Landslide Information Research Articles

SAM-CFFNet: SAM-Based Cross-Feature Fusion Network for Intelligent Identification of Landslides

Landslides are common hazardous geological events, and accurate and efficient landslide identification methods are important for hazard assessment and post-disaster response to geological disasters. Deep learning (DL) methods based on remote sensing data are currently widely used in landslide identification tasks. The recently proposed segment anything model (SAM) has shown strong generalization capabilities in zero-shot semantic segmentation. Nevertheless, SAM heavily relies on user-provided prompts, and performs poorly in identifying landslides on remote sensing images. In this study, we propose a SAM-based cross-feature fusion network (SAM-CFFNet) for the landslide identification task. The model utilizes SAM’s image encoder to extract multi-level features and our proposed cross-feature fusion decoder (CFFD) to generate high-precision segmentation results. The CFFD enhances landslide information through fine-tuning and cross-fusing multi-level features while leveraging a shallow feature extractor (SFE) to supplement texture details and improve recognition performance. SAM-CFFNet achieves high-precision landslide identification without the need for prompts while retaining SAM’s robust feature extraction capabilities. Experimental results on three open-source landslide datasets show that SAM-CFFNet outperformed other comparative models in terms of landslide identification accuracy and achieved an intersection over union (IoU) of 77.13%, 55.26%, and 73.87% on the three datasets, respectively. Our ablation studies confirm the effectiveness of each module designed in our model. Moreover, we validated the justification for our CFFD design through comparative analysis with diverse decoders. SAM-CFFNet achieves precise landslide identification using remote sensing images, demonstrating the potential application of the SAM-based model in geohazard analysis.

Landslide extraction from aerial imagery considering context association characteristics

Accurate extraction of landslide information is crucial for timely disaster emergency response, yet this process faces significant challenges due to the interference of bare land and vegetation cover as well as the variability in landslide scales. Furthermore, the scarcity of comprehensive landslide datasets has hindered the in-depth exploration of deep learning techniques in this domain. Addressing these issues, this study introduces a novel framework for landslide extraction that leverages the characteristics of context association. In the encoding phase, a two-branch multiscale context feature extraction module (TMCFM) is established, which captures the contextual relationships across different scales through an attention mechanism while concurrently extracting context information within the same scale. To refine the interplay between features at different levels, self-attention is utilized, allowing for the effective fusion of context information at both ends of the same scale through feature fusion exploiting static and dynamic context feature modules (FSDC). Building upon this, this study develops a deeply supervised classifier (DSC) that enhances the network’s discriminative capabilities in the prediction phase via six auxiliary branches. Additionally, this study contributes a new aerial image dataset for landslide extraction created through meticulous visual interpretation. The proposed method is thoroughly compared with 17 contemporary deep learning methods, demonstrating an increase in the intersection-over-union (IoU) metric of 0.92%-16.94% over these methods. The robustness and superiority of the proposed approach are further validated through various discussions and analyses.

Research on an Object-Oriented Intelligent Extraction Method For Landslide

Landslides are the most important type of geological disaster development in China. Heavy rainfall and earthquakes often cause mass landslides. Rapid extraction of postdisaster landslide information is an important method for realizing scientific and efficient emergency investigations and can provide a key basis for disaster assessment and emergency response decision-making. Rapid extraction of postdisaster landslide information is an important method for realizing scientific and efficient emergency investigations and can provide a key basis for disaster assessment and emergency response decision-making. Using Fujian as an example, based on basic topography, geology and other data, the multiscale segmentation method is used for image segmentation, and different features are selected. The applicability of the rule-based extraction method and support vector machine, random forest and other machine learning methods. The applicability of the rule-based extraction method and support vector machine, random forest and other machine learning methods in small-scale landslide extraction in the study area is compared. The results showed that the extraction accuracy of the rule-based extraction method was 90.86%, the extraction accuracy of the support vector machine was 65.95%, and the extraction accuracy of the random forest was 93.43%. A comparison of the three methods revealed that the random forest method was most suitable for landslide extraction in the study area.

A Multiscale Attention Segment Network-Based Semantic Segmentation Model for Landslide Remote Sensing Images

Landslide disasters have garnered significant attention due to their extensive devastating impact, leading to a growing emphasis on the prompt and precise identification and detection of landslides as a prominent area of research. Previous research has primarily relied on human–computer interactions and visual interpretation from remote sensing to identify landslides. However, these methods are time-consuming, labor-intensive, subjective, and have a low level of accuracy in extracting data. An essential task in deep learning, semantic segmentation, has been crucial to automated remote sensing image recognition tasks because of its end-to-end pixel-level classification capability. In this study, to mitigate the disadvantages of existing landslide detection methods, we propose a multiscale attention segment network (MsASNet) that acquires different scales of remote sensing image features, designs an encoder–decoder structure to strengthen the landslide boundary, and combines the channel attention mechanism to strengthen the feature extraction capability. The MsASNet model exhibited an average accuracy of 95.13% on the test set from Bijie’s landslide dataset, a mean accuracy of 91.45% on the test set from Chongqing’s landslide dataset, and a mean accuracy of 90.17% on the test set from Tianshui‘s landslide dataset, signifying its ability to extract landslide information efficiently and accurately in real time. Our proposed model may be used in efforts toward the prevention and control of geological disasters.

Deep learning approaches for landslide information recognition: Current scenario and opportunities

Deep learning approaches for landslide information recognition: Current scenario and opportunities

Displacement field reconstruction in landslide physical modeling by using a terrain laser scanner – Part 2: Application and large strain/displacement and water effect analysis

Deformation analysis is fundamental in geotechnical modeling. Nevertheless, there is still a lack of an effective method to obtain the deformation field under various experimental conditions. In this study, we introduce a process–based physical modeling of a pile–reinforced reservoir landslide and present an improved deformation analysis involving large strains and water effects. We collect multi–period point clouds using a terrain laser scanner and reconstruct its deformation field through a point cloud processing workflow. The results show that this method can accurately describe the landslide surface deformation at any time and area by both scalar and vector fields. The deformation fields in different profiles of the physical model and different stages of the evolutionary process provide adequate and detailed landslide information. We analyze the large strain upstream of the pile caused by the pile installation and the consequent violent deformation during the evolutionary process. Furthermore, our method effectively overcomes the challenges of identifying targets commonly encountered in geotechnical modeling where water effects are considered and targets are polluted, which facilitates the deformation analysis at the wading area in a reservoir landslide. Eventually, combining subsurface deformation as well as numerical modeling, we comprehensively analyze the kinematics and failure mechanisms of this complicated object involving landslides and pile foundations as well as water effects. This method is of great significance for any geotechnical modeling concerning large-strain analysis and water effects.

Landslide Hazard Area Identification Using Smorph-Slope Morphology Method in Manado City

Landsliding is one of the natural process that shape the earth’s surface to achieve slope stability and it will be hazard if they threat human life. This study was conducted to identify potential landslide hazard in Manado region using SMORPH method with digital elevation model (DEM) data. ASTER GDEM data provide more detail landslide information than SRTM data because of the higher spatial resolution. Result shows that 716 ha approximately of land have high landslide vulnerability and 1,176 ha approximately of land have medium landslide vulnerability.

A methodological framework of landslide quantitative risk assessment in areas with incomplete historical landslide information

A methodological framework of landslide quantitative risk assessment in areas with incomplete historical landslide information

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.

Improving Railway Alignment Selection in Mountainous Areas with Complex Vegetation: A Multisource Data Landslide Identification Approach for Assisted Decision-Making Research

In order to provide important assistance for the scientific and effective route selection of future planned railways in the research area and to quickly and accurately identify the distribution range of landslides, thereby proactively mitigating the impact of geological hazards on railways under earthquake conditions, this study aims to shift the risk threshold for geological hazards and provide a scientific basis for the accurate planning and route selection of railways in mountainous areas. Jiuzhaigou was selected as the research area and postearthquake surface deformation information in the study area was obtained through Sentinel-1 satellite radar data. Based on Sentinel-2 optical remote sensing imagery, the changes in vegetation indices in the study area before and after the earthquake were analyzed in depth. The concept of vegetation index difference was proposed as a characteristic parameter for landslide information interpretation and a method combining surface deformation information was developed for landslide information interpretation. According to this method, the study area experienced a deformation subsidence of up to 14.93 cm under the influence of the earthquake, with some areas experiencing an uplift of approximately 6.0 cm. The vegetation index difference in the research area ranged from −1.83502 to 1.45366. The total number of landslides extracted is 12.034 km2 and 164 landslide points are marked, with an overall recognition accuracy of 92.6% and a Kappa coefficient of 0.876. The research results provide new research ideas for landslide information interpretation and can be used to assist in the decision-making of mountain railroad alignment options.

Landslides triggered by an extraordinary rainfall event in Central Italy on September 15, 2022

Timely and systematic collection of landslide information after a triggering event is pivotal for the definition of landslide trends in response to climate change. On September 15, 2022, a large part of central Italy, particularly Marche and Umbria regions, was struck by an anomalous rainfall event that showed characteristics of a persistent convective system. An extraordinary cumulated rainfall of 419 mm was recorded by a rain gauge in the area in only 9 h. The rainfall triggered 1687 landslides in the area affected by the peak rainfall intensity and caused widespread flash floods and floods in the central and lower parts of the catchments. In this work, we describe the characteristics of the landslides identified during a field survey started immediately after the event. Most of the mass movements are shallow, and many are rapid (i.e., debris flows, earth flows) and widely affecting the road network. Landslide area spans from a few tens of square meters to 105 m2, with a median value of 87 m2. Field evidence revealed diffuse residual risk conditions, being a large proportion of landslides located in the immediate vicinity of infrastructures. Besides reporting the spatial distribution of landslides triggered by an extreme rainfall event, the data collected on landslides can be used to make comparisons with the distribution of landslides in the past, validation of landslide susceptibility models, and definition of the general interaction between landslides and structures/infrastructures.

The ITAlian rainfall-induced LandslIdes CAtalogue, an extensive and accurate spatio-temporal catalogue of rainfall-induced landslides in Italy

Abstract. Italy is frequently hit and damaged by landslides, resulting in substantial and widespread disruptions. In particular, slope failures have a high impact on the population, communication infrastructure, and economic and productive sectors. The hazard posed by landslides requires adequate responses for landslide risk mitigation, with special attention to the risk to the population. In 2006 the Italian Department of Civil Protection, an office of the Prime Minister, commissioned the Research Institute for Geo-Hydrological Protection (Istituto di Ricerca per la Protezione Idrogeologica), a research institute of the Italian National Research Council, to carry out operational forecasting of rainfall-induced landslides. Collecting landslide information in a catalogue is a preliminary action toward landslide forecasting. The use of spatially and temporally inaccurate landslide catalogues results in uncertain and unreliable operational landslide forecasting. Consequently, accurate catalogues are needed to reduce the uncertainties, which are to some extent unavoidable. To this end, over the last 15 years many researchers have been involved in compiling a catalogue called ITALICA (ITAlian rainfall-induced LandslIdes CAtalogue), which currently lists 6312 records with information on rainfall-induced landslides that occurred over the Italian territory between January 1996 and December 2021. Overall, more than one-third of the catalogue has very high geographic accuracy (less than 1 km2) and hourly temporal resolution. In contrast, less than 2 % of the catalogue has low and very low geographical accuracy and daily temporal resolution. This makes ITALICA the largest catalogue of rainfall-induced landslides accurately located in space and time available in Italy. Without this high level of accuracy, the precipitation responsible for the initiation of landslides cannot be reliably reconstructed, thus making the prediction of landslide occurrence ineffective. ITALICA can be accessed at https://doi.org/10.5281/zenodo.8009366 (Brunetti et al., 2023). ITALICA's information on rainfall-induced landslides in Italy places a special emphasis on their spatial and temporal locations, making the catalogue especially suitable for defining the rainfall conditions capable of triggering future landslides in the Italian territory. This information is fundamental for decision-making in landslide risk management.

Spatial and temporal landslide distributions using global and open landslide databases

Landslide databases are a potential tool for the analysis of landslide susceptibility, hazard, and risk. Additionally, the spatio-temporal distribution of landslides and their correlation with their triggering factors are inputs that facilitate the evaluation of landslide prediction models and the determination of thresholds necessary for early warning systems (EWS). This study presents an analysis of four widely known global databases—the International Disaster database (EM-DAT), the Disaster Inventory System (DesInventar), the Global Landslide Catalog (GLC), and the Global Fatal Landslide database (GFLD)—which contain relevant landslide information for different regions of the world. These databases were analysed and compared by means of the spatio-temporal distributions of their records. Subsequently, these databases were merged and depurated to obtain a more robust database, namely the Unified Global Landslide Database (UGLD), with 161 countries, 37,946 landslides, and 185,753 fatalities registered between 1903 and 2020. The merging process among the databases resulted in a small number of repeated landslides, indicating that the databases collect very different landslide information and complement each other. Finally, an update of the spatial and temporal analysis of landslides in the world was performed with the new database, in which patterns, trends, and the main triggers were presented and analysed. The results obtained from the analysis of the UGLD database show the American and Asian continents as the continents with the highest number of landslides and associated fatalities, showing a bimodal and unimodal annual temporal pattern, respectively. Regarding the most frequent triggers of landslides, rainfall, anthropogenic intervention, and earthquakes stand out.

Analysis of Geological Hazard Susceptibility of Landslides in Muli County Based on Random Forest Algorithm

Landslides seriously threaten human life and property. The rapid and accurate prediction of landslide geological hazard susceptibility is the key to disaster prevention and mitigation. Traditional landslide susceptibility evaluation methods have disadvantages in terms of factor classification and subjective weight determination. Based on this, this paper uses a random forest model built using Python language to predict the landslide susceptibility of Muli County in western Sichuan and outputs the factor weight and model accuracy. The results show that (1) the three most important factors are elevation, distance from the road, and average annual rainfall, and the sum of their weights is 67.54%; (2) the model’s performance is good, with ACC = 99.43%, precision = 99.3%, recall = 99.48%, and F1 = 99.39%; (3) the landslide development and susceptibility zoning factors are basically the same. Therefore, this model can effectively and accurately evaluate regional landslide susceptibility. However, there are some limitations: (1) the landslide information statistical table is incomplete; (2) there are demanding requirements in terms of training concentration relating to the definition of landslide and non-landslide point sets, and the landslide range should be accurately delineated according to field surveys.

A comprehensive and up-to-date web-based interactive 3D emergency response and visualization system using Cesium Digital Earth: taking landslide disaster as an example

ABSTRACT As with the fast advances in the technologies of big Earth data and information communication, Web-based 3D GIS system has come a long way from a few years ago. These advances reflect in many aspects of 3D GIS such as higher real-time performance, enhanced interactivity, more realistic 3D visualization effect and improved user interface. This paper aims to present a comprehensive and up-to-date 3D Web GIS for Emergency Response using the current vue.js web application framework and the well-known Cesium API, taking landslide disaster as an example. Building upon recent advances in WebGL technology, we developed a suite of enhanced 3D spatial analysis functions, including interactive route planning, instant text/image/video messaging being incorporated into both 3D WebGL page and mobile GIS applications, and progressive 3D construction and AR visualization using LiDAR and camera over local emergency network or internet. Moreover, professional functions such as landslide susceptibility mapping, landslide monitoring, spatial temporal contingency plan management, landslide information management, personnel and equipment management, and communication are all implemented and integrated in the 3D GIS system. Most of the functions of the system are implemented using open-source projects, which is beneficial to the development of the 3D GIS research community.

Landslide Susceptibility mapping using random forest and extreme gradient boosting: A case study of Fengjie, Chongqing

Landslide susceptibility analysis can provide theoretical support for landslide risk management. However, some susceptibility analyses are not sufficiently interpretable. Moreover, the accuracy of many research methods needs to be improved. Therefore, this study can supplement these deficiencies. This study aims to research the evaluation effects of random forest (RF) and extreme gradient boosting (XGBoost) classifier models on landslide susceptibility, and to compare their applicability in Fengjie County, Chongqing, a typical landslide‐prone area in southwest of China. Firstly, 1624 landslides information from 1980 to 2020 were obtained through field investigation, and a geospatial database of 16 conditional factors had been constructed. Secondly, non‐landslide points were selected to form a complete data set and RF and XGBoost models were established. Finally, the area under the ROC curve (AUC) value, accuracy, and F‐score were used to compare the two models. The results show that even though both classifiers have a highly accurate evaluation of landslide susceptibility, the RF model performs better. In comparison, the RF model has a higher AUC value of 0.866, and its accuracy and F‐score are approximately 2% higher than XGBoost. The land use, elevation, and lithology of Fengjie County contribute to the occurrence of landslides. This is due to human engineering activities (such as land reclamation, and housing construction) resulting in low slope stability and landslides in widely distributed sandstone, siltstone, and mudstone layers owing to their low permeability and planes of weakness.

ShapeFormer: A Shape-Enhanced Vision Transformer Model for Optical Remote Sensing Image Landslide Detection

Landslides pose a serious threat to human life, safety, and natural resources. Remote sensing images can be used to effectively monitor landslides at a large scale, which is of great significance for pre-disaster warning and post-disaster assistance. In recent years, deep learning based methods have made great progress in the field of remote sensing image landslide detection. In remote sensing images, landslides display a variety of scales and shapes. In this paper, to better extract and keep the multi-scale shape information of landslides, a shape-enhanced vision transformer (ShapeFormer) model is proposed. For the feature extraction, a pyramid vision transformer (PVT) model is introduced, which directly models the global information of local elements at different scales. To learn the shape information of different landslides, a shape feature extraction branch is designed, which uses the adjacent feature maps at different scales in the PVT model to improve the boundary information. After the feature extraction step, a decoder with deconvolutional layers follows, which combines the multiple features and gradually recovers the original resolution of the combined features. A softmax layer is connected with the combined features to acquire the final pixel-wise result. The proposed ShapeFormer model was tested on two public datasets—the Bijie dataset and the Nepal dataset—which have different spectral and spatial characteristics. The results, when compared with those of some of the state-of-the-art methods, show the potential of the proposed method for use with multi-source optical remote sensing data for landslide detection.

DCI-PGCN: Dual-Channel Interaction Portable Graph Convolutional Network for Landslide Detection

Landslide, a kind of destructive natural disaster, often occurs in the mountainous areas of China. Landslide information instant collection plays an important role in taking appropriate remedial measures and personnel evacuation. In recent years, the use of Convolutional Neural Network (CNN) for landslide regional detection achieved good performance, however, most CNN-based methods had no regard for the internal connection of the cover materials in the disaster occurrence area. Moreover, the information revealed by the internal deformation features was ignored, and the same surface object in the image presents different features under different illumination, environment and resolution, which makes it difficult to extract the structural features of landslide images. In this paper, we propose a novel graph convolutional network for landslide detection, inspired by attention mechanism’s ability to focus on selective information supplemented with both different channels. The global maximum node connection strategy with positive and negative connectivity makes the Graph Convolution Network (GCN) more portable, which is used as the basic unit of graph feature propagation to construct a multi-layer residual connection module. In order to learn interactively and spread graph information, channel dimension is added to make the boundary of features between classes more discriminative. Extensive experiments on Sichuan province and Bijie landslide datasets show that our proposed method outperforms other detection models and achieves high precision and accuracy. In addition, we also carried out landslide detection for Zhaotong of Yunnan Province on GF-2 original images to prove the effectiveness and applicability of the algorithm.

Distribution of large- and medium-scale loess landslides induced by the Haiyuan Earthquake in 1920 based on field investigation and interpretation of satellite images

Abstract Studying the distribution law and influencing factors of coseismic landslides has important scientific significance and engineering value for understanding the mechanism of seismic landslides and predicting the occurrence of seismic landslides. After a hundred years, these large- and medium-scale landslides induced by the 1920 Haiyuan earthquake are still well-preserved and have extremely high academic research value. About 620 loess seismic landslides induced by the Haiyuan earthquake in 1920 were investigated on site. On this basis, the shape differences between seismic landslides and gravity landslides were summarized; 605 landslides were identified by satellite images, and the Haiyuan earthquake-induced loess landslide database containing seismic information and landslide information was established. The distribution law and morphological characteristics of large- and medium-sized landslides induced by the Haiyuan earthquake were systematically counted according to the conditions of the landslide-intensive area, intensity, and fault upper and lower plates. The influencing factors of loess earthquake landslides were summarized, and the following conclusions were obtained: (1) 1,225 large- and medium-sized landslides were induced by the Haiyuan earthquake in 1920. These landslides have the characteristics of long sliding distance, large single scale, and strong disaster-causing. They are mainly distributed in three concentrated areas of Xiji, Haiyuan, and Pengyang. The landslide morphological characteristics of the three landslide-intensive areas are different because the landslide sliding mechanism caused by topography and lithology is different. (2) The landslide distribution has obvious clustering, zonation, and directivity, and has an obvious river distance effect and fault hanging wall effect. (3) The internal influencing factors such as stratum lithology, topography, fault location, and direction, and the role of water control the occurrence location and scale of landslide. The external factor of an earthquake is an important incentive and control factor for landslide occurrence.

"풍화 산지에서 토사 재해 발생에 영향을 미친 요소들에 대한 연구 - 2020년 충청북도 북부지역을 사례로 -"

Among the reports for damaged area by landslide in Chungju, Jecheon, and Danyang-gun in the northern region of Chungcheongbuk-do, where damage occurred due to heavy rainstorms from July 28 to August 11, 2020, the coordinates of the locations has been confirmed at 438 places. Analyses were conducted for those spots using the landslide information system of the Korea Forest Service, geological information from geobig data, KAKAO map, and satellite images from Google Earth. The most common bedrock in the affected area was granite, followed by gneiss, limestone and other sedimentary rocks, slate and phyllite. As for the slope, sedimentary rocks such as granite and limestone have the steepest slope, and gneiss, granite, slate and phyllite follows it. It seems that the impact by landslide on these types of rocks depended on slope and weathering pattern. The damage type examined in 375 sites out of 438 were due to erosion and sediment runoff from the slope, and debris flow occurred in 63 places. In 31 of them, it was found that the outflow of soil from the slope was connected to the debris flow. As a result of analyzing the rock types, most of the debris flow in the granite area appeared to be ‘runoff-dominant erosion’. It seems that ‘infiltration-triggered soil slip’ was strong in the limestone area. In this study, using Dynamic World, Google’s land cover change tracking service, surface cover changes were tracked between 2018 and 2021-2022 before landslide-related damage occurred, and it was found that small-scale vegetation changes in land use near the study area. It can be seen that changes in Korea’s forestry policy have had a significant impact on this. In particular, significant landslides have occurred after the considerable amount of time followed by anthropogenic change of vegetation and building of forest roads. Therefore, it seems that continuous management is essential to prevent landslide disasters after anthropogenic changes in land cover. In addition, field studies on the condition of soil weathering in mountainous areas and the process of material movement on slopes are essential.