High Susceptibility Zones Research Articles

Flood risk assessment under the shared socioeconomic pathways: a case of electricity bulk supply points in Greater Accra, Ghana

This study evaluates flood susceptibility and risk on Bulk Supply Points in the Greater Accra region (GAR) using a Frequency Ratio model based on 15 flood conditioning factors. The model explores the influence of natural, meteorological and anthropogenic factors on flooding occurrences under the Shared Socioeconomic Pathway (SSP) scenarios and assesses flood risks at Bulk Supply Points (BSPs). Flood susceptibility mapping was conducted for both current and future periods under various SSP scenarios. Results reveal that elevation, slope, soil type, distance from urban areas, and SPI are the most influential factors contributing to flooding susceptibility in the region. The current flood map, about 37% of the total area of GAR categorized under the moderate flood-susceptible zone category followed by about 30% categorized under the low flood-vulnerable zone. However, about 16% was categorized under the very high flood-vulnerable zone. The study projects increasing flood susceptibility under the SSP scenarios with intensification under SSP2 and SSP3 scenarios. For instance, the areas categorized as high and very high flood susceptibility zones are projected to expand to approximately 32% and 26% each by 2055 under SSP3. The study also assesses flood risks at Bulk Supply Points (BSPs), highlighting the escalating susceptibility of power assets to flooding under different scenarios. For instance, in the very high scenario, flooding is estimated to reach 640 h in 2045 and exceed 800 h in 2055—more than double the 2020 baseline. The analysis shows the bulk supply points face increasing flood susceptibility, with risks escalating most sharply under the severe climate change SSP3 and SSP5 scenarios. Over 75% of BSPs are expected to fall in the low- to medium-risk categories across SSPs while more than 50% of BSPs are within medium- to high-risk categories in all scenarios except SSP1, reflecting the impact of climate change. SSP3 and SSP5 stand out with over 60% of BSPs facing high or very high flooding risks by 2055. It indicates moderate resilience with proper adaptation but highlights potential disruptions in critical infrastructure, such as BSPs, during persistent flooding. The findings of the study are expected to inform Ghana’s contributions towards addressing Sustainable Development Goals (SDGs) 7, 11 and 13 in Ghana.

Assessment of landslide susceptibility in the Himalayan state of Tripura, India, using a Multi-Model Approach

Landslides are the down slope mass movement of soil, rocks, and debris due to a natural or human activities resulting in widespread hazard events in India. The most affected areas comprise 15 percent of its landmass which includes Tripura and eleven Himalayan states and parts of the Western and Eastern Ghats in India. In Tripura, landslides cause road blockage and destruction of settlements, bringing economic and life losses in every year. Thus, this research is focused on identifying landslide susceptible zones and the significant causative factors behind landslides. Assessment of Landslide Susceptibility (LS) identifies fifteen major causative factors under five broad groups; topographic, geotechnical, hydrological, environmental, and anthropogenic. With application of Analytical Hierarchical Process (AHP), Frequency Ratio (FR), and Random Forest (RF)-based models were performed to extract landslide susceptible zonation map for Tripura. This study reveals that the successive hill ranges formed by young sedimentary lithologic formations associated with deforestation, heavy rainfall during monsoon, and anthropogenic activities (road constructions and jhumming) are the responsible geo-conditions for triggering landslides. In this study, while the AHP and FR model show only 1.95% and 11.46% confined along the hilltop of Jampui, Sakhan, and Longtarai, the RF model designated Tripura’s 30% land area as high and very high landslide susceptible zones (LSZ), predominantly over hills, foothills, and low laying undulating land (tillas). For the accuracy assessment, the ROC curve is used, which shows that RF model appears to be the maximum accurate (0.810) one, followed by FR (0.806) and AHP (0.744).

Fatal landslides in Kencho, Shacha & Gozdi villages, Gofa zone, Ethiopia: A detailed investigation (Geological, Geotechnical, geophysical & geospatial) of the July 22, 2024 catastrophe and its socioeconomic repercussions

A landslide is one of the geological hazards that cause the most disaster in densely populated areas. The landslide that occurred in Gezie Gofa woreda, Gofa Zone, Kencho Shacha Gozdi village, killed more than 250 people. Two landslides occurred on July 22, 2024, at 8:30 and 10:40 AM The first landslide killed six people and demolished three houses. The second landslide killed more than 245 people, including those who came to the site to excavate the buried bodies during the first landslide. The primary objective of this research is to evaluate the landslide causative factors, model the landslide susceptibility, and characterize the landslide disaster on socioeconomic effects that occurred on July 22, 2024. The landslide inventory data, field surveys, laboratory analyses, and various geophysical surveys characterized the current and past landslides of the area. The landslide susceptibility model was modeled using a statistical approach in the GIS. environment. The socioeconomic effects were assessed using field surveys and systematic interviews with the victims. The conditioning factors selected for landslide susceptibility modeling are lithology, geological structures, groundwater, slope, land use or land cover, aspect, curvature, and elevation. The major triggering factor of the landslide in the area was heavy rainfall, which occurred on July 21–22, 2024, between 3:00 a.m.-8:30 a.m. The results reveal that the significant conditioning factors of the landslide in the study area are geological structures (both visible and inferred), groundwater, slopes, and human activities. The characterized socioeconomic effects include the destruction of agricultural land, the demolishing of houses, and the loss of human lives, as well as several people evacuated and sheltered under tents and churches. However, the most momentous disaster in the area is the loss of human life. Based on the research results, it would be better to relocate those people living in the high landslide susceptible zones, and all high landslide-prone areas and mountainous terrain in southern Ethiopia should be mapped, and the people should be aware of the landslide risk areas.

Landslide susceptibility mapping using frequency ratio and analytical hierarchy process method in Awabel Woreda, Ethiopia

A landslide is a serious geo-environmental problem that results in the death of life and the devastation of infrastructure, properties, and agricultural lands. This research aimed to identify landslide susceptibility zones in selected Kebels of Awabel Woreda, central Ethiopia. Frequency ratio (FR) and analytical hierarchy process (AHP) methods were used. 175 landslide inventory data collected from Google Earth and field data were collected for testing and training data sets. Using the analytical hierarchy process, all the thematic layers (stream distance, slope, aspect, rainfall, lineament density, elevation, lithology, soil, land use/land cover, and curvature) were reclassified and weighted based on their relative contribution to landslide occurrence with the help of experts’ knowledge. The results show that 11.85% and 20.52 % of the study fall under the very high and high susceptible zones, respectively, while the low susceptible zones cover 26.3% and 14.74% of the area. The landslide susceptibility zone identified by the frequency ratio model shows that (6.09%) and (16.9%) of the area covered very high and high susceptible zones, respectively, while 30.4% and 23.4% of the area covered low and very susceptible zones, respectively. The predicted landslide-susceptible areas were validated using existing landslide points with the help of the ROC tool in ArcGIS. Area under the curve (AUC) results were 84.5% for the AHP model and 73% for the frequency ratio model. The find of this study will provide an input for decision makers and land use planners in the future.

GIS based Soil Erosion Susceptibility Mapping in Middle Himalayas using MCDM: A Case Study of Rajouri District.

The Pirpanjal range, nestled within the Himalayan region, confronts substantial soil erosion challenges owing to its diverse topography and the inherent instability of geological formations. This study focuses on the Rajouri region, to assess soil erosion susceptibility and spatially prioritize vulnerable zones. Utilizing a Geographical Information System (GIS)-based approach, we employed the Analytic Hierarchy Process (AHP) and the Weighted Sum Method (WSM). Various datasets, including precipitation records, geological maps, soil maps, and satellite imagery, were incorporated to derive eleven critical factors. These factors encompassed topographical derivatives; land use and land cover (LULC), soil properties, drainage patterns, rainfall data, lithological characteristics, wetness index, and the vegetative health of the area. The methodology employed yielded unbiased and reliable ratings and weightages, as substantiated by a Consistency Ratio (CR) of 0.095. The results reveal that 41% of the total area in the study region is exceedingly vulnerable to soil erosion. Slope variations range from 0 to 67.14 degrees, delineating high and very high susceptibility zones spanning 531.79 km² (19.82%) and 316.20 km² (11.78%) of the area, respectively. Additionally, assessments based on the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI) underscore the severity of soil erosion, covering 40% and 25% of the highly susceptible zones, respectively. High drainage density and curvature zones were identified in 13% and 31% of the study area, respectively. This comprehensive study contributes valuable insights for the planning and implementation of effective soil conservation measures in the Rajouri region.

Refined Landslide Susceptibility Mapping Considering Land Use Changes and InSAR Deformation: A Case Study of Yulin City, Guangxi

Landslide susceptibility maps (LSMs) are valuable tools typically used by local authorities for land use management and planning activities, supporting decision-makers in urban and infrastructure planning. To address this, we proposed a refined method for landslide susceptibility assessment, which comprehensively considered both static and dynamic factors. Neural network methods were used for susceptibility analysis. Land use and land cover (LULC) change and InSAR deformation were then integrated into the traditional susceptibility zoning to obtain a refined susceptibility map with higher accuracy. Validation was conducted on the improved landslide susceptibility map using site landslide data. The results showed that the LULC were proven to be the core driving factors for landslide occurrence in the study area. The GRU model achieved the highest model performance (AUC = 0.886). The introduction of InSAR surface deformation and land use and land cover change data could rationalize the inappropriateness of traditional landslide susceptibility zoning, correcting the false positive and false negative areas in the traditional landslide susceptibility map caused by human activities. Ultimately, 12.25% of the study area was in high-susceptibility zones, with 3.10% of false positive and 0.74% of false negative areas being corrected. The proposed method enabled refined analysis of landslide susceptibility over large areas, providing technical support and disaster prevention and mitigation references for geological hazard susceptibility assessment and land management planning.

Freeze-thaw landslide susceptibility assessment and its future development on the seasonally frozen ground of the Qinghai-Tibet Plateau under warming-humidifying climate

Exploring freeze-thaw landslide susceptibility on the Qinghai-Tibet Plateau (QTP) under warming-humidifying climate is greatly important for preventing and mitigating the risks of landslide hazards on engineering facilities. This study proposed a random forest-based freeze-thaw landslide susceptibility assessment model, where annual rainfall, annual average air temperature (AAAT), slope gradient, normalized difference vegetation index (NDVI), elevation, lithology, and plan curvature were fully considered. Selecting a study area of 324 km2 on the seasonally frozen ground (SFG) of QTP with 1059 freeze-thaw landslides, the model accuracy was validated. Low, moderate, high, and very high susceptibility zones were precisely classified, which accounted for 27.0, 27.5, 28.3, and 17.2%, respectively. Furthermore, its future development was explored under warming, humidifying, and warming-humidifying climates. Results indicated that when the AAAT or annual rainfall increased by 1.16 °C or 20 mm, both high and very high susceptibility zones increased by 2.0 or 1.0%, respectively. When AAAT and annual rainfall simultaneously increased by 1.16 °C and 20 mm, a higher increase in the high and very high susceptibility zones of 2.8% occurred. It was noteworthy that climate warming transitioned low and moderate susceptibility zones into high and very high susceptibility zones. These areas where freeze-thaw landslide susceptibility changed featured the AAAT of 4.29–6.15 °C, annual rainfall of 528.9–552.3 mm, slope gradient of 16–25°, and elevation of 3750-3940 m. Compared to climate warming, the humidifying climate and warming-humidifying climate expanded moderate susceptibility zones, and areas where freeze-thaw landslide susceptibility changed featured the gentler slope gradients of 8–16°. This study can provide a better guidance for safe engineering constructions influenced by freeze-thaw landslides on the QTP.

Landslide susceptibility and building exposure assessment using machine learning models and geospatial analysis techniques

Landslides are among the most dangerous hazards in Asia, posing a significant threat to human lives, infrastructure, and sustainable development. Landslide susceptibility maps provide useful insights into hazard potential but lack quantitative exposure assessments to develop targeted mitigation strategies and resource allocation. This study aims to propose an integrated approach for landslide hazards and exposure evaluation using machine learning models on the Google Earth Engine environment and geospatial analysis techniques. A geospatial database was established to predict hazards and evaluate exposure, including data on topography, geology, hydrology, climate features, land use, and building data. The landslide susceptibility map was created with advanced machine-learning algorithms, including Classification And Regression Tree (CART), Random Forest (RF), Gradient Boosting (GB), and Support Vector Machine (SVM). The ROC curve and the Wilcoxon signed-rank test were employed to evaluate the performance differences among the CART, GB, RF, and SVM models. The results indicated that the RF model demonstrated the highest performance, leading to its selection for creating a landslide susceptibility map. Landslide exposure was evaluated by overlaying the landslide susceptibility map with building data to quantify the number of affected houses by landslides across districts and communes. The analysis results identified the Cho Don district as the most exposed, with 46,237 households located in high and very high landslide susceptibility zones, followed by Ba Be district (39,631 households), Bac Kan city (37,266 households), Bach Thong district (28,495 households), Cho Moi district (28,436 households), Na Ri district (17,723 households), Ngan Son district (14,142 households), and Pac Nam district (13,034 households). These findings enable the identification of the potential consequences of landslides on infrastructure, human settlements, and livelihoods, contributing to the promotion of disaster reduction and prevention strategies.

Flood susceptibility assessment of Jhelum River Basin: A comparative study of TOPSIS, VIKOR and EDAS methods

Floods are frequent natural hazards that cause widespread destruction, particularly in low-elevated areas. This study focuses on identifying flood susceptible zones in the Kashmir Valley, known for historical flooding attributed to the overflow of the Jhelum River. Various Multi-Criteria Decision Making (MCDM) techniques, including Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), Vise Kriterijumska Optimizacijai Compromission Resenje (VIKOR), and Evaluation Based on Distance from Average Solution (EDAS), were employed in this research. A total of 17 multidimensional factors were considered, and multicollinearity tests revealed no correlation among these factors. The results of the MCDM models indicate that areas along the Jhelum River are classified under very high flood susceptible zone. Specifically, Srinagar city is consistently classified under very high flood susceptible zone by all three models. Approximately 4.27 %, 9.67 %, and 5.39 % of the total area were identified as very high susceptible areas by TOPSIS, VIKOR, and EDAS, respectively. The models exhibited robust performance, as evidenced by the Area Under the Curve of the Receiver Operating Characteristics curve (AUC-ROC). Notably, VIKOR demonstrated excellent performance among the three models in generating flood susceptible maps. The favorable outcomes of these models underscore their potential application in similar regions facing comparable challenges. This study carries significant implications for policymakers, administrators, and local authorities involved in flood management within the Kashmir Valley. The insights provided can inform proactive measures and strategies to mitigate the impact of floods and enhance the overall resilience of the region.

Landslide Susceptibility Mapping Using Weight of Evidence Method in the Purwanegara District and Surrounding Areas, Banjarnegara Regency, Central Java Province, Indonesia

Abstract The research location is in the Purwanegara Subdistrict, Banjarnegara Regency, Province of Central Java. According to the Geological Agency of Indonesia, this location is a landslide-prone area, and many landslides occur yearly. Therefore, it needs a landslide susceptibility map is basic information for landslide mitigation in this area. The parameters used in this research are slope, lithology, distance from structural geology, distance from the river, land use, and rainfall. Using six parameters, the weight of evidence (WoE) method was used to develop a landslide susceptibility map. This research aims to make a zonation of landslide susceptibility in this area. The data were collected by field survey and visiting some related institutions. The landslide data used are 85 landslide events in the last three years. All data was then randomly divided into two groups: 70% as model data sets and 30% as validation data sets. The map of every parameter will be overlaid to create a landslide susceptibility zone. The Area Under Curve (AUC) method validated the final map. The result shows that landslide susceptibility zonation can be divided into four zones: very low susceptibility zone, low susceptibility zone, medium susceptibility zone, and high susceptibility zone. The very low susceptibility zone is mostly located in the north, the low susceptibility zone is mostly in the north and south, the moderate susceptibility zone is in the middle, and the high susceptibility zone is in the middle of the research’s location. The validation results of the success and prediction rates using the AUC method show an accuracy value of 0.86 and 0.82, respectively. According to the validation result, both can be classified as very good.

Landslide susceptibility mapping in East Ungaran, Indonesia: A comparative study using statistical methods

East Ungaran, is one of landslide prone areas in Semarang Regency, Indonesia. In addition to provide a more detail map of landslide susceptibility, the objective of this research was to compare performance of three widely used methods, which are the Weight of Evidence (WoE), Logistic Regression (LR) and combined Weight of Evidence (WoE) – Logistic Regression (LR), for landslide susceptibility mapping. Slope, elevation, lithology, land use, normalized difference vegetation index (NDVI), distance from lineament, distance from river, and distance from road were considered as landslide controlling parameters in the research area and were used as input variables in the landslide susceptibility zonation. The results showed that the slope, elevation, and distance from the road are significant parameters causing the landslides. The research area is divided into very low, low, moderate, and high landslide susceptibility zones. The WoE performs better than the LR, while the combined WoE-LR method performs the best among the three methods in predicting landslide susceptibility in this area. The landslide susceptibility map developed using the combined WoE-LR method is suggested to be used for landslide mitigation planning of this area.

Landslide susceptibility mapping using frequency ratio method along the Bhaderwah-Bani Road, Jammu and Kashmir, India

This study outlines a method for landslide susceptibility mapping utilizing remote sensing and geographic information systems (GIS). Since the terrain from Bhaderwah to Bani in J&K is prone to landslide threats, the study has been conducted there. The purpose of the current study is to determine and identify the significant terrain elements causing landslides. Nine contributing factors on landslides were taken into account throughout the analysis. Thematic data layers are created in the GIS domain based on local events. Topographic maps, satellite images, on-site observations and publicly available published maps are used to gather the landscape data. To determine the standardized scores of criteria expressing their factor of importance for a given decision problem in terms of thematic parameters, categories, and their normalized weights the Frequency ratio (FR) method is used to digitize these maps along with tabular data and to create a GIS database. A particular landslide susceptibility zonation (LSZ) map was created on a GIS platform by statistically combining weightages from several thematic maps. Using FR, the LSZ map was divided into five separate susceptibility zones. According to the study's finding, 8% of the area is in a very high susceptibility zone, with the remaining 12%, 15%, 23%, and 42% falling into high, moderate, low, and very low susceptibility zones, respectively. The outcomes of the current study might aid policy and decision makers in moving forward with regional development initiatives including land use planning.

Susceptibility Modeling and Potential Risk Analysis of Thermokarst Hazard in Qinghai–Tibet Plateau Permafrost Landscapes Using a New Interpretable Ensemble Learning Method

Climate change is causing permafrost in the Qinghai–Tibet Plateau to degrade, triggering thermokarst hazards and impacting the environment. Despite their ecological importance, the distribution and risks of thermokarst lakes are not well understood due to complex influencing factors. In this study, we introduced a new interpretable ensemble learning method designed to improve the global and local interpretation of susceptibility assessments for thermokarst lakes. Our primary aim was to offer scientific support for precisely evaluating areas prone to thermokarst lake formation. In the thermokarst lake susceptibility assessment, we identified ten conditioning factors related to the formation and distribution of thermokarst lakes. In this highly accurate stacking model, the primary learning units were the random forest (RF), extremely randomized trees (EXTs), extreme gradient boosting (XGBoost), and categorical boosting (CatBoost) algorithms. Meanwhile, gradient boosted decision trees (GBDTs) were employed as the secondary learning unit. Based on the stacking model, we assessed thermokarst lake susceptibility and validated accuracy through six evaluation indices. We examined the interpretability of the stacking model using three interpretation methods: accumulated local effects (ALE), local interpretable model-agnostic explanations (LIME), and Shapley additive explanations (SHAP). The results showed that the ensemble learning stacking model demonstrated superior performance and the highest prediction accuracy. Approximately 91.20% of the total thermokarst hazard points fell within the high and very high susceptible areas, encompassing 20.08% of the permafrost expanse in the QTP. The conclusive findings revealed that slope, elevation, the topographic wetness index (TWI), and precipitation were the primary factors influencing the assessment of thermokarst lake susceptibility. This comprehensive analysis extends to the broader impacts of thermokarst hazards, with the identified high and very high susceptibility zones affecting significant stretches of railway and highway infrastructure, substantial soil organic carbon reserves, and vast alpine grasslands. This interpretable ensemble learning model, which exhibits high accuracy, offers substantial practical significance for project route selection, construction, and operation in the QTP.

GIS-based statistical analysis for landslide susceptibility evaluation and zonation mapping: A case from Blue Nile Gorge, Gohatsion-Dejen road corridor, Central Ethiopia

Globally, landslides in mountainous regions are considered to be major geo-hazards which are frequent, intensive and devastating that result in extensive destruction to people's lives, land, property and other infrastructure. To mitigate landslide-associated risks, identification, evaluation, and delineation of susceptible areas is essential for proper strategic planning and management. The present study was conducted in Blue Nile Gorge in the highlands of Central Ethiopia, about 185 km from Addis-Ababa, along the Gohatsion-Dejen road corridor. Landslides are frequently reported in the study area, particularly during the rainy season. The objective of the present study was to evaluate landslide susceptibility and to prepare a landslide susceptibility zoning (LSZ) map. For this purpose a statistical information value (IV) model was used that took into account various factors including slope material, elevation, slope, aspect, land-use and land-cover, proximity to the road, and proximity to streams. Each of these factors was classified based on different conditions like landslide concentration, topography, geology, and land-use and land-cover types. The landslide inventory mapping was conducted through field observations and Google Earth image interpretation. Later, the influence of each causative factor on past landslides was used to calculate their respective IVs. The statistical IVs were further used to evaluate the landslide susceptibility. The LSZ map produced from this study shows that 22.3 % of the study area falls into a very high susceptibility zone, 23.1 % into a high, 18.8 % in moderate, 23.1 % into a low, and the remaining 12.8 % into a low susceptibility zone. The LSZ map was validated using the Receiver Operating Characteristics Curve (ROC), landslide density index (LDI), and past landslide data. All validation methods showed acceptable results. The prepared LSZ map can be used for safe land-use planning of the area.

Data driven forest fire susceptibility mapping in Bangladesh

Forests are essential natural resources that serve to facilitate economic activity while also providing an essential impact on climate regulation and the carbon cycle. In the Chittagong hill tracts, forest fires significantly contribute to the eradication of forests and the depletion of biodiversity. As a result, addressing the forest fire hazard is crucial for mitigating its potential negative effects. By implementing machine learning (ML) and deep learning (DL) algorithms, the primary aim of this research is to determine which areas within the CHT are most prone to forest fires. Artificial Neural Network (ANN), Classification and Regression Trees (CART), Categorical Boosting (Catboost), Convolutional Neural Network (CNN), Deep Neural Network (DNN), Gradient Boosting Decision Tree (GBDT), Random Forest (RF), Support Vector Machine (SVM), eXtreme Gradient Boosting (XGboost) are among these algorithms. This research investigates 267 past forest fire incidents while incorporating fifteen predictive parameters for forest fires. The utmost accuracy is demonstrated by the Catboost model, which has an AUC value of 0.83. Applied models the Bandarbon and Rangamati regions of the Chittagong hill tracts have been thoroughly analyzed and discovered to be extremely and moderately vulnerable to forest fires due to their intricate physical and social characteristics. The Catboost model indicates that 67.38% of the total regions and 0.59% of the regions in Chittagoong hill tracts are classified as moderate and high susceptible zones, respectively. The results of this study present the models utilized in the present analysis exhibit a high level of effectiveness. Furthermore, the findings indicated that the NDVI and annual temperature had the most influence. The application of both ML and DL approaches in the Chittagong hill tracts for the first time resulted in models that exhibited excellent accuracy and precision, combining perfectly with the actual conditions on the ground. This study will undeniably assist the local government in implementing efficient governance and preservation strategies for the sustainable maintenance of forest resources, while enhancing the well-being of individuals living in and around the forest. Forest fire susceptibility assessment can be utilized in various locations worldwide that include significant natural and human-induced characteristics.

Regional Landslide Susceptibility Assessment and Model Adaptability Research

Landslide susceptibility denotes the likelihood of a disaster event under specific conditions. The assessment of landslide susceptibility has transitioned from qualitative to quantitative methods. With the integration of information technology in geological hazard analysis, a range of quantitative models for assessing landslide susceptibility has emerged and is now widely used. To compare and evaluate the accuracy of these models, this study focuses on Xupu County in Hunan Province, applying several models, including the CF model, FR model, CF-LR coupled model, FR-LR coupled model, SVM model, and RF model, to assess regional landslide susceptibility. ROC curves are used to evaluate the reliability of the model’s predictions. The evaluation results reveal that the CF model (AUC = 0.756), FR model (AUC = 0.764), CF-LR model (AUC = 0.776), FR-LR model (AUC = 0.781), SVM model (AUC = 0.814), and RF model (AUC = 0.912) all have AUC values within the range of 0.7–0.9, indicating that the overall accuracy of the models is good and can provide a reference for landslide susceptibility zoning in the study area. Among these, the Random Forest model demonstrates the best accuracy for landslide susceptibility zoning in the study area. By extracting the extremely high susceptibility zones from the landslide susceptibility zonings obtained by six models, a comparative analysis of model adaptability was conducted. The results indicate that the Random Forest model has the best adaptability under specific conditions in Xupu County.

Susceptibility Mapping of Thaw Slumps Based on Neural Network Methods along the Qinghai–Tibet Engineering Corridor

Climate warming has induced the thawing of permafrost, which increases the probability of thaw slump occurrences in permafrost regions of the Qinghai–Tibet Engineering Corridor (QTEC). As a key and important corridor, thaw slump distribution is widespread, but research into effectively using neural networks to predict thaw slumping remains insufficient. This study automated the identification of thaw slumps within the QTEC and investigated their environmental factors and susceptibility assessment. We applied a deep learning-based semantic segmentation method, combining U-Net with ResNet101, to high spatial and temporal resolution images captured by the Gaofen-1 images. This methodology enabled the automatic delineation of 455 thaw slumps within the corridor area, covering 40,800 km², with corresponding precision, recall, and F1 scores of 0.864, 0.847, and 0.856, respectively. Subsequently, employing a radial basis function neural network model on this inventory of thaw slumps, we investigated environmental factors that could precipitate the occurrence of thaw slumps and generated sensitivity maps of thaw slumps along the QTEC. The model demonstrated high accuracy, and the area under the curve (AUC) value of the receiver operating characteristic (ROC) curve reached 0.95. The findings of the study indicate that these thaw slumps are predominantly located on slopes with gradients of 1–18°, distributed across mid-elevation regions ranging from 4500 to 5500 m above sea level. Temperature and precipitation were identified as the predominant factors that influenced the distribution of thaw slumps. Approximately 30.75% of the QTEC area was found to fall within high to extremely high susceptibility zones. Moreover, validation processes confirmed that 82.75% of the thaw slump distribution was located within areas of high or higher sensitivity within the QTEC.

Assessing locations susceptible to shallow landslide initiation during prolonged intense rainfall in the Lares, Utuado, and Naranjito municipalities of Puerto Rico

Assessing locations susceptible to shallow landslide initiation during prolonged intense rainfall in the Lares, Utuado, and Naranjito municipalities of Puerto Rico

Landslide Inventorization and Susceptibility Mapping in the SemiArid Kargil-Ladakh Region of Northwestern Himalaya

ABSTRACT Landslides often result in damage to property and loss of life in the Himalayan Region because of high seismicity and rugged terrain. To address this issue, the current study focused on developing a landslide inventory based susceptibility map for Kargil-Ladakh Region of North-Western Himalaya. The landslide susceptibility map for the region was generated with the help of Frequency Ratio (FR) Method, landslide inventory layer and eleven influencing factors such as; elevation, geomorphology, aspect, slope, plan curvature, lithology, rainfall, profile curvature, distance from road and lineaments, temperature, and land cover. The results reveal that approximately (50%) of the study-area lies under moderate to very high susceptibility zones. The low and very low susceptibility zones cover the remaining (50%) of the study-area. The findings highlight that several factors significantly contribute to landslide occurrences in the region, namely slope, road network, elevation, rainfall, and land cover. The northern part of the study area is particularly susceptible to landslides due to the presence of the national highway, a high concentration of settlements, and increased infrastructure development. Furthermore, areas surrounding local highways and community road networks show moderate landslide susceptibility. The results were validated with the help of the Receiver Operating Characteristic (ROC) curve, yielding a value of (82%) which is well within the acceptable limit. The results underscore the importance of developing site-specific landslide mitigation strategies to safeguard vulnerable communities in this strategically significant border region of India.

Landslide susceptibility assessment based on frequency ratio and semi-supervised heterogeneous ensemble learning model.

Epistemic uncertainty in data-driven landslide susceptibility assessment often tends to be increased by the limited accuracy of an individual model, as well as uncertainties associated with the selection of non-landslide samples. To address these issues, this paper centers on the landslide disaster in Ji'an City, China, and proposes a heterogeneous ensemble learning method incorporating frequency ratio (FR) and semi-supervised sample expansion. Based on the superimposed results of 12 environmental factor frequency ratios (FFR), non-landslide samples were selected and input into light gradient boosting machine (LightGBM), random forest (RF), and convolutional neural network (CNN) models for prediction along with historical landslide samples. The predicted probability values are integrated by four heterogeneous ensemble strategies to expand samples from high-confidence results. The model's performance is evaluated using the area under the receiver operating characteristic curve (AUC), partition frequency ratio (PFR), and other verification methods. The results demonstrate that the negative sample based on FFR sampling is more accurate than the random sampling method, and the FR-SSELR model based on frequency ratio sampling and semi-supervised ensemble strategy exhibits the highest performance (AUC = 0.971, ACC = 0.941). A more reasonable landslide susceptibility map was drawn based on this model, with the lowest percentage of landslides in the low and very low susceptibility zones (sum of PFR = 0.194), as well as the highest percentage of landslides in the high and very high susceptibility zones (sum of PFR = 6.800). Furthermore, the FR-SSELR model improved economic benefits by 3.82-14.2%, offering valuable guidance for decision-making regarding landslide management and the sustainability of Ji'an City.