Strategic protection of landslide vulnerable mountains for biodiversity conservation under land-cover and climate change impacts.

Natural disasters impose huge uncertainty and loss to human lives and economic activities. Landslides are one disaster that has become more prevalent because of anthropogenic disturbances, such as land-cover changes, land degradation, and expansion of infrastructure. These are further exacerbated by more extreme precipitation due to climate change, which is predicted to trigger more landslides and threaten sustainable development in vulnerable regions. Although biodiversity conservation and development are often regarded as having a trade-off relationship, here we present a global analysis of the area with co-benefits, where conservation through expanding protection and reducing deforestation can not only benefit biodiversity but also reduce landslide risks to human society. High overlap exists between landslide susceptibility and areas of endemism for mammals, birds, and amphibians, which are mostly concentrated in mountain regions. We identified 247 mountain ranges as areas with high vulnerability, having both exceptional biodiversity and landslide risks, accounting for 25.8% of the global mountainous areas. Another 31 biodiverse mountains are classified as future vulnerable mountains as they face increasing landslide risks because of predicted climate change and deforestation. None of these 278 mountains reach the Aichi Target 11 of 17% coverage by protected areas. Of the 278 mountains, 52 need immediate actions because of high vulnerability, severe threats from future deforestation and precipitation extremes, low protection, and high-population density and anthropogenic activities. These actions include protected area expansion, forest conservation, and restoration where it could be a cost-effective way to reduce the risks of landslides.

Analysis of the spatial distribution of the landslides triggered by the 1923 Great Kanto Earthquake, Japan.

The Great Kanto Earthquake that occurred in the southern part of Kanto district, Japan, on September 1, 1923, was reported to have triggered numerous landslides (over 89,080 slope failures over an area of 86.32 km2). This study investigated the relationship between the landslide occurrence caused by this earthquake and geomorphology, geology, soil, seismic ground motion, and coseismic deformation. We found that a higher landslide density was mainly related to a larger absolute curvature and a higher slope angle, as well as to several geological units (Neogene plutonic rock, accretionary prism, and metamorphic rocks). Moreover, we performed decision tree analyses, which showed that slope angle, geology, and coseismic deformation were correlated to landslide density in that order. However, no clear correlation was found between landslide density and seismic ground motion. These results suggest that landslide density was greater in areas of large slope angle or fragile geology in the area with strong shaking enough to trigger landslides.

Exploring the spatial patterns of landslide susceptibility assessment using interpretable Shapley method: Mechanisms of landslide formation in the Sichuan-Tibet region.

Machine learning models are often viewed as black boxes in landslide susceptibility assessment, lacking an analysis of how input features predict outcomes. This makes it challenging to understand the mechanisms and key factors behind landslides. To enhance the interpretability of machine learning models in wide-area landslide susceptibility assessments, this study uses the Shapely method to explore the contributions of feature factors from local, global, and spatial perspectives. Landslide susceptibility assessments were conducted using random forest (RF), support vector machine (SVM), and eXtreme Gradient Boosting (XGBoost) models, focusing on the geologically complex Sichuan-Tibet region. Initially, the study revealed the contributions of specific key feature factors to landslides from a local perspective. It then examines the overall impact of interactions among feature factors on landslide occurrence globally. Finally, it unveils the spatial distribution patterns of the contributions of various feature factors to landslide occurrence. The analysis indicates the following: (1) The XGBoost model excels in landslide susceptibility assessment, achieving accuracy, precision, recall, F1-score, and AUC values of 0.7815, 0.7858, 0.7962, 0.7910, and 0.86, respectively; (2) The Shapely method identifies the leading factors for landslides in the Sichuan-Tibet region as Elevation (3000-4000 m), PGA (1-2 g), NDVI (<0.5), and distance to rivers (<3 km); (3) Using the Shapely method, the study explains the contributions, interaction mechanisms, and spatial distribution patterns of landslide susceptibility feature factors across local, global, and spatial perspectives. These findings offer new avenues and methods for the in-depth exploration and scientific prediction of landslide risks.

Engineering and environmental assessment of soilbag-based slope stabilisation for sustainable landslide mitigation in mountainous area.

Changes in land use significantly impact landslide occurrence, particularly in mountainous areas in northern Thailand, where human activities such as urbanization, deforestation, and slope modifications alter natural slope angles, increasing susceptibility to landslides. To address this issue, an appropriate method using soilbags has been widely used for slope stabilisation in northern Thailand, but their effectiveness and sustainability require assessment. This research highlights the need to evaluate the stability of the soilbag-based method. In this study, a case study was conducted in northern Thailand, focusing on an area characterised by high-risk landslide potential. This research focuses on numerical evaluation the slope stability of soilbag-reinforced structures and discusses environmental sustainability. The study includes site investigations using an unmanned aerial photogrammetric survey for slope geometry evaluation and employing the microtremor survey technique for subsurface investigation. Soil and soilbag material parameters are obtained from existing literatures. Modelling incorporates hydrological data, slope geometry, subsurface conditions, and material parameters. Afterwards, the pore-water pressure results and safety factors are analysed. Finally, the sustainability of soilbags is discussed based on the Sustainable Development Goals (SDGs). The results demonstrate that soilbags effectively mitigate pore-water pressures, improve stability, and align with several SDGs objectives. This study enhances understanding of soilbags in slope stabilisation and introduces a sustainable landslide mitigation approach for landslide-prone regions.

Construction of ecological security pattern in combination with landslide sensitivity: A case study of Yan'an City, China.

The ecological security pattern can harmonize the relationship between natural environmental protection and socio-economic development. This study proposes a regional ecological security pattern optimization framework by integrating theory and practice with landslide sensitivity and landscape structure. Using Yan'an City as an example, this study optimizes the landscape layout of preliminary ecological sources. The landslide sensitivity index is generated using the information value model and then used to adjust the ecological resistance surface. The Minimum Cumulative Resistance (MCR) approach is used to extract ecological corridors, locate ecological nodes utilizing circuit theory, and outline crucial ecological control areas. The results demonstrate: (1) the ecological sources are primarily composed of forestlands, with a total area of 2,352.2400 km2, concentrated in the southwest, central, and southeast regions. The optimal landscape granularity for the source patches is 600 m. (2) Yan'an is divided into four landslide sensitivity level zones: extremely high, high, medium, and low, with the overall landslide sensitivity of the region being high. (3) The highest ecological resistance is observed in built-up land and the lowest in forestland. The total number of ecological corridors is 26, avoiding most of the highly sensitive areas of landslides. (4) The number of ecological pinch points is 61, while the ecological barrier points amounted to 54. The critical ecological control areas consist mainly of cropland, forestland, and grassland, and differentiated restoration strategies are proposed to address their unique characteristics. The findings of the research can offer scientific guidance for the practice of ecological security protection in geohazard-prone areas.

Investigating the dynamic nature of landslide susceptibility in the Indian Himalayan region.

Landslide susceptibility zonation (LSZ) mapping is used to delineate areas prone to landslides and is critical for effective landslide hazard management. The existing methodologies for generating such maps tend to neglect the influence of dynamic environmental variables on landslide occurrences, which may lead to obsolete and erroneous estimates of landslide susceptibility (LS) for a concerned area. Although recent studies have started to report the effects of Land Use/ Land Cover (LULC) variation on LSZ mapping, variations in other dynamic variables like rainfall, soil moisture, and evapotranspiration apart from LULC may also influence slope stability in mountainous regions. The present study investigates the impact of variations in these four variables on the LS distribution, of a selected Indian Himalayan region between 2017 and 2021. Random Forest (RF) susceptibility models are utilized for evaluating the LS for the selected years and geospatial technologies are employed for LS change detection. The results indicate up to 19% variations in the spatial extent for some of the zones of the generated LSZ maps. The research findings of this study are crucial since they reveal the impact of dynamic behavior on LS, which has not been previously documented in the literature.

Land use and climate change impacts on global soil erosion by water (2015-2070).

Soil erosion is a major global soil degradation threat to land, freshwater, and oceans. Wind and water are the major drivers, with water erosion over land being the focus of this work; excluding gullying and river bank erosion. Improving knowledge of the probable future rates of soil erosion, accelerated by human activity, is important both for policy makers engaged in land use decision-making and for earth-system modelers seeking to reduce uncertainty on global predictions. Here we predict future rates of erosion by modeling change in potential global soil erosion by water using three alternative (2.6, 4.5, and 8.5) Shared Socioeconomic Pathway and Representative Concentration Pathway (SSP-RCP) scenarios. Global predictions rely on a high spatial resolution Revised Universal Soil Loss Equation (RUSLE)-based semiempirical modeling approach (GloSEM). The baseline model (2015) predicts global potential soil erosion rates of [Formula: see text] Pg yr-1, with current conservation agriculture (CA) practices estimated to reduce this by ∼5%. Our future scenarios suggest that socioeconomic developments impacting land use will either decrease (SSP1-RCP2.6-10%) or increase (SSP2-RCP4.5 +2%, SSP5-RCP8.5 +10%) water erosion by 2070. Climate projections, for all global dynamics scenarios, indicate a trend, moving toward a more vigorous hydrological cycle, which could increase global water erosion (+30 to +66%). Accepting some degrees of uncertainty, our findings provide insights into how possible future socioeconomic development will affect soil erosion by water using a globally consistent approach. This preliminary evidence seeks to inform efforts such as those of the United Nations to assess global soil erosion and inform decision makers developing national strategies for soil conservation.

Insights into geospatial heterogeneity of landslide susceptibility based on the SHAP-XGBoost model.

The spatial heterogeneity of landslide influencing factors is the main reason for the poor generalizability of the susceptibility evaluation model. This study aimed to construct a comprehensive explanatory framework for landslide susceptibility evaluation models based on the SHAP (SHapley Additive explanation)-XGBoost (eXtreme Gradient Boosting) algorithm, analyze the regional characteristics and spatial heterogeneity of landslide influencing factors, and discuss the heterogeneity of the generalizability of the models under different landscapes. Firstly, we selected different regions in typical mountainous hilly region and constructed a geospatial database containing 12 landslide influencing factors such as elevation, annual average rainfall, slope, lithology, and NDVI through field surveys, satellite images, and a literature review. Subsequently, the landslide susceptibility evaluation model was constructed based on the XGBoost algorithm and spatial database, and the prediction results of the landslide susceptibility evaluation model were explained based on regional topography, geology, and hydrology using the SHAP algorithm. Finally, the model was generalized and applied to regions with both similar and very different topography, geology, meteorology, and vegetation, to explore the spatial heterogeneity of the generalizability of the model. The following conclusions were drawn: the spatial distribution of landslides is heterogeneous and complex, and the contribution of each influencing factor on the occurrence of landslides has obvious regional characteristics and spatial heterogeneity. The generalizability of the landslide susceptibility evaluation model is spatially heterogeneous and has better generalizability to regions with similar regional characteristics. Further explanation of the XGBoost landslide susceptibility evaluation model using the SHAP method allows quantitative analysis of the differences in how much various factors contribute to disasters due to spatial heterogeneity, from the perspective of global and local evaluation units. In summary, the integrated explanatory framework based on the SHAP-XGBoost model can quantify the contribution of influencing factors on landslide occurrence at both global and local levels, which is conducive to the construction and improvement of the influencing factor system of landslide susceptibility in different regions. It can also provide a reference for predicting potential landslide hazard-prone areas and for Explainable Artificial Intelligence (XAI) research.

A hybrid machine learning model for landslide-oriented risk assessment of long-distance pipelines.

Preparation of pipeline risk zoning is essential for pipeline construction and safe operation. Landslides are one of the main sources of risk to the safe operations of oil and gas pipelines in mountainous areas. This work aims to propose a quantitative assessment model of landslide-induced long-distance pipeline risk by analyzing historical landslide hazard data along oil and gas pipelines. Using the Changshou-Fuling-Wulong-Nanchuan (CN) gas pipeline dataset, two independent assessments were carried out: landslide susceptibility assessment and pipeline vulnerability assessment. Firstly, the study combined the recursive feature elimination and particle swarm optimization-AdaBoost method (RFE-PSO-AdaBoost) to develop a landslide susceptibility mapping model. The RFE method was used to select the conditioning factors, while PSO was used to tune the hyper-parameters. Secondly, considering the angular relationship between the pipelines and landslides, and the segmentation of the pipelines using the fuzzy clustering (FC), the CRITIC method (FC-CRITIC) was combined to develop a pipeline vulnerability assessment model. Accordingly, a pipeline risk map was obtained based on pipeline vulnerability and landslide susceptibility assessment. The study results show that almost 35.3% of the slope units were in extremely high susceptibility zones, 6.68% of the pipelines were in extremely high vulnerability areas, the southern and eastern pipelines segmented in the study area were located in high risk areas and coincided well with the distribution of landslides. The proposed hybrid machine learning model for landslide-oriented risk assessment of long-distance pipelines can provide a scientific and reasonable risk classification for new planning or in service pipelines to avoid landslide-oriented risk and ensure their safe operation in mountainous areas.

Recording Historical Variations in Agricultural Areas Caused by Landslides and River Floods between 1977 and 2012 through Participatory Mapping in the Foothills of the Atacama Desert, Chile.

In arid areas, rural communities can be affected by erosive phenomena caused by intense rainfall. By involving such communities in participatory mapping over the last few decades, our ability to analyse the effects of these phenomena has been enhanced. The aim of this study was to evaluate participatory mapping as a tool for spatially analysing agricultural variations caused by erosive phenomena, using local people to identify chronologies of physical events so we could analyse their effects on agriculture. The study was conducted in Laonzana, Tarapacá Valley, in northern Chile. We selected the participants for the participatory mapping using specific criteria, and carried out field activities in different phases, which allowed the identification, georeferencing and registration (through participatory mapping) of the information collected in the field and from the collective memories of the participants. Three periods were studied. This provided evidence for a decrease in the number of productive sites, these being limited to the vicinity of the village. The participatory mapping technique has become a useful tool in desert and mountainous areas with low population densities for recovering experiential information from communities.

Determination of landslide susceptibility with Analytic Hierarchy Process (AHP) and the role of forest ecosystem services on landslide susceptibility.

The analysis of landslide susceptibility is a crucial tool in the mitigation and management of ecological and economic hazards. The number of studies examining how the form and durability of forest areas affect landslide susceptibility is very limited. This study was conducted in the Marmara region of northwestern Türkiye, where forested areas and industrial zones are intertwined and dense. The landslide susceptibility map was produced by Analytic Hierarchy Process (AHP) method. In the context of AHP, a total of 12 different variables were employed, namely lithology, slope, curvatures, precipitations, aspect, distance to fault lines, distance to streams, distance to roads, land use, soil, elevation, and Normalized Difference Vegetation Index (NDVI). The performance analysis of the landslide susceptibility map was conducted using the Receiver Operating Characteristics (ROC) curve method. The AUC value was computed (0.809) for the landslide susceptibility map generated by using the AHP technique. Forest type maps were used to analyze the impact of forests on landslide susceptibility. In terms of forest structure, 4 main criteria were determined: stand structure, development stage, crown closure, and stand age. Each criterion was analyzed with Geographic Information Systems (GIS) by overlaying it with the landslide susceptibility map of the study area. The results showed that the risk of landslides was lowest in forests with more than one tree species, mature, development stage and of (e) > 52 cm, and crown closure of 41%-70% (2).

The role of PMCT for the assessment of the cause of death in natural disaster (landslide and flood): a Sicilian experience.

In this report, the authors provide a contribution of PMCT in assessing the cause of death due to natural disasters. Here, the PMCT findings of 43 subjects who died during both landslide and flood were described. The post-mortem imaging revealed, clearly, traumatic injuries and/or the presence of foreign material in airways allowing to assess the cause of death of each subject, together with external inspection and the collected circumstantial data. Particularly, the PMCT has been helpful for characterization and localization of the clogging substance in airways providing findings on bronchial branches involvement. Moreover, the investigation offered detailed data on skeletal injuries in all anatomic districts and put in evidence both the precise fracturing site and the characteristics of fracture stubs for each bone fracture. This report supports the recommendation of the virtual autopsy in a case with several victims, as in natural disasters, and its role as an alternative diagnostic investigation when the standard autopsy is not feasible.

Landslides associated with recent road constructions in the Río Lucma catchment, eastern Cordillera Blanca, Peru.

Extensive road construction works recently took place in the remote eastern part of the Peruvian Cordillera Blanca, aiming at a better connection of isolated mountain communities with regional administrative centres. Here we document and characterize landslides associated with these road construction efforts in the Río Lucma catchment, Peru. We show that a total area of 321,332 m2 has been affected by landslides along the 47.1 km of roads constructed between 2015 and 2018. While landslides downslope the roads (48.2%) and complex landslides crossing the roads (46.4%) were the most frequent landslide types in relation to the position of the road; slide-type movement (60.7%) prevails over the flow-type movement (39.3%). Timewise, we found that 75.0% of landslides were observed simultaneously with road construction work, while the remaining 25.0% occurred up to seven months after the roads had been constructed. We plotted the lagged occurrence of these subsequent landslides against precipitation data, showing that 85.7% of them were observed during the wet season (November to April). We conclude that the majority of mapped landslides were directly associated with road constructions and that the road constructions also may set preconditions for landslides, which mainly occurred during the subsequent wet season.

Identification of Landslides in Mountainous Area with the Combination of SBAS-InSAR and Yolo Model.

Landslides have been frequently occurring in the high mountainous areas in China and poses serious threats to peoples' lives and property, economic development, and national security. Detecting and monitoring quiescent or active landslides is important for predicting risks and mitigating losses. However, traditional ground survey methods, such as field investigation, GNSS, and total stations, are only suitable for field investigation at a specific site rather than identifying landslides over a large area, as they are expensive, time-consuming, and laborious. In this study, the feasibility of using SBAS-InSAR to detect landslides in the high mountainous areas along the Yunnan Myanmar border was tested first, with fifty-four IW mode Sentinel-1A ascending scenes from 12 January 2019 to 8 December 2020. Next, the Yolo deep-learning model with Gaofen-2 images captured on 5 December 2020 was tested. Finally, the two techniques were combined to achieve better performance, given each of them has intrinsic limitations on landslide detection. The experiment indicated that the combination could improve the match rate between detection results and references, which implied that the performance of landslide detection can be improved with the fusion of time series SAR images and optical images.

Combined SBAS-InSAR and PSO-RF Algorithm for Evaluating the Susceptibility Prediction of Landslide in Complex Mountainous Area: A Case Study of Ludian County, China.

In complex mountainous areas where earthquakes are frequent, landslide hazards pose a significant threat to human life and property due to their high degree of concealment, complex development mechanism, and abrupt nature. In view of the problems of the existing landslide hazard susceptibility evaluation model, such as poor effectiveness and inaccuracy of landslide hazard data and the need for experts to participate in the calculation of a large number of evaluation factor weight classification statistics. In this paper, a combined SBAS-InSAR (Small Baseline Subsets-Interferometric Synthetic Aperture Radar) and PSO-RF (Particle Swarm Optimization-Random Forest) algorithm was proposed to evaluate the susceptibility of landslide hazards in complex mountainous regions characterized by frequent earthquakes, deep river valleys, and large terrain height differences. First, the SBAS-InSAR technique was used to invert the surface deformation rates of the study area and identified potential landslide hazards. Second, the study area was divided into 412,585 grid cells, and the 16 selected environmental factors were analyzed comprehensively to identify the most effective evaluation factors. Last, 2722 landslide (1361 grid cells) and non-landslide (1361 grid cells) grid cells in the study area were randomly divided into a training dataset (70%) and a test dataset (30%). By analyzing real landslide and non-landslide data, the performances of the PSO-RF algorithm and three other machine learning algorithms, BP (back propagation), SVM (support vector machines), and RF (random forest) algorithms were compared. The results showed that 329 potential landslide hazards were updated using the surface deformation rates and existing landslide cataloguing data. Furthermore, the area under the curve (AUC) value and the accuracy (ACC) of the PSO-RF algorithm were 0.9567 and 0.8874, which were higher than those of the BP (0.8823 and 0.8274), SVM (0.8910 and 0.8311), and RF (0.9293 and 0.8531), respectively. In conclusion, the method put forth in this paper can be effectively updated landslide data sources and implemented a susceptibility prediction assessment of landslide disasters in intricate mountainous areas. The findings can serve as a strong reference for the prevention of landslide hazards and decision-making mitigation by government departments.

A Robust Deep-Learning Model for Landslide Susceptibility Mapping: A Case Study of Kurdistan Province, Iran.

We mapped landslide susceptibility in Kamyaran city of Kurdistan Province, Iran, using a robust deep-learning (DP) model based on a combination of extreme learning machine (ELM), deep belief network (DBN), back propagation (BP), and genetic algorithm (GA). A total of 118 landslide locations were recorded and divided in the training and testing datasets. We selected 25 conditioning factors, and of these, we specified the most important ones by an information gain ratio (IGR) technique. We assessed the performance of the DP model using statistical measures including sensitivity, specificity, accuracy, F1-measure, and area under-the-receiver operating characteristic curve (AUC). Three benchmark algorithms, i.e., support vector machine (SVM), REPTree, and NBTree, were used to check the applicability of the proposed model. The results by IGR concluded that of the 25 conditioning factors, only 16 factors were important for our modeling procedure, and of these, distance to road, road density, lithology and land use were the four most significant factors. Results based on the testing dataset revealed that the DP model had the highest accuracy (0.926) of the compared algorithms, followed by NBTree (0.917), REPTree (0.903), and SVM (0.894). The landslide susceptibility maps prepared from the DP model with AUC = 0.870 performed the best. We consider the DP model a suitable tool for landslide susceptibility mapping.

Landslide Susceptibility Mapping with Integrated SBAS-InSAR Technique: A Case Study of Dongchuan District, Yunnan (China).

Landslide susceptibility maps (LSM) are often used by government departments to carry out land use management and planning, which supports decision makers in urban and infrastructure planning. The accuracy of conventional landslide susceptibility maps is often affected by classification errors. Consequently, they become less reliable, which makes it difficult to meet the needs of decision-makers. Therefore, it is proposed in this paper to reduce classification errors and improve LSM reliability by integrating the Small Baseline Subsets-Interferometric Synthetic Aperture Radar (SBAS-InSAR) technique and LSM. By using the logistic regression model (LR) and the support vector machine model (SVM), experiments were conducted to generate LSM in the Dongchuan district. It was classified into five classes: very high susceptibility, high susceptibility, medium susceptibility, low susceptibility, and very low susceptibility. Then, the surface deformation rate of the Dongchuan area was obtained through the ascending and descending orbit sentinel-1A data from January 2018 to January 2021. To correct the classification errors, the SBAS-InSAR technique was integrated into LSM under the optimal model by constructing the contingency matrix. Finally, the LSMs obtained before and after correction were compared. Moreover, the correction results were validated and analyzed by combining remote sensing images, InSAR deformation results, and field surveys. According to the research results, the susceptibility class of 66,094 classification error cells (59.48 km2) was significantly improved in the LSM after the integration of the SBAS-InSAR correction. The enhanced susceptibility classes and the spectral characteristics of remote sensing images are highly consistent with the trends of InSAR cumulative deformation and the results of field investigation. It is suggested that integrating SBAS-InSAR and LSM is effective in correcting classification errors and further improving the reliability of LSM for landslide prediction. The LSM obtained by using this method plays an important role in guiding local government departments on disaster prevention and mitigation, which is conducive to eliminating the risk of landslides.

Landslide Susceptibility Mapping by Fusing Convolutional Neural Networks and Vision Transformer.

Landslide susceptibility mapping (LSM) is an important decision basis for regional landslide hazard risk management, territorial spatial planning and landslide decision making. The current convolutional neural network (CNN)-based landslide susceptibility mapping models do not adequately take into account the spatial nature of texture features, and vision transformer (ViT)-based LSM models have high requirements for the amount of training data. In this study, we overcome the shortcomings of CNN and ViT by fusing these two deep learning models (bottleneck transformer network (BoTNet) and convolutional vision transformer network (ConViT)), and the fused model was used to predict the probability of landslide occurrence. First, we integrated historical landslide data and landslide evaluation factors and analysed whether there was covariance in the landslide evaluation factors. Then, the testing accuracy and generalisation ability of the CNN, ViT, BoTNet and ConViT models were compared and analysed. Finally, four landslide susceptibility mapping models were used to predict the probability of landslide occurrence in Pingwu County, Sichuan Province, China. Among them, BoTNet and ConViT had the highest accuracy, both at 87.78%, an improvement of 1.11% compared to a single model, while ConViT had the highest F1-socre at 87.64%, an improvement of 1.28% compared to a single model. The results indicate that the fusion model of CNN and ViT has better LSM performance than the single model. Meanwhile, the evaluation results of this study can be used as one of the basic tools for landslide hazard risk quantification and disaster prevention in Pingwu County.