cover (PTC) were used as input of the landslide susceptibility mapping (LSM) models. Accuracy evaluation based on the areas under the receiver operating characteristic curves (AUC) showed that the LR model gives the highest success rate (78.2%) and prediction rate (76.4%), the SVM has the second-highest success rate (75.9%) and the WoE had the second-highest prediction rate (75.6%). Comparison results suggested that the LR and the SVM are proper models for LSM of the study area. The obtained susceptibility maps would benefit regional land planning and seismic landslide hazard mitigation in the study area.

Machine learning in earthquake- and typhoon-triggered landslide susceptibility mapping and critical factor identification

Landslides are one of the most devastating natural hazards worldwide. Landslides are triggered by different forces, such as earthquakes and typhoons, and have different characteristics in terms of distribution, influential factors, and process. The objectives of this study are to develop susceptibility maps using machine learning for two different triggering forces (earthquake and typhoon) and identify the main predisposing factors in mountainous regions of Pakistan and Taiwan. To compare different machine learning models for landslide susceptibility mapping, landslide susceptibility maps were developed using traditional (logistic regression) and modern techniques (decision tree). Results show that the spatial pattern of susceptibility map from logistic regression is continuously distributed, whereas that from the decision tree is crisp and sharp. From both models, consistent results show that the most important critical factors are completely different for both the earthquake- and typhoon-triggered landslides. For rainfall-triggered landslides in Taiwan, the most important factor of landslide susceptibility is the distance to the rivers, whereas, for earthquake-triggered landslides in Pakistan, the most important one is geological formations. Moreover, landslide susceptibility maps show that earthquake-triggered landslides tend to occur at the Muzaffarabad Formation, whereas rainstorm-induced landslides aggregate in the slope toe along the river.

Mamdani fuzzy inference systems and artificial neural networks for landslide susceptibility mapping

Two Artificial Intelligence (AI) methods, Fuzzy Inference System (FIS) and Artificial Neural Network (ANN), are applied to Landslide Susceptibility Mapping (LSM), to compare complementary aspects of the potentials of the two methods and to extract physical relationships from data. An index is proposed in order to rank and filter the FIS rules, selecting a certain number of readable rules for further interpretation of the physical relationships among variables. The area of study is Rolante river basin, southern Brazil. Eleven attributes are generated from a Digital Elevation Model (DEM), and landslide scars from an extreme rainfall event are used. Average accuracy and area under Receiver Operating Characteristic curve (AUC) resulted, respectively, in 81.27% and 0.8886 for FIS, and 89.45% and 0.9409 for ANN. ANN provides a map with more amplitude of outputs and less area classified as high susceptibility. Among the 40 (10%) best-ranked FIS rules, 13 have high susceptibility output, while 27 have low; a cause is that low susceptibility areas are larger on the map. Slope is highly connected to susceptibility. Elevation, when high (plateau) or low (floodplain), inhibits high susceptibility. Six attributes show the same fuzzy set for the 18 best-ranked rules, meaning this fuzzy set is common on the map. Overall findings point out that ANN is best suited for LSM map generation, but, based on them, using FIS is important to help researchers understand more about AI models for LSM and about landslide phenomenon.

Landslide susceptibility mapping by attentional factorization machines considering feature interactions

Landslide susceptibility mapping (LSM) is a commonly used approach to reduce landslide risk. However, conventional LSM methods generally only consider the influence of each single conditioning factor on landslide occurrence or absence, which neglects the interactions of different conditioning factors and may lead to biased LSM results. Therefore, this study aims to use a new machine learning model—attentional factorization machines (AFM)—to explicitly consider the influence of feature interactions in LSM to improve and obtain more reliable LSM results. The Anhua County in China is chosen as the study area. The area under the receiver operating characteristic curve (AUC) and statistical indicators are used to evaluate the performance of LSM models. For comparison, the common LSM models such as the logistic regression (LR) and random forest (RF) models are also used to conduct the LSM. The results show that the performance of AFM is a little better than RF in the AUC metric, whereas the LR model has the worst performance. Compared with general LSM models, AFM considers feature interactions by introducing an attention mechanism to learn the weight of different feature combinations, which not only ensures the model interpretability but also improves the model performance.

Exploring novel hybrid soft computing models for landslide susceptibility mapping in Son La hydropower reservoir basin

In this study, two novel hybrid models namely Bagging-based Rough Set (BRS) and AdaBoost-based Rough Set (ABRS) were used to generate landslide susceptibility maps of Son La hydropower reservoir basin, Vietnam. In total, 186 past landslide events and twelve landslides affecting factors (slope degree, slope aspect, elevation, curvature, focal flow, river density, rainfall, aquifer, weathering crust, lithology, fault density and road density) were considered in the modeling study. The landslide data was split into training (70%) and testing (30%) for the model’s development and validation. One R feature selection method was used to select and prioritize the landslide affecting factors based on their importance in model prediction. Performance of the hybrid developed models was evaluated and also compared with single rough set (RS) and support vector machine (SVM) models using various standard statistical measures including area under the curve (AUC)-receiver operating characteristics (ROC) curve. The results show that the developed hybrid model BRS (AUC = 0.845) is the most accurate model in comparison to other models (ABRS, SVM and RS) in predicting landslide susceptibility. Therefore, the BRS model can be used as an effective tool in the development of an accurate landslide susceptibility map of the hilly area.

Landslide susceptibility mapping in three Upazilas of Rangamati hill district Bangladesh: application and comparison of GIS-based machine learning methods

This study evaluates and compares three machine learning models: K-Nearest Neighbour (KNN), Random Forest (RF) and Extreme Gradient Boosting (XGBoost) for landslide susceptibility mapping for part of areas in Rangamati District, Bangladesh. The performance of these methods has been assessed by employing statistical methods such as the area under the curve (AUC) for success rate (SR) and prediction rate (PR), Kappa index, Qs index and Friedman's test. Results show that XGBoost had the best performance with the highest AUC for both SR (95.27%) and PR (90.63%), followed by RF (SR: 89.26%; PR: 84.74%) and KNN models (SR: 85.54%; PR: 81.02%). This study provides a useful analysis for the selection of the best model for landslide susceptibility mapping and that it will be helpful for disaster planning and risk reduction.

Application of a GIS-Based Slope Unit Method for Landslide Susceptibility Mapping in Helong City: Comparative Assessment of ICM, AHP, and RF Model

Landslides are one of the most extensive geological disasters in the world. The objective of this study was to assess the performances of different landslide susceptibility models information content method (ICM), analytical hierarchy process (AHP), and random forest (RF) model) and mapping unit (slope unit and grid unit) for landslide susceptibility mapping in the Helong city, Jilin province, northeastern China. First, a total of 159 landslides were mapped in the study area based on a geological hazard survey (1:50,000) of Helong city. Then, the slope units of the study area were divided by using the curvature watershed method. Next, eight influencing factors, namely, lithology, slope angle, slope aspect, rainfall, land use, seismic intensity, distance to river, and distance to fault, were selected to map the landslide susceptibility based on geological data, field survey, and landslide information. Afterward, landslide susceptibility modeling of landslide inventory data is performed for extracting and learning the symmetry latent in data patterns and relationships by three landslide susceptibility models and utilizing it to predict landslide susceptibility. Finally, the receiver operating characteristic (ROC) curve was used to compare the landslide susceptibility models. In addition, results based on grid units were calculated for comparison. The AUC (the area under the curve) result for ICM, AHP, and RF model was 87.1%, 80.5%, and 94.6% for slope units, and 83.4%, 70.9%, and 91.3% for grid units, respectively. Based on the overall assessments, the SU-RF model was the most suitable model for landslide susceptibility mapping. Consequently, these methods can be very useful for landslide hazard mitigation strategies.

Rainfall Induced Landslide Susceptibility Mapping Based on Bayesian Optimized Random Forest and Gradient Boosting Decision Tree Models—A Case Study of Shuicheng County, China

Among the most frequent and dangerous natural hazards, landslides often result in huge casualties and economic losses. Landslide susceptibility mapping (LSM) is an excellent approach for protecting and reducing the risks by landslides. This study aims to explore the performance of Bayesian optimization (BO) in the random forest (RF) and gradient boosting decision tree (GBDT) model for LSM and applied in Shuicheng County, China. Multiple data sources are used to obtain 17 conditioning factors of landslides, Borderline-SMOTE and Randomundersample methods are combined to solve the imbalanced sample problem. RF and GBDT models before and after BO are adopted to calculate the susceptibility value of landslides and produce LSMs and these models were compared and evaluated using multiple validation approach. The results demonstrated that the models we proposed all have high enough model accuracy to be applied to produce LSM, the performance of the RF is better than the GBDT model without BO, while after adopting the Bayesian optimized hyperparameters, the prediction accuracy of the RF and GBDT models is improved by 1% and 7%, respectively and the Bayesian optimized GBDT model is the best for LSM in this four models. In summary, the Bayesian optimized RF and GBDT models, especially the GBDT model we proposed for landslide susceptibility assessment and LSM construction has a very good application performance and development prospects.

Landslide susceptibility modeling based on ANFIS with teaching-learning-based optimization and Satin bowerbird optimizer

As threats of landslide hazards have become gradually more severe in recent decades, studies on landslide prevention and mitigation have attracted widespread attention in relevant domains. A hot research topic has been the ability to predict landslide susceptibility, which can be used to design schemes of land exploitation and urban development in mountainous areas. In this study, the teaching-learning-based optimization (TLBO) and satin bowerbird optimizer (SBO) algorithms were applied to optimize the adaptive neuro-fuzzy inference system (ANFIS) model for landslide susceptibility mapping. In the study area, 152 landslides were identified and randomly divided into two groups as training (70%) and validation (30%) dataset. Additionally, a total of fifteen landslide influencing factors were selected. The relative importance and weights of various influencing factors were determined using the step-wise weight assessment ratio analysis (SWARA) method. Finally, the comprehensive performance of the two models was validated and compared using various indexes, such as the root mean square error (RMSE), processing time, convergence, and area under receiver operating characteristic curves (AUROC). The results demonstrated that the AUROC values of the ANFIS, ANFIS-TLBO and ANFIS-SBO models with the training data were 0.808, 0.785 and 0.755, respectively. In terms of the validation dataset, the ANFIS-SBO model exhibited a higher AUROC value of 0.781, while the AUROC value of the ANFIS-TLBO and ANFIS models were 0.749 and 0.681, respectively. Moreover, the ANFIS-SBO model showed lower RMSE values for the validation dataset, indicating that the SBO algorithm had a better optimization capability. Meanwhile, the processing time and convergence of the ANFIS-SBO model were far superior to those of the ANFIS-TLBO model. Therefore, both the ensemble models proposed in this paper can generate adequate results, and the ANFIS-SBO model is recommended as the more suitable model for landslide susceptibility assessment in the study area considered due to its excellent accuracy and efficiency.

GIS-based ensemble soft computing models for landslide susceptibility mapping

Landslide susceptibility mapping has become one of the most important tools for the management of landslide hazards. In this study, we proposed a novel approach to improve the performance of Credal Decision Tree (CDT) by using four ensemble frameworks: Bagging, Dagging, Decorate, and Rotation Forest (RF) for landslide susceptibility mapping. A total number of 180 past and present landslides data of the Muong Lay district (Viet Nam) was analyzed and used for generating training and validation of the models. Several standard statistical performance evaluation metrics, such as negative predictive value, positive predictive value, root mean square error, accuracy, sensitivity, specificity, Kappa, Area Under the receiver operating Characteristic curve (AUC) were used to evaluate performance of the models. Results indicated that all the developed and applied models performed well (AUC: 0.842–0.886) but performance of the RF-CDT (AUC: 0.886) model is the best. Therefore, the RF-CDT ensemble model can be used for the correct landslide susceptibility mapping and for proper landslide management not only of the study area but also other hilly areas of the world.

Landslide susceptibility mapping for the Black Sea Region with spatial fuzzy multi-criteria decision analysis under semi-humid and humid terrestrial ecosystems

Landslide, which is a process experienced not only at the local and catchment scales but also the regional and national scales, adversely affects the natural environment, including the flora and fauna, and the socio-economic well-being of human communities. The main aim of the present study was to generate a landslide susceptibility mapping model for the semi-humid and humid terrestrial ecosystems of the Black Sea Region. The modelling was done by using a spatial multi-criterion analysis (SMCA) method based on the integration of fuzzy environment and geographical information system (GIS) techniques and also the use of the AHP approach by incorporating 9 environmental indicators and 27 sub-indicators. In order to allocate a weighting value for each indicator, the fuzzy-AHP approach was applied to determine efficiently sensitive levels of importance for the indicator. In addition, precipitation climatology was included in detail with respect to the temporal and spatial distribution of rainfall in the study area. It was determined that approximately 33% of the research area includes high or very high susceptibility to landslide, whereas about 37% of the research area is in the low or very low susceptibility classes. Most highly or very highly susceptibility areas were located in the western and central Black Sea regions. In conclusion, this study provides an alternative perspective a useful alternative method for landslide susceptibility mapping through the incorporation of fuzzy sets through AHP in conjunction with the Buckley approach in modelling.

Application of an Incomplete Landslide Inventory and One Class Classifier to Earthquake-Induced Landslide Susceptibility Mapping

Rapid and effective evaluation of landslide susceptibility after earthquakes is critical for various applications, such as emergency rescue, land planning, and disaster prevention. Current research suffers from the lack of a complete landslide inventory and sample selection uncertainty issues. To solve these problems, this study presents a landslide susceptibility mapping model that integrates one-class support vector machine (OCSVM) and an incomplete landslide inventory, which was established with the aid of change detection from bi-temporal Landsat images. Wenchuan County is selected as the study area to test the performance of the proposed method. The proposed method is also compared with standard two-class SVM that selects a sample randomly. Experimental results show that OCSVM can achieve better performance than SVM when only an incomplete landslide inventory is available. The findings of this study can be applied to determine regional landslide susceptibility after earthquakes and provide an essential reference for emergency response.

Machine learning and fractal theory models for landslide susceptibility mapping: Case study from the Jinsha River Basin

The quality of “non-landslide” negative samples may result in unreasonable prediction results for machine learning (ML) models. The aim of this study is to improve the performance of ML models by perfecting the quality of “non-landslide” samples in landslide susceptibility modelling so as to produce more reliable susceptibility maps. The middle and lower reaches of Jinsha River basin (MLRJB) were chosen as the study area, and the elevation, slope aspect, curvature, lithology, distance to faults, slope of slope, slope of aspect, precipitation, land use, and NDVI were considered as predisposing factors for landslide susceptibility mapping. Firstly, three “non-landslide” samples are randomly selected from the low-slope area, landslide-free area and very low susceptibility area based on fractal theory (FT) model generation, and then three sample scenarios are constructed with 4445 landslide positive samples. Next, the performance of cross-application of three sample scenarios in the support vector machines (SVM) and naïve Bayes (NB) models are compared and evaluated based on the statistical indicators such as accuracy, recall, precision, receiver operating characteristic (ROC) curve, and area under the ROC curve (AUC). The evaluation results show that the “non-landslide” negative samples generated on the basis of FT model are more reasonable and that the hybrid method supported by FT and ML models exhibits the highest prediction efficiency, around 94% overall accuracy produced by scenario-FT, followed by scenario-SS (87%) and scenario-RS (65%). Finally, with the validation of landslide and unstable slopes data, the landslide susceptibility map produced by the hybrid method composed of FT model and the SVM model is the ultimate output product for landslide prevention.

Landslide Susceptibility Mapping Combining Information Gain Ratio and Support Vector Machines: A Case Study from Wushan Segment in the Three Gorges Reservoir Area, China

Landslides are destructive geological hazards that occur all over the world. Due to the periodic regulation of reservoir water level, a large number of landslides occur in the Three Gorges Reservoir area (TGRA). The main objective of this study was to explore the preference of machine learning models for landslide susceptibility mapping in the TGRA. The Wushan segment of TGRA was selected as a case study. At first, 165 landslides were identified and a total of 14 landslide causal factors were constructed from different data sources. Multicollinearity analysis and information gain ratio (IGR) model were applied to select landslide causal factors. Subsequently, the landslide susceptibility mapping using the calculated results of four models, namely, support vector machines (SVM), artificial neural networks (ANN), classification and regression tree (CART), and logistic regression (LR). The accuracy of these four maps were evaluated using the receive operating characteristic (ROC) and the accuracy statistic. Results revealed that eliminating the inconsequential factors can perhaps improve the accuracy of landslide susceptibility modelling, and the SVM model had the best performance in this study, providing strong technical support for landslide susceptibility modelling in TGRA.

Combining Participatory Mapping, Cloud Computing and Machine Learning for Mapping Climate Induced Landslide Susceptibility in Lembeh Island, North Sulawesi

This study explored the use of participatory mapping and several machine learning algorithms (Naïve Bayes, GMO Maxent, SVM, CART, and Random Forest) to map climate induced landslide susceptibility in Lembeh Island, North Sulawesi, based on Earth Observation data available in Google Earth Engine. Participatory mapping on landslide incidents were conducted in three villages, i.e., Kareko, Pintu Kota, and Pasir Panjang. Data used include digital elevation model from SRTM, multispectral imageries from Sentinel 2, and precipitation from CHIRPS. Terrain modelling was done to DEM to come up with elevation, slope, curvature, and aspect. A cloud free mosaic of Sentinel Images was created using the median reducer and then NDVI was calculated. Precipitation data from CHIRPS was sampled and interpolated using kriging and reduced to maximum and mean. Each algorithm was trained using 70% participatory mapping data and then the prediction was tested for accuracy using the last 30%. Results showed that Random Forest, SVM, CART, and GMO Maxent gave 0.98 testing accuracy and Naïve Bayes only 0.90. The map showed that due to the terrain condition, Lembeh Island is prone to Landslide and even though previously BNPB already provide a landslide hazard risk map, there were many areas not included on that map. Therefore, the map could become an input for BNPB and the Bitung City for developing a mitigation and adaptation strategy. Machine learning and cloud computing along with participatory mapping could also complement mechanistic or multi-criteria analysis using GIS model for landslide susceptibility mapping.

Application of logistic regression (LR) and frequency ratio (FR) models for landslide susceptibility mapping in Relli Khola river basin of Darjeeling Himalaya, India

Landslide is one of the important disasters taking place on earth, which may be either a natural or man-made process. Landslides are more active and disastrous in hilly and mountainous regions. The present study aims to identify the landslide susceptibility areas in the Relli river basin in Darjeeling Himalaya using logistic regression (LR) and frequency ratio (FR) models. The GIS techniques have been used for landslide susceptibility mapping. A total number of 67 landslide locations have been identified from Google Earth images and multiple field surveys. 70% of landslide locations have been randomly selected and used as training data set for preparing landslide susceptibility map, and the remaining 30% have been used as validation data set. For the present study, 20 different factors like drainage density, drainage texture, infiltration number, stream frequency, stream junction frequency, stream power index, lithology, soil, relative relief, slope, maximum relief, drainage intensity, ruggedness number, rainfall, dissection index, aspect, relief class, and distance from stream, topographic wetness index and land use land cover have been used. The application of the logistic regression and frequency ratio model has demonstrated that the lower catchment of the basin has been widely dominated by the most landslide susceptibility areas than other parts of the catchment. Almost 6.92 sq km (4.05%) and 7.44 sq km (4.36%) areas out of 170.61 sq km area of the basin have been observed as very high and high landslide susceptibility categories, respectively, for FR model and 5.75 sq km (3.37%) and 1.86 sq km (1.09%) areas have been under very high and high landslide susceptibility zones for LR model. Finally, the ROC curve has been used to validate the models. The prediction capabilities of the models seem significant as the area under the curve value ranges from 75 to 81%.

Landslide susceptibility modeling using different artificial intelligence methods: a case study at Muong Lay district, Vietnam

Landslide is a natural hazard which causes huge loss of properties and human life in many places of the world. Mapping of landslide susceptibility is an important task for preventing and combating the landslides problems. Main objective of this study is to use different artificial intelligence methods namely support vector machines (SVM), artificial neural networks (ANN), logistic regression (LR), and reduced error-pruning tree (REPT) in the development of models for landslide susceptibility mapping of Muong Lay district of Vietnam. In total data of 217 landslide locations of the study area was used for the development and evaluation of the models. Nine landslide-conditioning factors were used for generating the datasets for training and validating the models. Results show that the SVM outperformed all other methods namely ANN, LR and REPT. Thus, it can be suggested that the SVM method is more useful in developing accurate and robust landslide prediction model.

A novel hybrid approach of landslide susceptibility modelling using rotation forest ensemble and different base classifiers

In the present study, Rotation Forest ensemble was integrated with different base classifiers to develop different hybrid models namely Rotation Forest based Support Vector Machines (RFSVM), Rotation Forest based Artificial Neural Networks (RFANN), Rotation Forest based Decision Trees (RFDT), and Rotation Forest based Naïve Bayes (RFNB) for landslide susceptibility modelling. The validity of these models was evaluated using statistical methods such as Root Mean Square Error (RMSE), Kappa index, accuracy, and the area under the success rate and predictive rate curves (AUC). Part of the landslide prone area of Pithoragarh district, Uttarakhand, Himalaya, India was selected as the study area. Results indicate that the RFDT is the best model showing the highest predictive capability (AUC = 0.741) in comparison to RFANN (AUC = 0.710), RFSVM (AUC = 0.701), and RFNB (AUC = 0.640) models. The present study would be helpful in the selection of best model for landslide susceptibility mapping.

An ensemble model for landslide susceptibility mapping in a forested area

This article proposes a new methodological approach using a combination of expert knowledge-based (analytic hierarchy process, AHP), bivariate (statistical index, SI) and multivariate (linear discriminant analysis, LDA) models for landslide susceptibility mapping (LSM) in Mazandran Province, Iran. Tolerance and variance inflation factor indicators were used for assessing multi-collinearity among parameters, and three (i.e. profile curvature, soil type and topography wetness index) of 18 factors were eliminated because of multi-collinearity issues. Fifteen geo-environmental conditioning factors including elevation, slope degree, slope aspect, plan curvature, slope length, convergence index, stream power index, distance from river, drainage density, distance from road, distance from fault, lithology, rainfall, land use/landcover and normalized difference vegetation index and 321 landslide locations (testing data set, 70% of total landslides) were used for modeling. The importance of factors showed that distance to road (AHP = 0.201, LDA = 0.301) was the most important factor in landslide occurrence. The validation results using validation data set (138 landslide locations, 30% of total landslides) and area under the receiver operating characteristic curve (AUROC) showed that the ensemble models AHP-LDR (83%), AHP-SI (95%) and SI-LDR (83%) had higher prediction accuracies than the individual AHP (82%), SI (82%) and LDA (79%) models and combination of AHP and SI models along with ALOS-PALSAR remote sensing data and geographic information system (GIS) technique provide powerful tool in LSM in the study area. The results of proposed novel methodological framework can be used by decision-makers and forest engineers for forest management spatially forest roads conservation that have key importance in sustainable development in local and regional scales.

Landslide susceptibility modelling using different advanced decision trees methods

ABSTRACTIn this paper, decision trees machine learning algorithms, namely Random Forest (RF), Alternating Decision Tree (ADT), and Logistic Model Tree (LMT), were applied for modelling of susceptibility of landslides at the Luc Yen district, Northern Vietnam. These methods were evaluated to compare the performance of models and for selection of the best model for landslide susceptibility mapping and prediction. In this study, data of 95 landslides events were analysed with 10 landslide affecting factors using the Correlation-Based Feature Selection (CFS). These factors are land use, elevation, slope, distance to roads, aspect, curvature, distance to faults, rainfall, lithology, and distance to rivers. Receiver Operating Characteristic (ROC) curve, statistical indices (sensitivity, specificity, and kappa), and Chi-square test were utilised for validating and comparing the models performance. The modelling results show that the performance of RF model (AUC = 0.839) is the best with the data at hand compared to the ADT model (0.827) and the LMT (0.809) model. The RF should be applied for the better landslide susceptibility mapping and management.