Landslide Inventory Map Research Articles

Automatic Landslide Inventory Mapping Approach Based on Change Detection Technique With Very-High-Resolution Images

Landslide inventory mapping (LIM) plays an important role in landslide susceptibility analysis. Many LIM approaches based on change detection techniques have been proposed, but with various drawbacks. For example, existing approaches have limited capability to capture the objects of varying shapes/sizes present in an area impacted by landslide. Many existing approaches are supervised and require parameter tuning. Moreover, some methods are prone to salt-and-pepper noise. To overcome these limitations, in this letter, an algorithm based on automatic adaptive region extension using very-high-resolution remote sensing images is developed. First, a simple yet effective k-means clustering method is used to generate training samples for landslide and nonlandslide classes, which refer to changed and unchanged areas, respectively. Second, an automatic adaptive region extension algorithm is developed and applied to each pixel of the postevent image, and the label of an extended region around a pixel is determined by the nearest distance between the central pixel and the changed or unchanged samples. Finally, the labels of a pixel are recorded because a pixel in different adaptive regions may be reassigned dissimilar labels, and the final label of the pixel is consistent with its maximum assigned label. To verify the performance of the proposed approach, we conducted experiments on two different landslide sites with VHR remote sensing images in Lantau Island, Hong Kong, China. Experimental results clearly demonstrate that the proposed approach has several advantages in improving the performance of LIM with VHR remote sensing images.

Spatial prediction of landslide susceptibility in western Serbia using hybrid support vector regression (SVR) with GWO, BAT and COA algorithms

In this study, we developed multiple hybrid machine-learning models to address parameter optimization limitations and enhance the spatial prediction of landslide susceptibility models. We created a geographic information system database, and our analysis results were used to prepare a landslide inventory map containing 359 landslide events identified from Google Earth, aerial photographs, and other validated sources. A support vector regression (SVR) machine-learning model was used to divide the landslide inventory into training (70%) and testing (30%) datasets. The landslide susceptibility map was produced using 14 causative factors. We applied the established gray wolf optimization (GWO) algorithm, bat algorithm (BA), and cuckoo optimization algorithm (COA) to fine-tune the parameters of the SVR model to improve its predictive accuracy. The resultant hybrid models, SVR-GWO, SVR-BA, and SVR-COA, were validated in terms of the area under curve (AUC) and root mean square error (RMSE). The AUC values for the SVR-GWO (0.733), SVR-BA (0.724), and SVR-COA (0.738) models indicate their good prediction rates for landslide susceptibility modeling. SVR-COA had the greatest accuracy, with an RMSE of 0.21687, and SVR-BA had the least accuracy, with an RMSE of 0.23046. The three optimized hybrid models outperformed the SVR model (AUC = 0.704, RMSE = 0.26689), confirming the ability of metaheuristic algorithms to improve model performance.

Assessing landslide susceptibility using machine learning models: a comparison between ANN, ANFIS, and ANFIS-ICA

The present study is aimed at conducting a comparative landslide susceptibility assessment in a landslide-prone subset area of the Tajan Watershed in northern Iran. For this aim, three probabilistic models were used namely: multilayer perceptron Artificial Neural Networks with a Back-Propagation algorithm (BPANN), Adaptive Neuro-Fuzzy Inference System (ANFIS), and the coupled ANFIS-Imperialist Competitive Algorithm (ANFIS-ICA). An all-inclusive landslide inventory map was prepared together with ten pivotal geo-environmental and anthropogenic landslide-controlling factors. Three indices including Pierce Skill Score (PSS), Cohen’s kappa, and the Area Under the Receiver Operating Characteristic curve (AUROC) were calculated from the confusion matrix and used to assess the performance of the models. Results in the validation stage revealed that the ensemble of ANFIS-ICA outperformed its counterparts with the respective PSS, kappa, and AUROC values of 0.766, 0.792, 0.966, followed by ANFIS (0.629, 0.666, 0.902), and ANN (0.603, 0.652, 0.866). Regarding the superior model (ANFIS-ICA), about 27% of the study area falls within high landslide susceptibility zones which needs to be considered for further risk mitigation measures and pragmatic actions. Our results pinpointed the outstanding performance of ANFIS-ICA ensemble in landslide susceptibility modeling at the watershed scale. Furthermore, it was found that the ANFIS-ICA model borrows most of its susceptibility pattern and performance from the distance to roads factor, although the total performance of the model is derived from the integration of all the factors. Moreover, this study attested to the advantages of hybrid algorithms and showed that the integration of machine learning models with evolutionary algorithms can be a new horizon to ensemble modeling. Utilization of well-adjusted ensemble models is pivotal for natural resource managers due particularly to their enhanced prediction powers which, in turn, can significantly reduce the social-economic losses emanated from failed predictions.

Modeling of Landslide susceptibility in a part of Abay Basin, northwestern Ethiopia

AbstractThe study area in northwestern Ethiopia is one of the most landslide-prone regions, which is characterized by frequent high landslide occurrences. To predict future landslide occurrence, preparing a landslide susceptibility mapping is imperative to manage the landslide hazard and reduce damages of properties and loss of lives. Geographic information system (GIS)-based frequency ratio (FR), information value (IV), certainty factor (CF), and logistic regression (LR) methods were applied. The landslide inventory map is prepared from historical records and Google Earth imagery interpretation. Thus, 717 landslides were mapped, of which 502 (70%) landslides were used to build landslide susceptibility models, and the remaining 215 (30%) landslides were used to model validation. Eleven factors such as lithology, land use/cover, distance to drainage, distance to lineament, normalized difference vegetation index, drainage density, rainfall, soil type, slope, aspect, and curvature were evaluated and their relationship with landslide occurrence was analyzed using the GIS tool. Then, landslide susceptibility maps of the study area are categorized into very low, low, moderate, high, and very high susceptibility classes. The four models were validated by the area under the curve (AUC) and landslide density. The results for the AUC are 93.9% for the CF model, which is better than 93.2% using IV, 92.7% using the FR model, and 87.9% using the LR model. Moreover, the statistical significance test between the models was performed using LR analysis by SPSS software. The result showed that the LR and CF models have higher statistical significance than the FR and IV methods. Although all statistical models indicated higher prediction accuracy, based on their statistical significance analysis result (Table 5), the LR model is relatively better followed by the CF model for regional land use planning, landslide hazard mitigation, and prevention purposes.

Landslides triggered by the 6 September 2018 Mw 6.6 Hokkaido, Japan: an updated inventory and retrospective hazard assessment

On September 6, 2018, an Mw 6.6 earthquake occurred in Hokkaido, Japan, triggering massive landslides. Based on the 0.3 m-resolution Pleiades-1 satellite image on the Google Earth platform, we constructed an updated co-seismic landslide inventory for this event, which permitted to map 12,586 landslides. These slope failures are featured by a large number, wide spread, large scales, and local dense distribution. Considering nine influence factors, i.e. elevation, slope angle, slope aspect, curvature, distance to road, distance to river, distance to the epicenter, lithology, and topography, the logistic regression model was applied to carry out a retrospective hazard assessment of earthquake-triggered landslides in the affected area, and the rationality of such prediction was verified by the real landslides mapped. Results show that the landslide hazard evaluation map is in good agreement with the actual landslide distribution. The success rate of the hazard assessment map is 87.0% compared to training samples, and 86.5% compared to testing samples, respectively. The updated landslide inventory is a useful complement to the existing landslide inventory maps and the landslide hazard assessment map is helpful for landslide disaster prevention and reduction and reconstruction planning in the affected area.

Deep convolutional neural network–based pixel-wise landslide inventory mapping

This paper reports a feasible alternative to compile a landslide inventory map (LIM) from remote sensing datasets using the application of an artificial intelligence–driven methodology. A deep convolutional neural network model, called LanDCNN, was developed to generate segmentation maps of landslides, and its performance was compared with the benchmark model, named U-Net, and other conventional object-based methods. The landslides that occurred in Lantau Island, Hong Kong, were taken as the case study, in which the pre- and post-landslide aerial images, and a rasterized digital terrain model (DTM) were used. The assessment reveals that LanDCNN trained with bitemporal images and DTM yields the smoothest and most semantically meaningfully LIM, compared to other methods. This LIM is the most balanced segmentation results, represented by the highest F1 measure among all analyzed cases. With the encoding capability of LanDCNN, the application of DTM as the input renders better LIM production, especially when the landslide signatures are relatively subtle. With the computational setup used in this study, LanDCNN requires ~ 3 min to map landslides from the datasets of approximately 25 km2 in area and with a resolution of 0.5 m. In short, the proposed landslide mapping framework, featured LanDCNN, is scalable to handle the vast amount of remote sensing data from different types of measurements within a short processing period.

Performance evaluation of ensemble learning techniques for landslide susceptibility mapping at the Jinping county, Southwest China

The objective of this study is to investigate different ensemble learning techniques namely Bagging, Boosting, and Stacking for LSM at the Jinping county, Southwest China. Two well-known machine learning classifiers such as C4.5 decision tree (C4.5) and artificial neural network (ANN) were served as base-learners. A total of five ensemble models, including the Bag-C4.5 model, the Boost-C4.5 model, the Bag-ANN model, the Boost-ANN model, and the Stacking C4.5-ANN model, were constructed by using various ensemble techniques and base-learners. A landslide inventory map and 12 landslide-related factors have been prepared as the spatial database for landslide modeling. The importance of factors was verified using the information gain (IG) method. It turns out that the distance to roads has the greatest contribution to landslide susceptibility assessment. Subsequently, various landslide models were evaluated regarding the goodness of fit, generalization capability, and robustness. The area under the ROC curve (AUC), statistical analysis, and stability index (SI) were used as performance metrics. Evaluation results showed that ensemble learning techniques significantly refined individual landslide models such as the C4.5 (AUC = 0.832) and ANN (AUC = 0.870). In particular, Boosting-based models, e.g., the Boost-C4.5 model (AUC = 0.945) and the Boost-ANN model (AUC = 0.903), gained a higher performance than the Stacking C4.5-ANN model (AUC = 0.900), the Bag-ANN (AUC = 0.892), and the Bag-C4.5 (AUC = 0.878). Additionally, the best modeling robustness was achieved by the Stacking C4.5-ANN method (SI = 1). The results indicate that the Boosting technique has great confidence in strengthening the predictive accuracy for LSM. Also, the Stacking can provide a promising method for stable and improved landslide modeling. Findings from this study may assist to refine the quality of LSM and facilitate risk management for the study area or other similar regions.

Comparative analysis of gradient boosting algorithms for landslide susceptibility mapping

The aim of the study is to compare four recent gradient boosting algorithms named as Gradient Boosting Machine (GBM), Categorical Boosting (CatBoost), Extreme Gradient Boosting (XGBoost), and Light Gradient Boosting Machine (LightGBM) for modelling landslide susceptibility (LS). In the first step of the study, the geodatabase including landslide inventory map and landslide conditioning factors was constructed. In the second step, chi-square (CHI) statistic-based feature selection (FS) technique was utilized to compute the importance of the landslide causative factors. In the third step, tree-based ensemble learning algorithms were applied to predict the potential distribution of landslide susceptibility. Also, the prediction performance of ensemble methods was compared to that of Random Forest (RF) ensemble method. Finally, the prediction capabilities of the methods were assessed using overall accuracy (Acc), area under the receiver operating characteristic curve (AUC), kappa index, root mean square error (RMSE), and F score measures. In order to further evaluation, the McNemar's test was utilized to assess statistical significance in the differences between the four gradient boosting models. The accuracy results indicated that the CatBoost model had the highest prediction capability (Acc= 0.8503 and AUC= 0.8975), followed by the XGBoost (Acc= 0.8336 and AUC= 0.8860), the LightGBM (Acc= 0.8244 and AUC= 0.8796) and the GBM (Acc= 0.8080 and AUC= 0.8685). On the other hand, the estimated accuracy measures considered in this study showed that the RF method had the lowest prediction capability of compared the others. Although the individual performances of the methods were found to be acceptable level, the CatBoost method showed the superior performance compared to others with respect to the AUC and Acc values estimated in this study. The results of the study confirmed that the relatively new ensemble learning techniques were efficient and robust for producing LS maps and furthermore, it is probably that these algorithms will be preferred more often in the future studies due to their robustness.

GIS-based landslide susceptibility modeling: A comparison between fuzzy multi-criteria and machine learning algorithms

Hazards and disasters have always negative impacts on the way of life. Landslide is an overwhelming natural as well as man-made disaster that causes loss of natural resources and human properties throughout the world. The present study aimed to assess and compare the prediction efficiency of different models in landslide susceptibility in the Kysuca river basin, Slovakia. In this regard, the fuzzy decision-making trial and evaluation laboratory combining with the analytic network process (FDEMATEL-ANP), Naïve Bayes (NB) classifier, and random forest (RF) classifier were considered. Initially, a landslide inventory map was produced with 2000 landslide and non-landslide points by randomly divided with a ratio of 70%:30% for training and testing, respectively. The geospatial database for assessing the landslide susceptibility was generated with the help of 16 landslide conditioning factors by allowing for topographical, hydrological, lithological, and land cover factors. The ReliefF method was considered for determining the significance of selected conditioning factors and inclusion in the model building. Consequently, the landslide susceptibility maps (LSMs) were generated using the FDEMATEL-ANP, Naïve Bayes (NB) classifier, and random forest (RF) classifier models. Finally, the area under curve (AUC) and different arithmetic evaluation were used for validating and comparing the results and models. The results revealed that random forest (RF) classifier is a promising and optimum model for landslide susceptibility in the study area with a very high value of area under curve (AUC = 0.954), lower value of mean absolute error (MAE = 0.1238) and root mean square error (RMSE = 0.2555), and higher value of Kappa index (K = 0.8435) and overall accuracy (OAC = 92.2%).

Application and comparison of different ensemble learning machines combining with a novel sampling strategy for shallow landslide susceptibility mapping

The existence of shallow landslide brings huge threats to the human lives and economic development, as the Lang County, Southeastern Tibet prone to landslide. Landslide susceptibility mapping (LSM) is considered as the key for the prevention of hazard. The primary goal of the present study is to assess and compare four models: classification and regression tree, gradient boosting decision tree (GBDT), adaptive boosting-decision tree and random forest for the performance of landslide susceptibility modeling. Firstly, a landslide inventory map consisting of 229 historical shallow landslide locations was prepared and the same number of non-landslide points was determined by k-means clustering. Secondly, 12 conditioning factors were considered in the landslide susceptibility modeling. The prediction performance of the four models were estimated by fivefold cross validation and relative operating characteristic curve (ROC), area under the ROC curve (AUC) and statistical measures. The results showed that the GBDT performed best in the training and validation dataset, with the highest prediction capability (AUC = 0.986 and 0.940), highest accuracy value (95.3% and 88.1%) and highest kappa index (0.904 and 0.772), respectively. Therefore, the GBDT was considered to be the most suitable model and applied to the whole study area for LSM. The results of this study also demonstrate that the performance can be enhanced with the use of ensemble learning. The sampling strategy of non-landslide points can be improved by combining with clustering analysis which are more reasonable.

Rainfall-Induced Shallow Landslide Susceptibility Mapping at Two Adjacent Catchments Using Advanced Machine Learning Algorithms

Landslides impact on human activities and socio-economic development, especially in mountainous areas. This study focuses on the comparison of the prediction capability of advanced machine learning techniques for the rainfall-induced shallow landslide susceptibility of Deokjeokri catchment and Karisanri catchment in South Korea. The influencing factors for landslides, i.e., topographic, hydrologic, soil, forest, and geologic factors, are prepared from various sources based on availability, and a multicollinearity test is also performed to select relevant causative factors. The landslide inventory maps of both catchments are obtained from historical information, aerial photographs and performed field surveys. In this study, Deokjeokri catchment is considered as a training area and Karisanri catchment as a testing area. The landslide inventories contain 748 landslide points in training and 219 points in testing areas. Three landslide susceptibility maps using machine learning models, i.e., Random Forest (RF), Extreme Gradient Boosting (XGBoost) and Deep Neural Network (DNN), are prepared and compared. The outcomes of the analyses are validated using the landslide inventory data. A receiver operating characteristic curve (ROC) method is used to verify the results of the models. The results of this study show that the training accuracy of RF is 0.756 and the testing accuracy is 0.703. Similarly, the training accuracy of XGBoost is 0.757 and testing accuracy is 0.74. The prediction of DNN revealed acceptable agreement between the susceptibility map and the existing landslides, with a training accuracy of 0.855 and testing accuracy of 0.802. The results showed that the DNN model achieved lower prediction error and higher accuracy results than other models for shallow landslide modeling in the study area.

An Application of Sentinel-1, Sentinel-2, and GNSS Data for Landslide Susceptibility Mapping

In this study, we used Sentinel-1 and Sentinel-2 data to delineate post-earthquake landslides within an object-based image analysis (OBIA). We used our resulting landslide inventory map for training the data-driven model of the frequency ratio (FR) for landslide susceptibility modelling and mapping considering eleven conditioning factors of soil type, slope angle, distance to roads, distance to rivers, rainfall, normalised difference vegetation index (NDVI), aspect, altitude, distance to faults, land cover, and lithology. A fuzzy analytic hierarchy process (FAHP) also was used for the susceptibility mapping using expert knowledge. Then, we integrated the data-driven model of the FR with the knowledge-based model of the FAHP to reduce the associated uncertainty in each approach. We validated our resulting landslide inventory map based on 30% of the global positioning system (GPS) points of an extensive field survey in the study area. The remaining 70% of the GPS points were used to validate the performance of the applied models and the resulting landslide susceptibility maps using the receiver operating characteristic (ROC) curves. Our resulting landslide inventory map got a precision of 94% and the AUCs (area under the curve) of the susceptibility maps showed 83%, 89%, and 96% for the F-AHP, FR, and the integrated model, respectively. The introduced methodology in this study can be used in the application of remote sensing data for landslide inventory and susceptibility mapping in other areas where earthquakes are considered as the main landslide-triggered factor.

Landslide susceptibility modelling using integrated evidential belief function based logistic regression method: A study from Southern Western Ghats, India

The present study aims to develop a spatially integrated evidential belief function based logistic regression model (EBF-LR) for landslide susceptibility mapping in a naturally sloping terrain of Southern Western Ghats in Kerala, India. For this, a landslide inventory map of 83 previous landslides was prepared using satellite imageries and verified in field check. Thereafter the landslide inventory was randomly divided into 70%–30% basis for model training and testing. Twelve landslide conditioning factors viz., lithology, land use/land cover, NDVI, slope angle, slope aspect, profile curvature, distance to stream, distance to roads, distance to lineaments, soil texture, topographic wetness index and average annual rainfall were considered for landslide susceptibility modelling. The resultant susceptibility maps were validated using receiver operating characteristics curve (ROC) with area under the curve (AUC) value, sensitivity, specificity, kappa index, mean absolute error (MAE) and root mean square error (RMSE). Analysis shows that the integrated EBF-LR model outweigh other conventional bivariate and multivariate approaches with a ROC-AUC value of 0.935, Kappa score 0.719, sensitivity 0.885, specificity 0.833 and least RMSE of 0.456 in the validation stage. The study also reveals that anthropogenic disturbances have a significant role on landslide initiation in the study area. Considering the ROC- AUC, MAE, RMSE and other validation measures, the accuracy of the proposed landslide susceptibility model is satisfactory and can be used for future land use planning and landslide mitigation in the study area.

Random Forest-Based Landslide Susceptibility Mapping in Coastal Regions of Artvin, Turkey

Natural disasters such as landslides often occur in the Eastern Black Sea region of Turkey owing to its geological, topographical, and climatic characteristics. Landslide events occur nearly every year in the Arhavi, Hopa, and Kemalpaşa districts located on the Black Sea coast in the Artvin province. In this study, the landslide susceptibility map of the Arhavi, Hopa, and Kemalpaşa districts was produced using the random forest (RF) model, which is widely used in the literature and yields more accurate results compared with other machine learning techniques. A total of 10 landslide-conditioning factors were considered for the susceptibility analysis, i.e., lithology, land cover, slope, aspect, elevation, curvature, topographic wetness index, and distances from faults, drainage networks, and roads. Furthermore, 70% of the landslides on the landslide inventory map were used for training, and the remaining 30% were used for validation. The RF-based model was validated using the area under the receiver operating characteristic (ROC) curve. Evaluation results indicated that the success and prediction rates of the model were 98.3% and 97.7%, respectively. Moreover, it was determined that incorrect land-use decisions, such as transforming forest areas into tea and hazelnut cultivation areas, induce the occurrence of landslides.

Susceptibility mapping of damming landslide based on slope unit using frequency ratio model

In the mountainous area of southwest China, there is a large number of large hydropower stations that have already been built or under construction. The construction and the safety of the operation of hydropower stations directly affect the residents’ safety of life and property on both sides of the downstream river. Landslide hazards are one of the main problems affecting the construction and operation of hydropower stations. Therefore, it is currently more urgent to evaluate landslide susceptibility within the basin. The goal of this paper is to carry out landslide susceptibility mapping that affects the construction and operation of hydropower stations. In the case of a large research area and lack of landslide hazard data, this paper selects the sample of blocking river landslide in a larger area with similar topography and geological conditions and applies the quantitative method to carry out landslide sensitivity mapping. In this paper, the slope unit works as the evaluation unit, 191 landslide sample data were obtained, and nine landslide conditioning factors (slope, aspect, curvature, lithology, distance to fault, distance to the river, land use, soil, and the topographic wetness index (TWI)) were obtained. Combined with the quantitative frequency model (FR model), the landslide susceptibility map was obtained. The research process and results show that (1) when the study area is large, the landslide data is scarce, and it is difficult to get the landslide inventory map; the landslide samples can be selected in a larger area with similar topography and geological conditions. It is feasible to use the analysis data of landslide samples to evaluate the landslide susceptibility; (2) for specific industry applications such as hydropower station operation safety, the landslide evaluation range can be reduced by the spatial correlation between landslide aspect and river flow direction, which is beneficial to the management department to reduce the scope of attention; and (3) the verification results show that the slope unit can be used for quantitative evaluation of landslide sensitivity based on the FR model and can obtain the highest evaluation accuracy. The main advantage of this method is its capability to carry out landslide susceptibility mapping with quantitative methods and to ensure that higher mapping accuracy is obtained in areas where landslide hazard data is lacking.

Landslide susceptibility mapping using hybridized block modular intelligence model

Landslide susceptibility map (LSM) provide useful tool for decision makers in hazard mitigation concerns. In the present paper, a novel hybrid block-based neural network model (HBNN) for the purpose of producing high-resolution LSM was developed. This hybrid approach was found through the mixture of expert modular structures and divide-and-conquer strategy incorporated with genetic algorithm (GA). The introduced HBNN then was applied on southern part of Guilan province (north of Iran) using 14 causative factors covering topographic and geomorphologic features, and geological and tectonical factors as well as hydrology, land data, and climate conditions. The landslide inventory map was provided using a synergy work from monitored events, interpretation of aerial photographs, and carried out geotechnical investigations in the area as well as field surveys. To insight, the predictability of proposed HBNN was compared with two developed models using multilayer perceptrons (MLPs) and generalized feed forward neural network (GFFN). The generated LSM was validated using receiver operating characteristic (ROC) curves, statistical error indices, and sensitivity and weight analyses as well as monitored landslides. Based on the compared metrics, HBNN with 86.52% and 90.15% in prediction and success rate as well as 89.36% for precision-recall curve demonstrated more consistent tool for future landslide susceptibility zonation. This implies on capability of developed HBNN in producing higher resolution and more reliable LSM for urban and land-use planners.

Performance Evaluation of GIS-Based Artificial Intelligence Approaches for Landslide Susceptibility Modeling and Spatial Patterns Analysis

The main purpose of this study was to apply the novel bivariate weights-of-evidence-based SysFor (SF) for landslide susceptibility mapping, and two machine learning techniques, namely the naïve Bayes (NB) and Radial basis function networks (RBFNetwork), as benchmark models. Firstly, by using aerial photos and geological field surveys, the 263 landslide locations in the study area were obtained. Next, the identified landslides were randomly classified according to the ratio of 70/30 to construct training data and validation models, respectively. Secondly, based on the landslide inventory map, combined with the geological and geomorphological characteristics of the study area, 14 affecting factors of the landslide were determined. The predictive ability of the selected factors was evaluated using the LSVM model. Using the WoE model, the relationship between landslides and affecting factors was analyzed by positive and negative correlation methods. The above three hybrid models were then used to map landslide susceptibility. Thirdly, the ROC curve and various statistical data (SE, 95% CI and MAE) were used to verify and compare the predictive power of the model. Compared with the other two models, the Sysfor model had a larger area under the curve (AUC) of 0.876 (training dataset) and 0.783 (validation dataset). Finally, by quantitatively comparing the susceptibility values of each pixel, the differences in spatial morphology of landslide susceptibility maps were compared, and the model was found to have limitations and effectiveness. The landslide susceptibility maps obtained by the three models are reasonable, and the landslide susceptibility maps generated by the SysFor model have the highest comprehensive performance. The results obtained in this paper can help local governments in land use planning, disaster reduction and environmental protection.

GIS-based evaluation of landslide susceptibility using hybrid computational intelligence models

Landslides have caused huge economic and human losses in China. Mapping of landslide susceptibility is an important tool to prevent and control landslide disasters. The purpose of this study is to make use of a hybrid machine learning approach by combining the reduced-error pruning trees (Rept) with a series of ensemble techniques (Bagging, Dagging, and Real Adaboost) and compare the performance of each combination for landslide susceptibility modeling. The combination of Rept model and Real Adaboost (RRept)is a novel application in the field of landslide susceptibility. Firstly, a landslide inventory map was prepared with 298 determined historical landslides events in the study area, 209 landslides (70%) were randomly selected for the training dataset and the remaining 89 landslides (30%) were used for validation dataset. On this basis, 16 landslide influencing factors were included in the landslide susceptibility evaluation (slope angle, elevation, slope aspect, sediment transport index (STI), topographical wetness index (TWI), stream power index (SPI), profile curvature, plan curvature, distance to rivers, distance to roads, distance to faults, soil, normalized difference vegetation index (NDVI), landuse, lithology and rainfall). Secondly, the correlation attribute evaluation (CAE) method was used to select the most important factors for the proposed landslide susceptibility model. The results show that all the factors contribute to the occurrence of landslide. Slope angle, Landuse, Elevation, Distance to roads, Soil and Lithology have the greatest influence on the occurrence of landslide. The receiver operating characteristics (ROC), standard error (SE), 95% confidence interval and mean absolute error (MAE) were then used to validate and compare the performance of the model. The best model should have the largest AUC value, the smallest SE, the narrowest 95% CI and the smallest MAE. The results show that the three hybrid models perform better than the Rept model alone. For the training data set, the RRept model has highest AUC value (0.927), the smallest SE (0.121), the narrowest 95% confidence interval (0.898–0.95) and the lowest MAE (0.20). For the validation data set, the RRept model has the highest AUC value (0.745), the narrowest 95% confidence interval (0.674–0.807) and the lowest MAE (0.33). The RRept has the highest predictive power for landslide susceptibility evaluation. The results show that a hybrid method improves the prediction ability of the base Rept model. In addition, the RRept model is a promising comprehensive model that can be applied to landslide susceptibility mapping.

Comparative Analysis of Pixel and Object Based Classification Approach for Rapid Landslide Delineation with the Aid of Open Source Tools in Garhwal Himalaya

Abstract Landslides are the natural geomorphic processes that are essential for the landscape development. Loss of tremendous amount of life, natural resources as well as property has made this phenomenon a natural disaster. Number of researches have been done to understand this phenomenon and how to overcome this calamity with the aid of upcoming modern information services and technological advancements. Despite all these, there are still certain undocumented slope failure events due to inaccessibility and lack of a proper database. With the advancement of satellite remote sensing and geographical information system (GIS), it has certainly become easier to monitor and prepare landslide database especially in rough and rugged terrain of the Himalaya for delineation of risk zones. Free availability of high-resolution images and an open source efficient software have certainly been proved advantageous for this purpose. This paper aims at quick and accurate landslide inventory mapping, using high resolution Sentinel 2 data along with normalized difference vegetation index (NDVI), over which unsupervised and object-based image analysis (OBIA) was done to extract landslide features in an efficient manner. Both these processes were achieved with the aid of open source SAGA (System for Automated Geoscientific Analyses) software. The performances of these classifications were analyzed for their quantitative reliability in extracting the landslide features in two different areas (test area 1 and 2). In this, the object based image classification outstands, with the overall accuracy percentage difference of 17.11% higher than pixel-based classification approach in test area 1 and about 21% higher in test area 2.

Assessing the predictive capability of ensemble tree methods for landslide susceptibility mapping using XGBoost, gradient boosting machine, and random forest

Decision tree-based classifier ensemble methods are a machine learning (ML) technique that combines several tree models to produce an effective or optimum predictive model, and that allows well-predictive performance especially compared to a single model. Thus, selecting a proper ML algorithm help us to understand possible future occurrences by analyzing the past more accurate. The main purpose of this study is to produce landslide susceptibility map of the Ayancik district of Sinop province, situated in the Black Sea region of Turkey using three featured regression tree-based ensemble methods including gradient boosting machines (GBM), extreme gradient boosting (XGBoost), and random forest (RF). Fifteen landslide causative factors and 105 landslide locations occurred in the region were used. The landslide inventory map was randomly divided into training (70%) and testing (30%) dataset to construct the RF, XGBoost and GBM prediction models. Symmetrical uncertainty measure was utilized to determine the most important causative factors, and then the selected features were used to construct susceptibility prediction models. The performance of the ensemble models was validated using different accuracy metrics including Area under the curve (AUC), overall accuracy (OA), Root mean square error (RMSE), and Kappa coefficient. Also, the Wilcoxon signed-rank test was used to assess differences between optimum models. The accuracy results showed that the model of XgBoost_Opt model (the model created by optimum factor combination) has the highest prediction capability (OA = 0.8501 and AUC = 0.8976), followed by the RF_opt (OA = 0.8336 and AUC = 0.8860) and GBM_Opt (OA = 0.8244 and AUC = 0.8796). When the Wilcoxon sign-rank test results were analyzed, XgBoost_Opt model, which is the best subset combinations, were confirmed to be statistically significant considering other models. The results showed that, the XGBoost method according to optimum model achieved lower prediction error and higher accuracy results than the other ensemble methods.