Landslide Disaster Prevention Research Articles

Advances in Deep Learning Recognition of Landslides Based on Remote Sensing Images

Against the backdrop of global warming and increased rainfall, the hazards and potential risks of landslides are increasing. The rapid generation of a landslide inventory is of great significance for landslide disaster prevention and reduction. Deep learning has been widely applied in landslide identification due to its advantages in terms of its deeper model structure, high efficiency, and high accuracy. This article first provides an overview of deep learning technology and its basic principles, as well as the current status of landslide remote sensing databases. Then, classic landslide deep learning recognition models such as AlexNet, ResNet, YOLO, Mask R-CNN, U-Net, Transformer, EfficientNet, DeeplabV3+ and PSPNet were introduced, and the advantages and limitations of each model were extensively analyzed. Finally, the current constraints of deep learning in landslide identification were summarized, and the development direction of deep learning in landslide identification was analyzed. The purpose of this article is to promote the in-depth development of landslide identification research in order to provide academic references for the prevention and mitigation of landslide disasters and post-disaster rescue work. The research results indicate that deep learning methods have the characteristics of high efficiency and accuracy in automatic landslide recognition, and more attention should be paid to the development of emerging deep learning models in landslide recognition in the future.

Fast physically-based probabilistic modelling of rainfall-induced shallow landslide susceptibility at the regional scale considering geotechnical uncertainties and different hydrological conditions

The inherent uncertainty in hydro-geotechnical parameters presents a significant challenge for accurately predicting rainfall-triggered shallow landslides in mountainous regions. In this study, a novel probabilistic framework was developed and implemented in the “Py.GIS-FSLAM-FORM” software, designed to address the complexities associated with parameter uncertainty, correlation, and distribution. By combining the Fast Shallow Landslide Assessment Model (FSLAM) with the First-Order Reliability Method (FORM), we have enhanced the traditional probabilistic approach to create more accurate landslide susceptibility maps. This study emphasizes the uncertainly of geotechnical parameters and the critical influence of hydrological conditions on landslide susceptibility, especially focusing on the interaction between antecedent recharge (qa) and event rainfall (Pe). In our study area (Val d’Aran, Spain), the probabilistically based results revealed that areas of very high susceptibility are significantly affected by event rainfall, particularly on slopes of 30–40 degrees and aspects between 100 and 250 degrees. The variability in geotechnical parameters, especially the coefficient of variation (COV) in cohesion and friction angle, plays a crucial role in landslide susceptibility assessment, with increased COVs leading to greater landslide uncertainty. Additionally, cross-negative correlations and non-normal distributions of geotechnical parameters substantially influence the spatial distribution of landslides, notably when combining antecedent recharge with event rainfall. These results highlight the importance of incorporating parameter variability and hydrological conditions in susceptibility models to improve the precision of regional landslide forecasts. While the study was performed in Val d'Aran, its methodologies and conclusions are relevant to mountainous areas worldwide, offering insights for refining landslide prediction models and susceptibility assessments, contributing to global efforts in landslide disaster prevention.

An interpretable model for landslide susceptibility assessment based on Optuna hyperparameter optimization and Random Forest

This study proposed an interpretable model that combines Random Forest (RF), Optuna hyperparameter optimization, and SHapley Additive exPlanations (SHAP) to achieve optimal landslide susceptibility evaluation and provide explanations in the northwest region of Yunnan Province in China. First, an inventory of 4447 landslides and 23 related factors was considered for the landslide susceptibility assessment. Subsequently, a hyperparameter-optimized RF model was developed using the Optuna framework and the training dataset to generate landslide susceptibility maps. The performance of the models were evaluated using accuracy (ACC), precision (PPV), recall (TPR), F1-score (F1), and the Area Under the Curve (AUC) based on the Receiver Operating Characteristic. Furthermore, the interpretability of the model was enhanced through the implementation of SHAP. The proposed model demonstrated outstanding performance on the test set, achieving an ACC of 0.7792, PPV of 0.7448, TPR of 0.8769, F1 of 0.8055, and an AUC of 0.8387. The interpretability analysis revealed that elevation, population density, distance from roads, and normalized difference vegetation index were the primary factors influencing landslide occurrences in the study area. This study provides a comprehensive framework for evaluating landslide susceptibility in specific regions and offers invaluable insights for the prevention and management of landslide disasters.

Enhancing Seismic Landslide Susceptibility Analysis for Sustainable Disaster Risk Management through Machine Learning

The accuracy of Seismic Landslide Susceptibility Maps (SLSMs) is imperative for the prevention of seismic landslide disasters. This study enhances the precision of SLSMs by integrating nine distinct machine learning methodologies with the GeoDetector version 0.0.4 tool to filter both numerical and physical factors contributing to landslide susceptibility. The dataset comprises 2317 landslide instances triggered by the 2013 Minxian Ms = 6.6 earthquake, from which redundant factors were pruned using the Recursive Feature Elimination technique. Subsequent evaluations of the optimized factors, both individually and in combination, were conducted through Frequency Ratio analysis and Factor Interaction assessment. The study juxtaposes the Area Under the Receiver Operating Characteristic Curve (AUC) and the accuracy of nine machine learning models before and after factor optimization. The findings indicate an increase in AUC from a maximum of 0.989 to 0.992 in the Random Forest model, and an 8.37% increase in AUC for the SVM model, signifying a notable enhancement in the stability across all models. The establishment of the SLSM notably elevated the frequency ratio in high-risk zones from 50.40 to 85.14, underscoring the efficacy of combining machine learning and detector optimization techniques in sustainable practices. This research proposes a universal framework aimed at eliminating redundancy and noise in SLSMs and hazard risk assessments, thereby facilitating sustainable geological disaster risk management.

A landslide susceptibility assessment method considering the similarity of geographic environments based on graph neural network

Landslide susceptibility assessment (LSA) is vital for landslide mitigation and management. Existing LSA methods only consider local environmental characteristics associated with landslides, while neglecting the characteristics of remote yet interconnected geographic environments. In addition, the datasets for landslide samples do not consider the equilibrium among various geographic environmental characteristics. Ultimately, the landslide susceptibility mapping (LSM) generated is unreliable. We propose an LSA method based on the similarity of geographic environments and the equilibrium of landslide samples. We selected the Bailong River Basin (BRB) as the study area. Our proposed LSA method has four steps: (1) We constructed slope units using the curvature watershed method (CWM) based on the digital elevation model (DEM) to generate geographic nodes and selected 12 landslide influencing factors (LIFs) through multicollinearity analysis to extract geographic node features. (2) We constructed an association graph of geographic environment similarity constraints based on geographic node features using the cosine similarity method to correlate landslide and geographical environments. (3) We constructed sample datasets for an equal number of landslide and non-landslides based on various geographic environments. (4) We constructed a graph neural network (GNN) model based on the association graph to predict the susceptibility of unknown geographic nodes, obtaining a reliable LSM. Compared with random forest (RF), support vector machine (SVM), and multi-layer perceptron (MLP) models, our proposed model framework outperforms in the LSA. The area under the receiver operating characteristic curve (AUC) values improved by 6.2%, 5.7%, and 1.8%, respectively. Specifically, our model can predict landslides across space in areas with few samples. Our proposed model offers an effective approach and essential technical support for landslide disaster investigation and prevention.

Landslides of China's Qinling

AbstractThe Qinling Mountains in China frequently experience geological disasters, with large‐scale landslides being particularly prominent, causing severe economic losses to the local area. To gain a comprehensive understanding of the geological disaters distribution in the region, we conducted extensive research on the entire Qinling Mountains, covering an area of approximately 380,000 km2. By employing methods such as literature review, data collection, and interpretation of remote sensing images, we have successfully created a database of landslides. The inventory of landslides includes a total of 169,888 large‐scale landslides, covering a combined area of approximately 1575 km2. The average size of these landslides is approximately 92,734 m2. The scale of these landslides varies widely, with the smallest individual landslide covering an area of 166.25 m2 and the largest reaching 12.9 km2. Upon examining areas with frequent landslides, it was observed that landslides are usually densely distributed along riverbanks or within valleys. Landslide development is also dense in areas prone to frequent historical earthquakes. This comprehensive database provides essential data to support the analysis of spatial distribution patterns of large‐scale landslides in the Qinling Mountains. It also facilitates landslide assessments and serves as a reference for the prevention and control of landslide disasters in the area.

Influence of Calculation Parameters on the Slope Stability of the Historical Rasos Cemetery in Vilnius (Lithuania)

Many objects of cultural and religious importance are located on naturally or artificially shaped hills. In such cases, the stability of the slopes determines both the safety of users and the facilities themselves and thus the preservation of cultural heritage. The analysis of the slope stability requires the assessment of the soil shear strength, defined as the ratio of resistance forces to driving forces. An important issue in the slope stability analysis is the proper determination of soil mechanical parameters, including their changes due to soil moisture. This paper presents an assessment of the slope stability of the Rasos Cemetery in Vilnius (Lithuania), where, due to a partial slope sliding, some of the tombstones were destroyed by being covered with soil. An analysis of the slope stability (factor of safety) was performed for two cross-sections using six calculation schemes. In each of them, calculations were carried out for three groups of soil parameters and for four design approaches (characteristic values, DA2, DA3a, and DA3b in accordance with the geotechnical standard applicable in the European Union—Eurocode 7: Geotechnical design). For the calculation approach (DA3a) for different groups of soil parameters, the values of the slope stability factors of safety were practically below 1.0. If the value of the factor of safety is less than 1.0, the slope is unstable. In the range of values 1.0–1.3, there is a serious risk of slope instability, and only when the factor of safety value reaches above 1.3 is the slope classified as stable. The comparison of the slope stability factors obtained with different design approaches (in relation to the DA3a) shows that the most unfavorable stability factor is approximately 43% lower than the result obtained from the characteristic values, approximately 21% lower than the result obtained using the DA2 design approach, and approximately 24% lower than the result obtained using the DA3b design approach. The analytical and numerical method (GEO5) analysis showed that, regardless of the adopted calculation approach, the required slope stability factor was not obtained. The article shows that selecting the correct geological cross-section to determine the critical slope slip surface and adopting the relevant design approach is crucial for slope sustainability. Based on the analysis, a proprietary solution for the construction work was proposed, which will ensure both the full stability of the slope and the safety of all graves and consequently landslide disaster prevention.

Dynamic landslide susceptibility mapping based on the PS-InSAR deformation intensity.

In order to meet the needs of refined landslide risk management, the extended correlation framework of dynamic susceptibility modeling desiderates to be further explored. This work considered the Wanzhou channel of the Three Gorges Reservoir Area as the experimental site, with a transportation channel with significant economic value to carry out innovative research in two stages. (i) Five machine learning models logistic regression (LR), multilayer perceptron neural network (MLPNN), support vector machine (SVM), random forest (RF), and decision tree (DT) were used to explore landslide susceptibility distribution based on detailed landslide boundaries. (ii) Based on the PS-InSAR technology, the dynamic factor of deformation intensity was obtained. Subsequently, the dynamic factor was combined with proposed static factors (topography conditions, geological conditions, hydrological conditions, and human activities) to generate dynamic landslide susceptibility mapping (DLSM). The receiver operating characteristic (ROC) curve, accuracy, precision, recall, and F1 score were proposed as evaluation metrics. Compared with ignoring the dynamic factor, the predictive accuracy of some models was further improved when considering the dynamic factor. Especially the DT model, the area under the curve of ROC (AUC) value increased by 2%, and obtained the highest AUC value (93.1%). The susceptibility results of introducing the dynamic factor are more in line with the spatial distribution of actual landslides. The research framework proposed in this study has important reference significance for the dynamic management and prevention of landslide disasters in the study area.

A Landslide Displacement Prediction Model Based on the ICEEMDAN Method and the TCN–BiLSTM Combined Neural Network

Influenced by autochthonous geological conditions and external environmental changes, the evolution of landslides is mostly nonlinear. This article proposes a combined neural network prediction model that combines a temporal convolutional neural network (TCN) and a bidirectional long short-term memory neural network (BiLSTM) to address the shortcomings of traditional recurrent neural networks in predicting displacement-fluctuation-type landslides. Based on the idea of time series decomposition, the improved complete ensemble empirical mode decomposition with an adaptive noise method (ICEEMDAN) was used to decompose displacement time series data into trend and fluctuation terms. Trend displacement is mainly influenced by the internal geological conditions of a landslide, and polynomial fitting is used to determine the future trend displacement; The displacement of the fluctuation term is mainly influenced by the external environment of landslides. This article selects three types of landslide-influencing factors: rainfall, groundwater level elevation, and the historical displacement of landslides. It uses a combination of gray correlation (GRG) and mutual information (MIC) correlation modules for feature screening. Then, TCN is used to extract landslide characteristic factors, and BiLSTM captures the relationship between features and displacement to achieve the prediction of wave term displacement. Finally, the trend term and fluctuation term displacement prediction values are reconstructed to obtain the total displacement prediction value. The results indicate that the ICEEMDAN–TCN–BiLSTM model proposed in this article can accurately predict landslide displacement and has high engineering application value, which is helpful for planning and constructing landslide disaster prevention projects.

Research on landslide susceptibility prediction model based on LSTM-RF-MDBN.

The occurrence of landslide disasters causes huge economic losses and casualties. Although many achievements have been made in predicting the probability of landslide disasters, various factors such as the scale and spatial location of landslide geological disasters should still be fully considered. Further research on how to quantitatively characterize the susceptibility of landslide geological disasters is necessarily important. To this end, taking the Wenchuan earthquake as the research area and extracting eight influencing factors, including terrain information entropy (Ht), lithology, distance from rivers, distance from faults, vegetation coverage (NDVI), distance from roads, peak ground motion acceleration (PGA), and annual rainfall, a landslide susceptibility prediction model was hereby established based on LSTM-RF-MDBN, a landslide susceptibility prediction map was drawn, and the spatial distribution characteristics of landslide disasters were analyzed. The results showed that (1) LSTM had good prediction results for the eight influencing factors, with an average prediction accuracy of 85%; (2) compared with models such as DNN and LR for predicting landslide disaster points, the AUC value of RF for predicting landslide point positions reached 0.88, presenting a higher accuracy compared to other models; (3) the AUC value of the landslide susceptibility prediction model based on LSTM-RF-MDBN reached 0.965, which had a high accuracy in predicting landslide susceptibility. Overall, the research results can provide a scientific basis for selecting the best strategy for landslide disaster warning, prevention, and mitigation.

Investigation of Model Uncertainty in Rainfall-Induced Landslide Prediction under Changing Climate Conditions

Climate change can exacerbate the occurrence of extreme precipitation events, thereby affecting both the frequency and intensity of rainfall-induced landslides. It is important to study the threat of rainfall-induced landslides under future climate conditions for the formulation of disaster prevention and mitigation policies. Due to the complexity of the climate system, there is great uncertainty in the climate variables simulated by a global climate model (GCM), which will be further propagated in landslide prediction. In this study, we investigate the spatial and temporal trends of future landslide hazards in China under climate change, using data from a multi-model ensemble of GCMs based on two scenarios, RCP4.5 and RCP8.5. The uncertainty characteristics are then estimated based on signal-to-noise ratios (SNRs) and the ratio of agreement in sign (RAS). The results show that the uncertainty of landslide prediction is mainly dominated by the GCM ensemble and the RCP scenario settings. Spatially, the uncertainty of landslide prediction is high in the western areas of China and low in the eastern areas of China. Temporally, the uncertainty of landslide prediction is evolving, with characteristics of high uncertainty in the near future and characteristics of low uncertainty in the distant future. The annual average SNRs in the 21st century are 0.44 and 0.50 in RCP4.5 and RCP8.5, respectively, and the RAS of landslide prediction in Southeastern China is only 50–60%. This indicates that more than half of the patterns show trends that are opposite to those of the ensemble, suggesting that their landslide change trends are not universally recognized in the pattern ensemble. Considering the uncertainty of climate change in landslide prediction can enable studies to provide a more comprehensive picture of the possible range of future landslide changes, effectively improving the reliability of landslide hazard prediction and disaster prevention.

Hazard assessment modeling and software development of earthquake-triggered landslides in the Sichuan–Yunnan area, China

Abstract. To enhance the timeliness and accuracy of spatial prediction of coseismic landslides, we propose an improved three-stage spatial prediction strategy and develop corresponding hazard assessment software named Mat.LShazard V1.0. Based on this software, we evaluate the applicability of this improved spatial prediction strategy in six earthquake events that have occurred near the Sichuan–Yunnan region, including the Wenchuan, Ludian, Lushan, Jiuzhaigou, Minxian, and Yushu earthquakes. The results indicate that in the first stage (immediately after the quake event), except for the 2013 Minxian earthquake, the area under the curve (AUC) values of the modeling performance are above 0.8. Among them, the AUC value of the Wenchuan earthquake is the highest, reaching 0.947. The prediction results in the first stage can meet the requirements of emergency rescue by immediately obtaining the overall predicted information of the possible coseismic landslide locations in the quake-affected area. In the second and third stages, with the improvement of landslide data quality, the prediction ability of the model based on the entire landslide database is gradually improved. Based on the entire landslide database, the AUC value of the six events exceeds 0.9, indicating a very high prediction accuracy. For the second and third stages, the predicted landslide area (Ap) is relatively consistent with the observed landslide area (Ao). However, based on the incomplete landslide data in the meizoseismal area, Ap is much smaller than Ao. When the prediction model based on complete landslide data is built, Ap is nearly identical to Ao. This study provides a new application tool for coseismic landslide disaster prevention and mitigation in different stages of emergency rescue, temporary resettlement, and late reconstruction after a major earthquake.

Multi-source monitoring data helps revealing and quantifying the excavation-induced deterioration of rock mass

There exist many high-steep rock slopes in Southwest China due to the construction of large-scale hydropower projects. The study on excavation-induced damage of high rock slopes is vital for disaster prevention and mitigation. Due to the complexity of high-steep rock slopes, assessing the excavation-induced deterioration of rock mass is often challenging. This study used the space-borne Interferometric synthetic aperture radar (InSAR) and multipoint displacement meters to monitor the slope displacements before and after the excavation. Based on the multi-source monitoring data and the parameter estimation method, we evaluated the changes in rock mass properties due to slope excavation. We employed an improved particle swarm optimization algorithm with dynamic topology and adaptive parameter adjustment and a neural network-based surrogate model in the parameter estimation. The degradation of rock mass strength and stiffness coincides with the increased sensitivity of slope deformation to rainfall infiltration. An in-depth analysis indicates that the absence of vegetation protection and excavation-induced deterioration of rock masses are responsible for the noticeable slope deformation during the rainy season. As slope excavation profoundly influences slope stability and deformation, we should pay closer attention to the excavation way and support style. This study also demonstrates the benefits of using multi-source monitoring data in slope deformation analysis. The findings can aid in comprehending the deformation evolution and instability mechanisms of rock slopes excavated at the hydropower sites, thereby contributing to the prevention and mitigation of landslide disasters.

Early detection of potential landslides along high‐speed railway lines: A pressing issue

AbstractEarly detection of landslides is important for prevention and mitigation of landslide disasters. Especially, accurately identifying potential landslides along high‐speed railway is becoming a pressing issue for operation safety of high‐speed railway. Here, 161 Sentinel‐1A satellite images from March 2017 to September 2022 were acquired to detect potential landslides along the Lanzhou‐Urumqi high‐speed railway (LUHR) in Qinghai. Results show that rainfall is the main cause for the transition from secondary creep (steady state) to tertiary creep (accelerated) at the Jiujiawan landslide. The hot spot analysis method was used to cluster the average line‐of‐sight deformation velocity along the LUHR, effectively extracting five deformation zones that may pose a threat. Optical satellite image analysis revealed four potential landslides and one unstable slope. Furthermore, the rescaled range analysis documented that the Hurst exponent is greater than 0.8 for all four potential landslide instability points, implying that the identified potential landslide areas are still in motion and may pose a threat to the operation of the LUHR. On this basis, we propose a framework for early identification and monitoring of landslides along the high‐speed railway lines that supports landslide hazard managing with the goal to reduce landslide risk.

A landslide susceptibility assessment method based on auto-encoder improved deep belief network

Abstract The landslide susceptibility assessment is an essential part of landslide disaster risk identification and prevention. However, the binarization of the hidden layer limits the parameterization ability of the conditional probability of visible layer, making the training process of restricted Boltzmann machine more difficult and further limiting the accuracy and efficiency of deep belief network (DBN) model in landslide susceptibility assessment. Therefore, this study proposed a landslide susceptibility assessment method based on Auto-Encoder (AE)-modified DBN. Zhouqu County, Gansu Province in the People’s Republic of China, was selected as the study area. Historical landslides in Zhouqu County were identified using small baseline subset interferometric synthetic aperture radar technology and optical image. Landslide factors were screened based on a geographical detector and stepwise regression method. The Logcosh loss function and determinant coefficient R 2 index were used to evaluate the training process of the AE model, and the balanced cross entropy loss function was used to evaluate the entire network training process. In addition, the area under the curve (AUC) of the synthetical index model (SIM), support vector machine (SVM), and multilayer perceptron (MLP) were compared and evaluated. The results indicated that the proposed model could significantly improve the accuracy of landslide susceptibility assessment. The AUC value of the proposed model was 0.31, 0.12, and 0.11 higher than that of SIM, SVM, and MLP, respectively. Therefore, the improved DBN model based on AE proposed is reliable for early landslide identification and prediction.

Rapid Landslide Extraction from High-Resolution Remote Sensing Images Using SHAP-OPT-XGBoost

Landslides, the second largest geological hazard after earthquakes, result in significant loss of life and property. Extracting landslide information quickly and accurately is the basis of landslide disaster prevention. Fengjie County, Chongqing, China, is a typical landslide-prone area in the Three Gorges Reservoir Area. In this study, we newly integrate Shapley Additive Explanation (SHAP) and Optuna (OPT) hyperparameter tuning into four basic machine learning algorithms: Gradient Boosting Decision Tree (GBDT), Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), and Additive Boosting (AdaBoost). We construct four new models (SHAP-OPT-GBDT, SHAP-OPT-XGBoost, SHAP-OPT-LightGBM, and SHAP-OPT-AdaBoost) and apply the four new models to landslide extraction for the first time. Firstly, high-resolution remote sensing images were preprocessed, landslide and non-landslide samples were constructed, and an initial feature set with 48 features was built. Secondly, SHAP was used to select features with significant contributions, and the important features were selected. Finally, Optuna, the Bayesian optimization technique, was utilized to automatically select the basic models’ best hyperparameters. The experimental results show that the accuracy (ACC) of these four SHAP-OPT models was above 92% and the training time was less than 1.3 s using mediocre computational hardware. Furthermore, SHAP-OPT-XGBoost achieved the highest accuracy (96.26%). Landslide distribution information in Fengjie County from 2013 to 2020 can be extracted by SHAP-OPT-XGBoost accurately and quickly.

Landslide Risk Mapping Using the Weight-of-Evidence Method in the Datong Mining Area, Qinghai Province

Qinghai is rich in mineral resources, but frequent and large-scale mineral mining has caused secondary damage to the fragile primary surface and produced a large number of landslide disasters. In complex geological environments such as glacier ablation and frequent tectonic movements, a complete quantitative evaluation method for landslide risk in high-cold mining areas has not yet been formed. In view of this, this article uses the field survey and remote sensing data of the Datong mining area in Qinghai Province in 2012 as the basic data. We comprehensively considered five first-level factors (13 s-level factors) including topography, lithological structure, mining engineering activities, land use, and dynamic deformation as evaluation indicators for landslide susceptibility in mining areas, and used the Topographic Wetness Index (TWI) and the Human Engineering Activity Intensity (HEAI) to quantitatively estimate the hazard of landslide according to the landslide trigger mechanism. The weight-of-evidence approach was used for landslide hazard and risk mapping under different landslide--inducing conditions. The results indicate that the extremely high-hazard areas induced by human engineering activities account for 14% of the total area, and the extremely high-risk areas account for 13% of the total area in the Datong mining area, and the area of the extremely high-risk area is large; the landslide risk assessment mapping model constructed in this study can effectively evaluate the probability of slope instability caused by rainfall and human engineering activities. The effective value of the receiver operating characteristic (ROC) curve of the sensitivity assessment model reaches 0.863, and the evaluation results are consistent with reality; using the weight-of-evidence model for landslide risk assessment is more in line with the actual situation in alpine mining areas, and is more suitable for guiding landslide risk management and disaster prevention and mitigation in mining areas.

Hazard Assessment of Rainfall–Induced Landslide Considering the Synergistic Effect of Natural Factors and Human Activities

Landslide hazard assessment is essential for determining the probability of landslide occurrence in a specific spatial and temporal range. The hazard assessment of potential landslides could support landslide disaster early warning and disaster prevention decisions, which have important guiding significance for urban construction and sustainable development. Due to the lack of consideration of the synergistic effect of multiple factors and geographic scene heterogeneity, the accuracy of existing landslide hazard assessment methods still needs to be improved, and the interpretability and applicability of existing models still need to be improved. In this paper, we propose a landslide hazard assessment method considering the synergistic effect of multiple factors, including natural factors and human activities, and the heterogeneity of geographic scenes. On this basis, we carry out experimental verification on rainfall–induced landslides in Dehong Prefecture, Yunnan Province, China. Firstly, rainfall–induced landslide hazards’ characteristics and impact factors are analyzed and classified. The whole study area is divided into some homogeneous sub–regions using regional dynamic constraint clustering based on the similarity of underlying environmental variables. Then, considering the spatial autocorrelation between various landslide conditioning and trigger factors, a local weighted random forest model is developed to evaluate the rainfall–induced landslide hazards comprehensively. Experimental results show that the proposed method has higher accuracy and interpretability than the existing representative methods and can provide useful references for preventing landslide hazards.

Examination of the Effects of Different Frequencies on Rock Fracturing via Laboratory-Scale Variable Amplitude Fatigue Loading Experiments

In this study, the characteristics of the fracture evolution of argillaceous shale under increasing-amplitude loading were investigated. The GCTS RTR-2000 test system and in-situ acoustic emission (AE) monitoring were employed to execute the tests. The following results were observed. (1) The strength, deformation, and fatigue life increased with the frequency, and the morphology of the hysteresis curve changed regularly with time. (2) The cumulative damage of the rock at the location in which the stress amplitude suddenly increased exceeded that at the fatigue loading stage. The AE count and AE energy were affected by the loading frequency. (3) The secant modulus exhibited different values for different loading frequencies; the smaller the loading frequency, the fewer loading stages the samples experienced, and the faster the secant modulus decreased. The change in Poisson’s ratio over the entire process was composed of a steady growth stage and a rapid growth stage. (4) The rock exhibited two stages of damage evolution, with rapid damage accumulation occurring at the beginning of the loading and relatively smooth damage occurring thereafter. This study developed a cumulative fatigue damage model that can adequately fit the accumulated damage during the fracturing process. The experiments revealed that variable amplitude fatigue loading at different frequencies significantly influences the damage deterioration and the failure law of the rock. The results are expected to improve the understanding of the frequency effect on the fracture behavior and help predict the lifespan of rock structures. This is of great significance to the promotion of slope management, landslide disaster prevention, and mine reuse at the West Open-pit Mine.

An Adaptive Identification Method for Potential Landslide Hazards Based on Multisource Data

The effectiveness of landslide disaster prevention depends largely on the quality of early identification of potential hazards, and how to comprehensively, deeply, and accurately identify such hazards has become a major difficulty in landslide disaster management. Existing deep learning methods for potential landslide hazard identification often use fixed-size window modeling and ignore the different window sizes required by landslides of different scales. To address this problem, we propose an adaptive identification method for potential landslide hazards based on multisource data. Taking Yongping County, China, as the study area, we create a multisource factor dataset based on the landslide disaster background in terms of topography, geology, human activities, hydrology, and vegetation as the sample for the identification model after processing. Moreover, we combine differential interferometric synthetic aperture radar (D-InSAR) and multitemporal InSAR (MT-InSAR) to process the surface deformation of the study area, and we measure the deformation richness based on the average of the pixel deformation difference within the current window of a pixel point in the image. Therefore, convolutional neural networks (CNNs) with different window sizes are adaptively selected. The results show that the precision of adaptive identification of potential landslide hazards in the study area is 85.30%, the recall is 83.03%, and the F1 score is 84.15%. The recognition rate for potential hazards reaches 80%, which is better than the fixed-window modeling result and proves the effectiveness of the proposed method. This method can help to improve intelligent identification systems for potential landslide hazards, and also contribute to the identification of other potential geological hazards, such as mudslides and collapses.