Landslide Susceptibility Assessment Research Articles

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.

FSLAM: A QGIS plugin for fast regional susceptibility assessment of rainfall-induced landslides

Shallow slope failures triggered by rainfall commonly pose considerable risks in mountainous areas. In order to delineate areas where landslides are more prone to occur within a region, we have designed and developed a Python QGIS plugin named Fast Shallow Landslide Assessment Model (FSLAM). The plugin integrates a simplified hydrological model and a geotechnical model based on the infinite slope theory and contains two principal modules: runoff and slope stability modelling. It can output up to 15 raster maps describing the hydrological and stability conditions in a short computational time. Firstly, we explain the design of graphical user interface and the elements of the plugin. Then, the Berguedà area in NE Spain is used as case study to present the procedure of the plugin application. The results show that the accuracy of landslide susceptibility assessment performed by FSLAM-plugin is high and the computing time is only a few minutes.

Landslide susceptibility assessment using triangular fuzzy number-analytic hierarchy processing (TFN-AHP), contributing weight (CW) and random forest weighted frequency ratio (RF weighted FR) at the Pengyang county, Northwest China

Landslide susceptibility assessment using triangular fuzzy number-analytic hierarchy processing (TFN-AHP), contributing weight (CW) and random forest weighted frequency ratio (RF weighted FR) at the Pengyang county, Northwest China

Evaluation of Parameter Selection in the Bivariate Statistical-based Landslide Susceptibility Modeling (Case Study: the Citarik Sub-watershed, Indonesia)

A landslide susceptibility mapping is essential for landslide hazard mitigation to reduce the associated risk. This paper aims to present the results of the landslide susceptibility modeling in the Citarik sub-watershed using three bivariate statistical-based methods, i.e., frequency ratio (FR), information value (IV), and weight of evidence (WoE). The main objective of this study is to evaluate the significance of the threshold of the area under curve (AUC) value in parameter selection. In this study, 118 landslide pixels were compiled from Google Earth images, unmanned aircraft vehicle (UAV) aerial photos taken just after the landslide, official landslide reports, and field observation. Thirteen landslide causative factors were prepared in Geographic Information System (GIS) environment, derived from various satellite images and maps. The landslide data were divided into two groups, 70% of data as training data and the rest as test data. Two scenarios that involve a different number of parameters were compared to explain the threshold of the AUC value in parameter selection and model accuracy. The result of this study shows that the AUC value threshold of 0.6 for parameter selection cannot be applied in all cases, and the performance of both two scenarios was excellent in assessing landslide susceptibility in this study area. Those three landslide susceptibility zonation maps of the best scenario showed that the sub-watershed's northern, northeastern, south-eastern, and southern parts were under high to very high susceptibility to landslides, including the Cimanggung area where a recent deadly double landslide occurred.

Comprensión de la sensibilidad a las propiedades del suelo y condiciones de lluvia de dos modelos de estabilidad de taludes basados en la física

Physically-based models have been used to assess landslide susceptibility, hazard, and risk in many regions worldwide. They have also been regarded as valuable tools for landslide prediction and the development or improvement of landslide early warning systems. They are usually validated to demonstrate their predictive capacity, but they are not deeply studied regularly to understand the sensitivity of the input variables and the behavior of the models under many different rainfall scenarios. In this research paper, we studied two distributed physically-based models for shallow landslides: SLIP and Iverson. For this, the first-order second-moment (FOSM) method was used to calculate the contribution of random input variables (soil strength, unit weight, and permeability parameters) to the variance of the factor of safety. Different intensity and duration rainfall events were simulated to assess the response of the models to those rainfall conditions in terms of the factor of safety and failure probability. The results showed that the shear strength (cohesion and friction angle, in order of significance) parameters have the largest contribution to the variance in both models, but they vary depending on geological, geotechnical, and topographic conditions. The Iverson and SLIP models respond in different ways to the variation of rainfall conditions: for shorter durations (e.g. ≤ 8 h), increasing the intensity caused more unstable areas in the SLIP model, while for longer durations the unstable areas were considerably higher for the Iverson model. Understanding those behaviors can be useful for practical and appropriate implementation of the models in landslide assessment projects.

The importance of investigating causative factors and training data selection for accurate landslide susceptibility assessment: the case of Ain Lahcen commune (Tetouan, Northern Morocco)

Landslide susceptibility maps (LSMs) rely on statistical association for weighting the impact of each predictive factor class. However, the effects of using different training datasets and causative factors combinations on the accuracy of the results are rarely discussed. Accordingly, we investigated the effects of training data and computation technique selection on the accuracy and performance of LSMs. Results show that including relict landslides in the training data diminishes the accuracy of the LSMs compared to models that exclude such processes from the input. As for the effect of choosing different combinations of predictive variables, it was shown to be less significant. After investigating association and causality between predictive variables, the best performing models produced using frequency ratio and logistic regression tend to be more accurate than those including all variables. As for the artificial neural networks, it seems to outperform other models. Indeed, such investigations are beneficial for improving LSMs.

Review of landslide susceptibility assessment based on knowledge mapping

Review of landslide susceptibility assessment based on knowledge mapping

Landform classification: a high-performing mapping unit partitioning tool for landslide susceptibility assessment—a test in the Imera River basin (northern Sicily, Italy)

Landform classification: a high-performing mapping unit partitioning tool for landslide susceptibility assessment—a test in the Imera River basin (northern Sicily, Italy)

Landslide susceptibility assessment on the left side of the Izvorul Muntelui Lake bank, Romania

Landslide susceptibility assessment on the left side of the Izvorul Muntelui Lake bank, Romania

Landslide Susceptibility Assessment by Mapping and Diachronic Analysis: Case of Bushwira

Landslide Susceptibility Assessment by Mapping and Diachronic Analysis: Case of Bushwira

Estimating the Geological Strength Index (GSI) in Regional Seismic‐Landslide Zonation Using the Empirical Regression Model

The assessment of the strength parameters of geological formations in regional scale which encounters thousands of slopes is a complicated approach and time consuming and needs huge field work. This issue is an important research topic concerning the regional seismic‐landslide susceptibility analysis or hazard zonation. An empirical regression model was presented to estimate the Geological Strength Index (GSI) with an implication on geological quadrangle of Gorgan region at Alborz mountains range (north of Iran). Two main sets of data were applied in this study: (1) geomorphological data including the slope height, aspect, and distance from faults and distance from thrusts and (2) the physical and mechanical properties of rocks including the unit weight, uniaxial compressive strength (σci), and the petrographic constant (mί) of intact rock. The first group was extracted from a 1 : 100,000 digital geologic map and 10 m digital elevation model (DEM) and the second group was obtained from the Hoek–Brown failure criterion recommended tables. Linear regression equations were generated applying data collected from 294 studied stations using SPSS software. The regression equation predicted GSI in terms of (1) the distance from faults, (2) the distance from thrusts, and (3) the uniaxial compressive strength (σci). The equation had an R2 value of 0.739 and thus fit well to the data. The new method in its present state was recommended for the estimation of the GSI values in regional scale conditions for the assessment of landslide susceptibility and hazard mapping or post events landslide occurrence prediction in the case of probable big earthquakes in Alborz area that is required for emergency responses. The results indicated that the estimation error was about ±30 while the average error was within +5 and −5 and average error percentage was about 3%.

Assessment of landslide susceptibility using DBSCAN-AHD and LD-EV methods

Assessment of landslide susceptibility using DBSCAN-AHD and LD-EV methods

Regional-scale landslide susceptibility assessment for the hilly state of Uttarakhand, NW Himalaya, India

Regional-scale landslide susceptibility assessment for the hilly state of Uttarakhand, NW Himalaya, India

Susceptibility Assessment for Landslide Initiated along Power Transmission Lines

The power network has a long transmission span and passes through wide areas with complex topography setting and various human engineering activities. They lead to frequent landslide hazards, which cause serious threats to the safe operation of the power transmission system. Thus, it is of great significance to carry out landslide susceptibility assessment for disaster prevention and mitigation of power network. We, therefore, undertake an extensive analysis and comparison study between different data-driven methods using a case study from China. Several susceptibility mapping results were generated by applying a multivariate statistical method (logistic regression (LR)) and a machine learning technique (random forest (RF)) separately with two different mapping-units and predictor sets of differing configurations. The models’ accuracies, advantages and limitations are summarized and discussed using a range of evaluation criteria, including the confusion matrix, statistical indexes, and the estimation of the area under the receiver operating characteristic curve (AUROC). The outcome showed that machine learning method is well suitable for the landslide susceptibility assessment along transmission network over grid cell units, and the accuracy of susceptibility models is evolving rapidly from statistical-based models toward machine learning techniques. However, the multivariate statistical logistic regression methods perform better when computed over heterogeneous slope terrain units, probably because the number of units is significantly reduced. Besides, the high model predictive performances cannot guarantee a high plausibility and applicability of subsequent landslide susceptibility maps. The selection of mapping unit can produce greater differences on the generated susceptibility maps than that resulting from the selection of modeling methods. The study also provided a practical example for landslide susceptibility assessment along the power transmission network and its potential application in hazard early warning, prevention, and mitigation.

Comparison of landslide susceptibility models and their robustness analysis: a case study from the NW Himalayas, Pakistan

Machine learning methods are considered as most effective approaches to accomplish landslide susceptibility analysis around the globe. Landslide susceptibility maps (LSMs) have been frequently executed by statistical models in NW Himalaya. However, the comparison and applications of the statistical models with modern machine learning techniques has not been fully explored in this region. Hence, this study aims to compare the predicted performance of statistical and popular machine learning models to explore robust landslide prediction model in the landslide-prone area of NW Himalaya and investigate the compensations and limitations of these models to grasp a more precise and consistent result. This study presented machine learning approaches based on the artificial neural network (ANN), support vector machine (SVM) and logistic regression (LR) and the statistical methods based on the frequency ratio (FR), information value (InfoV) and weight of evidence (WoE). For this purpose, first an inventory map of 1507 landslides was prepared and randomly divided into training (70%) and testing (30%) dataset. Furthermore, 12 landslide conditioning factors (LCFs) were extracted from geospatial dataset to prepare thematic layers in ArcGIS. Thereafter, factor analysis was performed to eliminate colinear and least important variables which can mislead the results. The results showed that all selected LCFs are noncolinear and have significant contribution on landslides initiation, however, lithology, slope angle, annual rainfall and landuse were most influential factors. For modeling purpose, landslide inventory was correlated against all LCFs and trained into six models to produce respective LSMs. Finally, the performance of produced LSM models was validated and compared through area under receiver operating characteristic curve (AUROC), Accuracy, Recall, F 1-score and Cohen’s Kappa coefficients to assess the robustness of employed models. The results exhibit that the performance scores of machine learning models were considerably superior than statistical models. While, the AUROC values based on validation dataset indicate that LR (0.89) has better prediction ability followed by SVM (0.86), ANN (0.84), FR (0.83), InfoV (0.82) and WoE (0.81) in this study. Therefore, it is reasoned out that the machine learning methods are more reliable in generating adequate LSMs. However, the LR is recommended as most efficient model for predicting landslide susceptible zones in study region and thus can be considered as robust model for landslide susceptibility assessment in similar geo-environmental regimes.

Landslide susceptibility assessment in the Himalayan range based along Kasauli – Parwanoo road corridor using weight of evidence, information value, and frequency ratio

The Himalayan region is prone to landslides. Rainfall-induced slope failure activities in the Indian Himalaya cause considerable damage, posing a serious risk to life and property. This study attributes information regarding landslide triggering parameters to further delineate landslide susceptibility maps of Himachal Pradesh in the Himalayan region. The landslide inventory map was created using information from field visits, Linear Imaging Self-Scanning Sensor (LISS III), and Google Earth. Thereafter, eight landslide causative factors, viz. slope, aspect, curvature, elevation, Landuse Landcover (LULC), soil, lithology, and drainage density were performed by employing the weight of evidence (WOE), information value method (IVM) and the frequency ratio (FR) methods. Using the ArcGIS reclassification tool, the final landslide susceptibility zonation (LSZ) maps were categorized into five susceptibility zones: “very low, low, medium, high, and very high.” The success rate for WOE, FR, and IVM models was determined as 76.27%, 78.20%, and 76.09% respectively, depicting that the FR model based LSZ map is more accurate. According to the FR model based LSZ map, the highly susceptible classes in the case of aspect, curvature, and lithology are southeast, concave, and TBS, respectively. The sparsely vegetated areas are more prone to landslides than the other LULC areas. The higher elevations, ranging from 1191 to 1434, 1434–1655, and 1655–1876 m, are more vulnerable to landslide activities as compared to low elevations. The slope classes 30–45 and 45–60, medium drainage density class and fine loamy class of soils are more likely to landslides. The prepared susceptibility zone map can be used for future mitigation planning in the high and very high susceptibility zones in order to reduce landslide-related human and economic losses.

Generation of Homogeneous Slope Units Using a Novel Object-Oriented Multi-Resolution Segmentation Method

The generation of map units is a fundamental step for an appropriate assessment of landslide susceptibility. Recent studies have indicated that the terrain relief-based slope units perform better in homogeneity compared with the grid units. However, it is difficult at present to generate high-precision and high-matching slope units by traditional methods. The problem commonly concentrates in the plain areas without obvious terrain reliefs and the junction of sudden changes in terrain. In this paper, we propose a novel object-oriented segmentation method for generating homogeneous slope units. Herein, the multi-resolution segmentation algorithm in the image processing field is introduced, enabling the integration of terrain boundary conditions and image segmentation conditions in slope units. In order to illustrate the performances of the proposed method, Kitakyushu region in Japan is selected as a case study. The results show that the proposed method generates satisfactory slope units that satisfactorily reproduce the actual terrain relief, with the best within-unit and between-unit homogeneities compared with the previous methods, in particular at the plain areas. We also verify the effectiveness of the presented method through the sensitivity analysis using different resolutions of digital elevation models (DEMs) data of the region. It is reported that the presented approach is notably advanced in the requirements of the quality of DEM data, as the presented approach is less sensitive to DEM spatial resolution compared with other available methods.

Assessment of landslide behaviour in colluvium deposit at Doi Chang, Thailand

The paper presents the case study of the recurrent slope movement in colluvium deposits at Doi Chang, Thailand. A thorough site investigation confirmed the slope movement rate corresponding to slow creep during dry season, while in the rainy season, its velocity remarkably increased. Despite frequent repair, the movement rate was sufficient to result in the recurrent damage of infrastructures like roads and buildings, causing economic loss and public concerns. Furthermore, surface mapping revealed that the hill's topography led to the concentration of flowing water in a particular area. This resulted in a high level of groundwater table, especially during the rainy season. The inclinometer installed in that area suggested an average movement rate of 20.5 mm/month in the wet season. In contrast, during the dry season, it was limited within 2 mm/month, indicating that the increase in the rate of slope movement in the colluvium deposit was primarily due to the rising groundwater table. Field and laboratory tests were conducted to determine the properties of the colluvium deposit. Landslide susceptibility assessment was performed using infinite slope model and later integrated with GIS to evaluate the factor of safety (FS) over a large area. The FS decreased below 1 when the groundwater level rose to 0.3 cm below the ground surface, and using GIS, based on infinite slope model, the potential risk zone were delineated.

Spatiotemporal Landslide Susceptibility Mapping Incorporating the Effects of Heavy Rainfall: A Case Study of the Heavy Rainfall in August 2021 in Kitakyushu, Fukuoka, Japan

Landslide represents an increasing menace causing huge casualties and economic losses, and rainfall is a predominant factor inducing landslides. Landslide susceptibility assessment (LSA) is a commonly used and effective method to prevent landslide risk, however, the LSA does not analyze the impact of the rainfall on landslides which is significant and non-negligible. Therefore, the spatiotemporal LSA considering the inducing effect of rainfall is proposed to improve accuracy and applicability. In this study, the influencing factors are selected using the chi-square test, out-of-bag error and multicollinearity test. The spatial LSA are thus obtained using the random forest (RF) model, deep belief networks model and support vector machine, and compared using receiver operating characteristic curve and seed cell area index to determine the optimal assessment result. According to the heavy rainfall characteristics in the study area, the rainfall period is divided into four stages, and the effective rainfall model is employed to generate the rainfall impact (RI) maps of the four stages. The spatiotemporal LSAs are obtained by coupling the optimal spatial LSA and various RI maps and verified using the landslide warning map. The results demonstrate that the optimal spatiotemporal LSA is obtained using the spatial LSA of the RF model and temporal LSA of the rainfall data in the peak stage. It can predict the area where rainfall-induced landslides are likely to occur and prevent landslide risk.

Reducing Local Correlations Among Causal Factor Classifications as a Strategy to Improve Landslide Susceptibility Mapping

A landslide susceptibility map (LSM) is the basis of hazard and risk assessment, guiding land planning and utilization, early warning of disaster, etc. Researchers are often overly keen on hybridizing state-of-the-art models or exploring new mathematical susceptibility models to improve the accuracy of the susceptibility map in terms of a receiver operator characteristic curve. Correlation analysis of the causal factors is a necessary routine process before susceptibility modeling to ensure that the overall correlation among all factors is low. However, this overall correlation analysis is insufficient to detect a high local correlation among the causal factor classes. The objective of this study is to answer three questions: 1) Is there a high correlation between causal factors in some parts locally? 2) Does it affect the accuracy of landslide susceptibility assessment? and 3) How can this influence be eliminated? To this aim, Wanzhou County was taken as the test site, where landslide susceptibility assessment based on 12 causal factors has been previously performed using the frequency ratio (FR) model and random forest (RF) model. In this work, we conducted a local spatial correlation analysis of the “altitude” and “rivers” factors and found a sizeable spatial overlap between altitude-class-1 and rivers-class-1. The “altitude” and “rivers” factors were reclassified, and then the FR model and RF model were used to reevaluate the susceptibility and analyze the accuracy loss caused by the local spatial correlation of the two factors. The results demonstrated that the accuracy of LSMs was markedly enhanced after reclassification of “altitude” and “rivers,” especially for the RF model–based LSM. This research shed new light on the local correlation of causal factors arising from a particular geomorphology and their impact on susceptibility.