Regional Landslide Susceptibility Assessment Research Articles

A python system for regional landslide susceptibility assessment by integrating machine learning models and its application

Landslide susceptibility assessment is considered the first step in landslide risk assessment, but current studies mostly rely on GIS platforms or other software for data preprocessing. The modeling process is relatively complicated and multi-models cannot be integrated. With regard to this issue, this study develops a Python system for automatic assessment of regional landslide susceptibility. The Python system implements landslide susceptibility assessment through three modules: geographic data processing, machine learning modeling and result evaluation analysis. For geographic data processing, ten landslide influencing factors can be used to construct an evaluation factor dataset and reclassify the thematic maps based on the frequency ratio method. Four built-in machine learning models (logistic regression (LR), multi-layer perceptron (MLP), support vector machine (SVM) and extreme gradient boosting (XGBoost)) are integrated into the system to complete susceptibility modeling and calculation. Additionally, receiver operating characteristic (ROC) curves can be automatically generated to evaluate the accuracy. The system was then applied into Lantian County in Shaanxi Province as a demonstration example. The results show that the areas under the ROC curve (AUC) of the four models are 0.838 (LR)、0.882 (SVM)、0.809 (MLP) and 0.812 (XGBoost), respectively, indicating that the SVM model was the most suitable model for landslide susceptibility assessment in Lantian County in the Loess Plateau of China. The system has now been made open source on Github, which can effectively improve the efficiency of regional landslide susceptibility assessment, especially provide tools for data processing and modeling for non-professionals.

Regional landslide susceptibility assessment based on improved semi-supervised clustering and deep learning

Regional landslide susceptibility assessment based on improved semi-supervised clustering and deep learning

Large-Scale Landslide Susceptibility Mapping Using an Integrated Machine Learning Model: A Case Study in the Lvliang Mountains of China

Integration of different models may improve the performance of landslide susceptibility assessment, but few studies have tested it. The present study aims at exploring the way to integrating different models and comparing the results among integrated and individual models. Our objective is to answer this question: Will the integrated model have higher accuracy compared with individual model? The Lvliang mountains area, a landslide-prone area in China, was taken as the study area, and ten factors were considered in the influencing factors system. Three basic machine learning models (the back propagation (BP), support vector machine (SVM), and random forest (RF) models) were integrated by an objective function where the weight coefficients among different models were computed by the gray wolf optimization (GWO) algorithm. 80 and 20% of the landslide data were randomly selected as the training and testing samples, respectively, and different landslide susceptibility maps were generated based on the GIS platform. The results illustrated that the accuracy expressed by the area under the receiver operating characteristic curve (AUC) of the BP-SVM-RF integrated model was the highest (0.7898), which was better than that of the BP (0.6929), SVM (0.6582), RF (0.7258), BP-SVM (0.7360), BP-RF (0.7569), and SVM-RF models (0.7298). The experimental results authenticated the effectiveness of the BP-SVM-RF method, which can be a reliable model for the regional landslide susceptibility assessment of the study area. Moreover, the proposed procedure can be a good option to integrate different models to seek an “optimal” result.

A Comparison of an Adaptive Neuro-Fuzzy and Frequency Ratio Model to Landslide-Susceptibility Mapping along Forest Road Networks

In this research, we used the integration of frequency ratio and adaptive neuro-fuzzy modeling (ANFIS) to predict landslide susceptibility along forest road networks in the Hyrcanian Forest, northern Iran. We began our study by first mapping landslide locations during an extensive field survey. In addition, we then selected landslide-conditioning factors, such as slope, aspect, altitude, rainfall, geology, soil, road age, and slip position from the available Geographic Information System (GIS) data. Following this, we developed Adaptive Neuro-Fuzzy Inference System (ANFIS) models with two different membership functions (MFs) in order to generate landslide susceptibility maps. We applied a frequency ratio model to the landslide susceptibility mapping and compared the results with the probabilistic ANFIS model. Finally, we calculated map accuracy by evaluating receiver-operating characteristics (ROC). The validation results yielded 70.7% accuracy using the triangular MF model, 67.8% accuracy using the Gaussian MF model, and 68.8% accuracy using the frequency ratio model. Our results indicated that the ANFIS is an effective tool for regional landslide susceptibility assessment, and the maps produced in the study area can be used for natural hazard management in the landslide-prone area of the Hyrcanian region.

A new approach to assess landslide susceptibility based on slope failure mechanisms

Susceptibility assessments are crucial in identifying hazards in densely populated mountain areas featuring active internal and external dynamic geological processes. A comprehensive analysis of the interactions among ecological, hydrological, and geotechnical factors was performed to develop a quantitative method for assessing regional landslide susceptibility based on slope failure and landslide formation mechanisms. Using slope geomorphic units (SGUs) that can sufficiently describe the comprehensive features of landforms and hazard-forming factors related to landslides, an integrated model was built to analyze potential slip planes and slope stability by considering the different reinforcement effects of roots on soil, the root anchoring capacities of various vegetation types, and the corresponding vegetation weight loads. Using this SGU-based method, the occurrence possibility of regional-scale landslides was determined by analyzing the propagation directions and maximum moving distances of landslide materials in unstable areas. Taking the landslides in the Dadu River Basin, China, as a case study, a landslide susceptibility map was obtained and validated by conducting a field study and remote sensing interpretation of actual landslides; the assessment results were in accordance with the actual disaster situation. The distributions of zones with high or very high susceptibility (i.e., high-altitude valleys with steep slopes, abundant rainfall, and severe soil erosion) are closely correlated with the slope topography and stability. These findings suggest that the proposed assessment methodology can provide scientific support for preventing or mitigating landslide hazards and may serve as pertinent guidance for regional landslide susceptibility assessments in the Dadu River Basin and beyond.

Comparing Forward Conditional Analysis and Forward Logistic Regression Methods in a Landslide Susceptibility Assessment: A Case Study in Sicily

Forward logistic regression and conditional analysis have been compared to assess landslide susceptibility across the whole territory of the Sicilian region (about 25,000 km2) using previously existing data and a nested tiered approach. These approaches were aimed at singling out a statistical correlation between the spatial distribution of landslides that have affected the Sicilian region in the past, and a set of controlling factors: outcropping lithology, rainfall, landform classification, soil use, and steepness. The landslide inventory used the proposal of building the models like the official one obtained in the PAI (hydro geologic asset plan) project, amounting to more than 33,000 events. The 11 types featured in PAI were grouped into 4 macro-typologies, depending on the inherent conditions believed to generate various kinds of failures and their kinematic evolution. The study has confirmed that it is possible to carry out a regional landslide susceptibility assessment based solely on existing data (i.e., factor maps and the landslide archive), saving a considerable amount of time and money. For scarp landslides, where the selected factors (steepness, landform classification, and lithology) are more discriminate, models show excellent performance: areas under receiver operating characteristic (ROC) (AUCs) average > 0.9, while hillslope landslide results are highly satisfactory (average AUCs of about 0.8). The stochastic approach makes it possible to classify the Sicilian territory depending on its propensity to landslides in order to identify those municipalities which are most susceptible at this level of study, and are potentially worthy of more specific studies, as required by European-level protocols.

The added value of a regional landslide susceptibility assessment: The western branch of the East African Rift

Predicting landslide occurrence is of key importance for understanding the geomorphological development of mountain environments as well as to assess the potential risk posed by landsliding to human societies in such environments. Global landslide susceptibility models use a generic model formulation to predict landslide susceptibility anywhere on the planet from openly available data. Regional models, on the contrary, use local information on landslide occurrence to constrain model parameters and may also benefit from better spatial information with respect to controlling factors. This study aims to investigate the added value of the construction of regional landslide susceptibility models (versus global and continental models) in the western branch of the East African Rift, a data-scarce landslide-prone tropical environment. First, a comprehensive landslide database containing 6446 instances was compiled for the study area using Google Earth imagery. Second, three regional data-driven landslide susceptibility models were developed. Third, the efforts to construct these regional models were quantified by analysing how their quality is impacted by (1) the use of more accurate, regional peak ground acceleration and geology data, and (2) an increasing inventory size. Fourth, regional and global/continental models were compared in terms of predictive power and geomorphological plausibility. We observe that global/continental landslide susceptibility models are capable of identifying landslide-prone areas, but lack prediction power and geomorphological plausibility when compared to our regional models. Importantly, this difference in quality is not driven by the use of more accurate and detailed geology and peak ground acceleration data, but rather by the use of a detailed regional landslide inventory to calibrate the models. We also show that the model quality only increases marginally beyond a certain inventory size. We conclude that the regional landslide susceptibility assessment does provide an added value compared to existing global models in terms of geomorphological plausibility and model performance, whereby the largest gain is to be found in the construction of a regional landslide inventory, rather than the investment in more detailed covariates or the application of more complex modelling techniques. The latter suggests that the role of controlling variables depends, to some extent, on the regional context: making adequate susceptibility predictions proves difficult when local conditions are not accounted for by means of a regional inventory.

Preliminary regional landslide susceptibility assessment using limited data

In this paper a heuristic approach for preliminary regional landslide susceptibility assessment using limited amount of data is presented. It is called arbitrary polynomial method and takes into account 5 landslide conditioning parameters: lithology, slope inclination, average annual rainfall, land use and maximum expected seismic intensity. According to the method, in the first stage, a gradation is performed for each of the carefully selected conditioning parameters by assigning so called rating points to the grid cells on which the region is divided. Values of the rating points vary from 0 to 3 and depend on the parameter’s character and importance for landslide development within the region of interest. A so called Total Landslide Susceptibility Rating (TLSR) model is obtained by summing the individual rating points of each parameter and dividing the region into five susceptibility zones according to Jenks natural breaks classification. Verification of the TLSR model is then performed by overlaying the landslide inventory map of the selected region over the prepared susceptibility map. The sensitivity of the model can be additionally tested by multiplying the conditioning parameter’s rating points by sensitivity coefficients. In this way, additional landslide susceptibility models are obtained, named Weighted Total Landslide Susceptibility Rating (WTLSR) models. As a practical example of the method, two TLSR models are presented here for the Polog region in Republic of Macedonia, for return periods of maximum expected seismic intensity for 100 and 500 years. With over 74% of mapped landslides falling in zones of high and very high susceptibility, the results are considered satisfactory for regional scale landslide modelling and are comparable with more advanced quantitative methods. Additional WTLSR models were prepared, and their correlation identified the best model. The presented approach is considered to be very convenient for conducting preliminary regional landslide susceptibility assessments with the ability to fine-tune the results. Due to its simplicity, it can be applied to additional landslide conditioning parameters other than the one presented in the paper, depending on the region of interest and available data sources. It is especially practical for use in developing countries, where various organizational, technical and economic constraints prevent application of more advanced data driven methods. Limitations and restrictions of the approach are also discussed.

Regional landslide susceptibility assessment using multi-stage remote sensing data along the coastal range highway in northeastern Taiwan

Typhoons Megi (2010) and Saola (2012) brought torrential rainfall which triggered regional landslides and flooding hazards along Provincial Highway No. 9 in northeastern Taiwan. To reduce property loss and saving lives, this study combines multi-hazard susceptibility assessment with environmental geology map a rock mass rating system (RMR), remote sensing analysis, and micro-topography interpretation to develop an integrated landslide hazard assessment approach and reflect the intrinsic state of slopeland from the past toward the future. First, the degree of hazard as indicated by historical landslides was used to determine many landslide regions in the past. Secondly, geo-mechanical classification of rock outcroppings was performed by in-situ investigation along the vulnerable road sections. Finally, a high-resolution digital elevation model was extracted from airborne LiDAR and multi-temporal remote sensing images which was analyzed to discover possible catastrophic landslide hotspot shortly. The results of the analysis showed that 37% of the road sections in the study area were highly susceptible to landslide hazards. The spatial distribution of the road sections revealed that those characterized by high susceptibility were located near the boundaries of fault zones and in areas of lithologic dissimilarity. Headward erosion of gullies and concave-shaped topographic features had an adverse effect and was the dominant factor triggering landslides. Regional landslide reactivation on this coastal highway are almost related to the past landslide region based on hazard statistics. The final results of field validation demonstrated that an accuracy of 91% could be achieved for forecasting geohazard followed by intense rainfall events and typhoons.

Physically Based Susceptibility Assessment of Rainfall-Induced Shallow Landslides Using a Fuzzy Point Estimate Method

The physically based model has been widely used in rainfall-induced shallow landslide susceptibility analysis because of its capacity to reproduce the physical processes governing landslide occurrence and a higher predictive capability. However, one of the difficulties in applying the physically based model is that uncertainties arising from spatial variability, measurement errors, and incomplete information apply to the input parameters and analysis procedure. Uncertainties have been recognized as an important cause of mismatch between predicted and observed distributions of landslide occurrence. Therefore, probabilistic analysis has been used to quantify the uncertainties. However, some uncertainties, because of incomplete information, cannot be managed satisfactorily using a probabilistic approach. Fuzzy set theory is applicable in this case. In this study, in order to handle uncertainty propagation through a physical model, fuzzy set theory, coupled with the vertex method and the point estimate method, was adopted for regional landslide susceptibility assessment. The proposed approach was used to evaluate susceptibility to rainfall-induced shallow landslides for a regional study area, and the analysis results were compared with landslide inventory to evaluate the performance of the proposed approach. The AUC values arising from the landslide susceptibility analyses using the proposed approach and probabilistic analysis were 0.734 and 0.736, respectively. However, when the COV values of the input parameters were reduced, the AUC values of the proposed approach and the probabilistic analysis were reduced to 0.722 and 0.688, respectively. It means that the performance of the fuzzy approach is similar to that of probabilistic analysis but is more robust against variation of input parameters. Thus, at catchment scale, the fuzzy approach can respond appropriately to the uncertainties inherent in physically based landslide susceptibility analysis, and is especially advantageous when the amount of quality data is very limited.

An integrative approach for regional landslide susceptibility assessment using weight of evidence method: a case study of Yom River Basin, Phrae Province, Northern Thailand

A landslide susceptibility assessment was conducted for the Yom River catchment, North Thailand, using a weight of evidence approach. In total, 1630 landslide events were detected using remote sensing techniques. An integrated workflow based on robust statistical threshold criteria was applied to reveal the landslide-controlling factors that can be utilised on a regional scale. Initially, approximately 15 different factors were considered within this study. After a sensitivity and plausibility analysis, a final susceptibility map was generated based on the three most essential factors, which fulfilled the statistical requirements of independence, prediction power and plausibility. The final map was subdivided into five susceptibility zones using quantitative classification. The map provides a suitable and reliable starting point for further detailed mass movement analysis in the Yom River Basin and can be used to support strategic spatial planning measures on a regional scale.

Application of analytical hierarchy process and least-squares method for landslide susceptibility assessment along the Zhong-Wu natural gas pipeline, China

As one of the major problems of geo-engineering, landslides often influence the safety of linear engineering projects that cross mountainous areas. Therefore, when selecting suitable routes for such projects, it is important to assess their susceptibility to landslides. In this paper, we used a natural gas pipeline in the northeast of the Yunnan-Guizhou Plateau of China as a case study to analyze landslide susceptibility. Based on engineering geological analogy, the analytical hierarchy process, and the least-squares method, a regional landslide susceptibility assessment model was developed and was programmed using GIS ArcEngine components under the Visual Studio.NET environment. The landslide susceptibility along the Zhong-Wu natural gas pipeline from Zhongxian County to Wuhan was assessed based on this model and classified into five levels: very safe, safe, moderate, susceptible, and very susceptible. The high accuracy and prediction capability of the model were confirmed by comparing the model results with past landslide data and performing a prediction test. The results indicated that the assessment model used in this study is reliable and can be used for landslide susceptibility assessment and route selection in other areas.

Application of a neuro-fuzzy model to landslide-susceptibility mapping for shallow landslides in a tropical hilly area

This paper presents landslide-susceptibility mapping using an adaptive neuro-fuzzy inference system (ANFIS) using a geographic information system (GIS) environment. In the first stage, landslide locations from the study area were identified by interpreting aerial photographs and supported by an extensive field survey. In the second stage, landslide-related conditioning factors such as altitude, slope angle, plan curvature, distance to drainage, distance to road, soil texture and stream power index (SPI) were extracted from the topographic and soil maps. Then, landslide-susceptible areas were analyzed by the ANFIS approach and mapped using landslide-conditioning factors. In particular, various membership functions (MFs) were applied for the landslide-susceptibility mapping and their results were compared with the field-verified landslide locations. Additionally, the receiver operating characteristics (ROC) curve for all landslide susceptibility maps were drawn and the areas under curve values were calculated. The ROC curve technique is based on the plotting of model sensitivity — true positive fraction values calculated for different threshold values, versus model specificity — true negative fraction values, on a graph. Landslide test locations that were not used during the ANFIS modeling purpose were used to validate the landslide susceptibility maps. The validation results revealed that the susceptibility maps constructed by the ANFIS predictive models using triangular, trapezoidal, generalized bell and polynomial MFs produced reasonable results (84.39%), which can be used for preliminary land-use planning. Finally, the authors concluded that ANFIS is a very useful and an effective tool in regional landslide susceptibility assessment.

Manifestation of an adaptive neuro-fuzzy model on landslide susceptibility mapping: Klang valley, Malaysia

The purpose of the present paper is to manifest the results of the neuro-fuzzy model using remote sensing data and GIS for landslide susceptibility analysis in a part of the Klang Valley areas i Malaysia. Landslide locations in the study area were identified by interpreting aerial photographs and satellite images, supported by extensive field surveys. SPOT 5 satellite imagery was used to map vegetation index. Maps of topography, lineaments, NDVI and land cover were constructed from the spatial datasets. Seven landslide conditioning factors such as altitude, slope angle, plan curvature, distance from drainage, soil type, distance from faults and NDVI were extracted from the spatial database. These factors were analyzed using a neuro-fuzzy model (adaptive neuro-fuzzy inference system, ANFIS) to construct the landslide susceptibility maps. During the model development works, total 5 landslide susceptibility models were obtained by using ANFIS results. For verification, the results of the analyses were then compared with the field-verified landslide locations. Additionally, the ROC curves for all landslide susceptibility models were drawn and the area under curve values was calculated. Landslide locations were used to validate results of the landslide susceptibility map and the verification results showed 98% accuracy for the model 5 employing all parameters produced in the present study as the landslide conditioning factors. The validation results showed sufficient agreement between the obtained susceptibility map and the existing data on landslide areas. Qualitatively, the model yields reasonable results which can be used for preliminary landuse planning purposes. As a conclusion, the ANFIS is a very useful tool for regional landslide susceptibility assessments.

Assessment of landslide susceptibility for a landslide-prone area (north of Yenice, NW Turkey) by fuzzy approach

Regional landslide susceptibility assessments pose complex problems. To solve these problems, numerous approaches, such as statistical analysis, geotechnical engineering approach, geomorphologic approach and fuzzy logic, have been employed. However, all the available methods for regional landslide susceptibility assessments have some uncertainties due to a lack of knowledge and variability. Minimizing these uncertainties provides realistic approaches. Use of the fuzzy logic approach to produce a landslide susceptibility map of a landslide-prone area in NW Turkey is the main purpose of the present study. For this purpose, the study includes five main stages, these being the preparation of a landslide inventory of the study area, the application of factor analysis, the extraction of fuzzy if-then rules, the use of a geographical information system, and the control of the reliability of the resulting landslide susceptibility map. Slope angle, slope aspect, land use, weathering depth, water conditions and topographical elevation were considered as landslide conditioning factors for the study area. A total of 23 if-then rules was extracted from the field data. Employing these rules, fuzzified index maps representing each parameter were obtained. Finally, combining these maps, the landslide susceptibility map of the area was prepared. When compared with the landslide susceptibility map, the landslides identified in the area were found to be located in the very high- and high-susceptibility zones. As far as the performance of the fuzzy approach for processing is concerned, the images appear to be quite satisfactory, the zones determined on the map being zones of relative susceptibility.