Search

Following the earthquake, prompt evaluation of the distribution of coseismic landslides and estimation of potential disaster losses are crucial for emergency response and resettlement planning. The Luding earthquake of 2022 offers a valuable opportunity to conduct a rapid assessment of coseismic landslides using various models. In this study, we utilize the Logistic Regression (LR)-based Xu2019 model, a new-generation model developed in China, alongside the Newmark model to perform the rapid hazard assessment of coseismic landslides. Assessing the accuracy and applicability of these two models based on the coseismic landslides from the Luding earthquake, we find that within intensity area of IX, the high probability area identified by the Newmark model aligns closely with the actual distribution of landslides. However, the Newmark model's prediction is overestimated in the intensity area of VIII. For the Xu2019 model, the prediction results are in good agreement with the distribution of actual landslides. Most landslides are located in high probability areas, such as Detuo town, Wandong, and Xingfu villages, indicating that the model has a higher prediction accuracy. Overall, two models have good practical utility in emergency hazard assessment of coseismic landslides. However, the Newmark model requires multi-input parameters and the assignment of these parameters will increase the uncertainty and subjectivity in the practical application of the modeling assessment. [ABSTRACT FROM AUTHOR]

Objective Strong earthquake-induced landslides are characterized by large number, wide distribution and large scale, and seriously threaten people's lives and property. Landslide susceptibility mapping (LSM) can quickly predict the spatial distribution of prone areas, which is highly important for reducing the risk of post-earthquake disasters. However, in the studies of coseismic landslide LSMs, how to select negative landslide samples and integrate machine learning models to improve the evaluation accuracy still needs further investigation. Methods In this study, the landslides induced by the Wenchuan earthquake in mountainous areas are selected as a case study. First, 10 landslide influencing factors, such as topography, geological environment, and seismic parameters, are selected to analyse the spatial distribution of landslides. Then, collinearity analysis is used to test data redundancy, nonnegative sample points from the sampling strategies are randomly selected in the extremely low susceptibility regions by the frequency ratio (FR) method. Finally, gradient boosting decision tree (GBDT), random forest (RF), and their optimal models are used to predict coseismic landslide susceptibility, conduct a comparative study of the models and carry out an accuracy assessment. Results The results show that ① the spatial distribution of landslides is controlled by multiple factors, and ② the accuracy of the models is FR-RF(AUC=0.943)>FR-GBDT(AUC=0.926)>RF(AUC=0.901)>GBDT(AUC=0.856). ③ Selecting negative landslide samples in low susceptibility areas could significantly improve the accuracy of LSMs. Conclusion The research results can provide a reference for selecting negative landslide samples and constructing evaluation models, as well as for providing theoretical support for post-earthquake disaster prevention and mitigation.

Earthquake-triggered landslides (ETLs) feature large quantities, extensive distributions, and enormous losses to human lives and critical infrastructures. Near-real spatial prediction of ETLs can rapidly predict the locations of coseismic landslides just after a violent earthquake and is a vital technical support for emergency response. However, near-real prediction of ETLs has always been a great challenge with relatively low accuracy. This work proposes an ensemble prediction model of EnPr by integrating machine learning tree models and a deep learning convolutional neural network. EnPr exhibits relatively strong prediction and generalization performance and achieves relatively accurate prediction of ETLs. Six great seismic events occurring from 2008 to 2022 on the southeastern margin of the Tibetan Plateau are selected to conduct ETL prediction. In a chronological order, the 2008 Ms 8.0 Wenchuan, 2010 Ms 7.1 Yushu, 2013 Ms 7.0 Lushan, and 2014 Ms 6.5 Ludian earthquakes are employed for model training and learning. The 2017 Ms 7.0 Jiuzhaigou and 2022 Ms 6.1 Lushan earthquakes are adopted for ETL prediction. The prediction accuracy merits of ACC and AUC attain 91.28% and 0.85, respectively, for the Jiuzhaigou earthquake. The values of ACC and AUC achieve 93.78% and 0.88, respectively, for the Lushan earthquake. The proposed EnPr algorithm outperforms the algorithms of XGBoost, random forest (RF), extremely randomized trees (ET), convolutional neural network (CNN), and Transformer. Moreover, this work reveals that seismic intensity, high and steep relief, pre-seismic fault tectonics, and pre-earthquake road construction have played significant roles in coseismic landslide occurrence and distribution. The EnPr model uses globally accessible open datasets and can therefore be used worldwide for new large seismic events in the future. [ABSTRACT FROM AUTHOR]

This study uses automatic linear regression (LINEAR) and artificial neural network (ANN) models to statistically analyze the area of landslides triggered by the 2021 SW Chelgard earthquake (ML = 6) based on controlling parameters. We recorded and mapped the number of 632 landslides into four groups (based on the Hungr et al. 2014): rock avalanche-rock fall, debris avalanche-debris flow, rock slump, and slide earth flow-soil slump using remote sensing method, satellite images (before and after the earthquake), and field observation. The spatial distribution of landslides showed that the highest values of the landslide area percentage (LAP %) and of the landslide number density (LND, N/km2) occurred in the northern part of the fault on the hanging wall. The ANN models with R2 = 0.51–0.80 provided more accurate predictions of landslide area (LA, m2) than the LINEAR models, with R2 = 0.40–0.61 using multiple parameters. The LINEAR models revealed that the most influential controlling parameters for landslides were the topographic factors and ANN models showed that seismic parameters are effective on the coseismic landslides (e.g., the distance from the epicenter on the rock slumps; the PGA on debris avalanches- debris flow; the distance from the rupture surface of the fault and Ia on the rock avalanches-rockfall and slide earth flow-soil slump). Therefore, the classification of coseismic landslides can be helpful for predicting the LA more accurately and better understanding the failure mechanism. [ABSTRACT FROM AUTHOR]

Near real-time spatial prediction of earthquake-induced landslides (EQILs) can rapidly forecast the occurrence position of widespread landslides just after a violent earthquake; thus, EQIL prediction is very crucial to the 72-hour ‘golden window’ for survivors. This work focuses on a series of earthquake events from 2008 to 2022 occurring in the Tibetan Plateau, a famous seismically-active zone, and proposes a novel interpretable self-supervised learning (ISeL) method for the near real-time spatial prediction of EQILs. This new method innovatively introduces swap noise at the unsupervised mechanism, which can improve the generalization performance and transferability of the model, and can effectively reduce false alarm and improve accuracy through supervised fine-tuning. An interpretable module is built based on a self-attention mechanism to reveal the importance and contribution of various influencing factors to EQIL spatial distribution. Experimental results demonstrate that the ISeL model is superior to the excellent state-of-the-art machine learning and deep learning methods. Furthermore, according to the interpretable module in the ISeL method, the critical controlling and triggering factors are revealed. The ISeL method can also be applied in other earthquake-frequent regions worldwide because of its good generalization and transferability.

Automatic recognition of numerous coseismic landslides after a violent earthquake is crucial for emergency rescue and post-disaster reconstruction. Currently, deep learning techniques have achieved state-of-the-art performance in coseismic landslide recognition. However, Convolutional Neural Networks (CNNs) often lose detailed information during downsampling and cannot adequately learn changeable shapes, colors, and sizes of coseismic landslides. In addition, complicated backgrounds, e.g., bare slopes and dry riverbeds, are easily misidentified as coseismic landslides. Focusing on the above difficulties, this work proposes a Gated Dual-Stream Convolutional Neural Network (GDSNet) for landslide recognition, which contains two branches and a feature fusion module. One branch, called as CPSConv, can extract detailed landslide information of shapes, sizes, spectra, and textures and guarantee the accurate identification of small landslides. Another branch utilizes a gated convolution strategy to adjust feature weights and importance and to enhance landslide features and suppress background features. The feature aggregation and fusion module fuses the features from two branches to effectively improve the recognition accuracy of coseismic landslides. The GDSNet is applied to the landslide recognition of four earthquakes by model training and testing. In test dataset, compared with 9 state-of-the-art models, the mIoU, F1, Kappa coefficient values by the GDSNet are improved by at least 12.30%, 8.12%, and 16.25%, respectively. The recognition accuracy of small landslides is improved by 1.08%-37.68% than the other 9 deep learning models.

On August 8, 2017, a magnitude Mw6.5 (Ms7.0) earthquake occurred in Jiuzhaigou County, Aba Prefecture, in the northern part of Sichuan Province, China, with a focal depth of 20 km and an epicenter located at (33.2°N, 103.8°E). Due to the significant magnitude of the earthquake, a large number of coseismic landslides were triggered. Despite previous research conducted by experts on the landslides caused by the Jiuzhaigou earthquake, the actual number of landslides has been severely underestimated in the previously published papers. Through field surveys and visual interpretation of high-resolution remote sensing images before and after the mainshock, we have established a detailed inventory of earthquake-induced landslides. The results indicate that the event caused a minimum of 9428 landslides covering a total area of 18.82 km2. These landslides are mainly distributed in the IX intensity area of the earthquake. The landslides mainly consist of medium-sized landslides and debris flows. They predominantly occur in areas with an altitude ranging from 2600 m to 3600 m, with slopes greater than 30° and facing east and southeast. The Lower Carboniferous and Middle Carboniferous formations are more prone to triggering landslides, and landslides are more concentrated within 1 km of roads and in forested areas. Additionally, as the distance from roads and the epicenter increases, the values of LAP and LND decrease, indicating a positive correlation between the two. There are more landslides within 2 km from the fault and within a range of 6 km–9 km from the epicenter. In conclusion, this study provides a comprehensive landslide inventory with broader coverage and increased accuracy. It also conducts a comprehensive analysis of the spatial distribution patterns of landslides. This contributes to a deeper understanding of the causes of coseismic landslides and further research on the impact of landslides in affected areas.

Earthquake-triggered landslides (ETLs) feature large quantities, extensive distributions, and enormous losses to human lives and critical infrastructures. Near-real spatial prediction of ETLs can rapidly predict the locations of coseismic landslides just after a violent earthquake and is a vital technical support for emergency response. However, near-real prediction of ETLs has always been a great challenge with relatively low accuracy. This work proposes an ensemble prediction model of EnPr by integrating machine learning tree models and a deep learning convolutional neural network. EnPr exhibits relatively strong prediction and generalization performance and achieves relatively accurate prediction of ETLs. Six great seismic events occurring from 2008 to 2022 on the southeastern margin of the Tibetan Plateau are selected to conduct ETL prediction. In a chronological order, the 2008 Ms 8.0 Wenchuan, 2010 Ms 7.1 Yushu, 2013 Ms 7.0 Lushan, and 2014 Ms 6.5 Ludian earthquakes are employed for model training and learning. The 2017 Ms 7.0 Jiuzhaigou and 2022 Ms 6.1 Lushan earthquakes are adopted for ETL prediction. The prediction accuracy merits of ACC and AUC attain 91.28% and 0.85, respectively, for the Jiuzhaigou earthquake. The values of ACC and AUC achieve 93.78% and 0.88, respectively, for the Lushan earthquake. The proposed EnPr algorithm outperforms the algorithms of XGBoost, random forest (RF), extremely randomized trees (ET), convolutional neural network (CNN), and Transformer. Moreover, this work reveals that seismic intensity, high and steep relief, pre-seismic fault tectonics, and pre-earthquake road construction have played significant roles in coseismic landslide occurrence and distribution. The EnPr model uses globally accessible open datasets and can therefore be used worldwide for new large seismic events in the future.

The June 10, 2022 Ms5.8, Ms6.0 and Ms5.2 earthquake swarm of Maerkang, Sichuan, China has triggered a lot of landslides. The primary purpose of this paper is to prepare a detailed coseismic landslides inventory and ancient landslide inventory in the earthquake swarm-affected area. Based on pre- and post-quake high-resolution optical satellite images (planet), we delineated 650 individual coseismic landslides in the VI-intensity area and above, occupying an area of about 1.2 ​km2. The largest landslide covers about 70217 ​m2 in horizontal projection, while the smallest one is just 81 ​m2. In addition, based on the series of high-precise images on the Google Earth Platform, 759 ancient landslides were also delineated with an area of 117 ​km2. The maximum area reaches about 2 ​km2 and the minimum is 2580 ​m2. The two inventories provide a basis for landslide spatial distribution and hazard assessment, disaster prevention and mitigation of earthquake-triggered landslides in similar regions. We opened these two landslide inventories to facilitate researchers to carry out more in-depth work.

Accurate assessment of seismic landslides hazard is a prerequisite and foundation for post-disaster relief of earthquakes. An Ms 5.7 earthquake occurring on September 7, 2012, in Yiliang County, Yunnan Province, China, triggered hundreds of landslides. To explore the characteristics of coseismic landslides caused by this moderate-strong earthquake and their significance in predicting seismic landslides regionally, this study uses an artificial visual interpretation method based on a planet image with 5-m resolution to obtain the information of the coseismic landslides and establishes a coseismic landslide database containing data on 232 landslides. Nine influencing factors of landslides were selected for this study: elevation, relative elevation, slope angle, aspect, slope position, distance to river system, distance to faults, strata, and peak ground acceleration. The real probability of coseismic landslide occurrence is calculated by combining the Bayesian probability and logistic regression model. Based on the coseismic landslides, the probabilities of landslide occurrence under different peak ground acceleration are predicted using a logistic regression model. Finally, the model established in this paper is used to calculate the landslide probability of the Ludian Ms 6.5 earthquake that occurred in August 2014, 78.9 km away from the macro-epicenter of the Yiliang earthquake. The probability is verified by the real coseismic landslides of this earthquake, which confirms the reliability of the method presented in this paper. This study proves that the model established according to the seismic landslides triggered by one earthquake has a good effect on the seismic landslides hazard assessment of similar magnitude, and can provide a reference for seismic landslides prediction of moderate-strong earthquakes in this region. [ABSTRACT FROM AUTHOR]

The rapid and accurate extraction of wide-area coseismic landslide locations is critical in earthquake emergencies. At present, the extraction of coseismic landslides is mainly based on post-earthquake site investigation or the interpretation of human–computer interactions based on remote sensing images. However, the identification efficiency is low, which seriously delays the earthquake emergency response. On the basis of the available multisource and multiscale remote sensing data, numerous studies have been carried out on the methods of coseismic landslide extraction, such as pixel analysis, object-oriented analysis, change detection, and machine learning. However, the effectiveness of coseismic landslide extraction was low in wide areas with complex topographic and geomorphic backgrounds. Therefore, this paper offers a comprehensive study of the factors influencing coseismic landslides and researches rapid and accurate wide-area coseismic landslide extraction methods with multisource remote sensing and geoscience technology. These techniques include digital elevation modeling (DEM) and its derived slopes and aspects. An embedded multichannel spectral–topographic feature fusion model for coseismic landslide extraction based on DeepLab V3+ is proposed, and a knowledge-enhanced deep learning information extraction method integrating geological knowledge is formed. Using the Jiuzhaigou Ms7.0 earthquake (seismic intensity VIII) in Sichuan Province, China, a comparison of landslide extraction models and strategies is carried out. The results show that the model proposed in this paper achieves the best balance in the accuracy and efficiency of wide-area extractions. Using multiple feature data of coseismic landslides, the problem of mixed pixels is solved. The rate of the misidentification of landslides as clouds, snow, buildings, and roads is significantly lower than in other methods. The identified landslide boundaries are smoother and more accurate, and the connectivity is better. Compared with other methods, ours can more accurately eliminate landslides not triggered by the Jiuzhaigou earthquake. While using the image block strategy to ensure extraction efficiency, it also improves the extraction accuracy of wide-area coseismic landslides in complex backgrounds. [ABSTRACT FROM AUTHOR]

On 5th September 2022，an Mw 6.6 earthquake struck Luding and it is the largest earthquake on the Xianshuihe fault, Eastern Tibet in the past 40 years. It is of great significance for investigating the tectonic activity and assessing future seismic hazards in the region. Methods: In this study, we used Sentinel-1 and ALOS-2 synthetic aperture radar (SAR) images to retrieve coseismic surface displacements and then to determine the fault geometry parameters and slip distribution with a dislocation model in an elastic halfspace. Results: The results show that the 2022 Luding Mw 6.6 earthquake is a left-lateral sliding event with a maximum surface displacement of 15 cm and 21 cm for the Sentinel-1 and ALOS-2 ra‐ dar images, respectively. The fault ruptured along an NNW-SSE strike, and westward at a dip of 72°. The slip was concentrated at depths of 4-12 km with a maximum fault slip of 2.23 m occurring at a depth of 5.8 km. Conclusions: By analyzing the distribution of the coseismic landslides interpreted by optical images, we found that the coseismic landslides were mainly located on the west side of the fault, which is consistent with the aftershock distribution and can be considered due to hanging wall effects. [ABSTRACT FROM AUTHOR]

Objectives: On 5th September 2022, an Ms 6.8 earthquake struck the Luding County, Ganzi Prefecture, Sichuan Province, China. This earthquake triggered extensive geological hazards in the moun‐ tainous area, leading to serious casualties. Rapidly and accurately obtaining the spatial distribution of the in‐ duced geological hazards is crucial for emergency decision-making and rescue after an earthquake. Methods: Based on the global coseismic landslide database and deep learning algorithm, this paper built a near real-time prediction model of spatial distribution probability of coseismic landslides, and obtained the prediction results of the geological hazards induced by the Luding earthquake within 2 hours after the event. Through the post-earthquake unmanned aerial vehicle （UAV） and satellite remote sensing images, ma‐ chine learning and deep learning algorithms were used to realize the automated recognition of large-scale geological hazards. A total of 3 633 earthquake-induced landslides with an area of 13.78 km2 were interpreted. Finally, the model was optimized by integrating these landslide data, and the prediction results of coseismic landslides with a broader area and higher accuracy were achieved. Results: The results show that the coseis‐ mic landslide prediction model can realize a rapid capture of spatial distribution of post-earthquake geologi‐ cal hazards, filling the blank period before the acquisition of post-earthquake remote sensing images and providing support for post-disaster emergency rescue. Conclusions: Intelligent identification technologies based on UAV and satellite remote sensing images are effective means to rapidly obtain the vital informa‐ tion of large-scale geological hazards. The achievements obtained in this paper played an important role in the emergency rescue after the Luding earthquake. [ABSTRACT FROM AUTHOR]

Near real-time spatial prediction of earthquake-induced landslides (EQILs) can rapidly forecast the occurrence position of widespread landslides just after a violent earthquake; thus, EQIL prediction is very crucial to the 72-hour 'golden window' for survivors. This work focuses on a series of earthquake events from 2008 to 2022 occurring in the Tibetan Plateau, a famous seismically-active zone, and proposes a novel interpretable self-supervised learning (ISeL) method for the near real-time spatial prediction of EQILs. This new method innovatively introduces swap noise at the unsupervised mechanism, which can improve the generalization performance and transferability of the model, and can effectively reduce false alarm and improve accuracy through supervised fine-tuning. An interpretable module is built based on a self-attention mechanism to reveal the importance and contribution of various influencing factors to EQIL spatial distribution. Experimental results demonstrate that the ISeL model is superior to the excellent state-of-the-art machine learning and deep learning methods. Furthermore, according to the interpretable module in the ISeL method, the critical controlling and triggering factors are revealed. The ISeL method can also be applied in other earthquake-frequent regions worldwide because of its good generalization and transferability. [ABSTRACT FROM AUTHOR]

Near real-time prediction of earthquake-triggered landslides can rapidly forecast the spatial distribution of coseismic landslides just after a great earthquake, and provide effective support for emergency response. However, the prediction of earthquake-triggered landslides has always been a great challenge because of low accuracy and high false alarms. This work proposes a novel fuzzy deep learning (FuDL) model for near real-time earthquake-triggered landslide spatial prediction. Fuzzy learning theory is for the first time employed in earthquake-triggered landslide prediction. The FuDL has high generalization and robustness, effectively improving the accuracy of earthquake-triggered landslide prediction. Eighteen earthquake-triggered landslide inventories worldwide from 2008 to 2022 are employed to conduct ETL prediction. According to the chronological order, 15 earthquake-triggered landslides from 2008 to 2018 are adopted to train the FuDL model, and 3 earthquake-triggered landslides from 2019 to 2022 are utilized for near real-time earthquake-triggered landslide prediction. Furthermore, this work reveals that ground movement, relatively steep and high topography, and strong seismic intensity are critical factors affecting the spatial distribution of earthquake-triggered landslides. In addition, this work conducted a detailed analysis of the distribution patterns of earthquake-triggered landslides on a global scale. • A novel FuDL model is suggested and achieves high accuracy and good generalization. • FuDL model outperforms the state-of-the-art machine learning or deep learning models. • Topography, ground shake, and seismic intensity dominate the distribution of ETLs. [ABSTRACT FROM AUTHOR]

At 17:00 (UTC+8) on 1 June 2022, an Ms 6.1 reverse earthquake struck Lushan County, Ya'an City, Sichuan Province. This earthquake event had a focal depth of 10 km and the epicenter was located at 30.37° N and 102.94° E. The purpose of this study is to document a comprehensive coseismic landslide inventory for this event and analyze the distribution pattern and factors controlling the landslides. After careful visual interpretations, this quake event was determined to have in total triggered about 2352 landslides in an area of 3900 km2, including both shallow disrupted landslides and collapses, for which the spatial distribution was statistically related to regional topography, geology, and seismicity. Notably, a vast majority of the landslides were located on the NW plate of the seismogenic fault, and were distributed in the area with a seismic intensity of VII. In addition, coseismic landslides were more likely to appear in areas with high altitude, relief, and large slope. The landslide area density (LAD) increased with an increase in the above factors and is explained by an exponential relationship, indicating that the occurrence of coseismic landslides in this area was more easily affected by topographic factors than seismic factors. Most small-scale landslides were clustered in the ridge area, which shows the seismic amplification effects of mountain slopes. Due to the impact of seismic wave propagation direction, hillslopes facing northeast-east (NE-E) were more prone to collapse than southwest-facing ones. Based on the distribution pattern of the landslides, we suggest that the seismogenic fault of this event was NW dipping. These findings indicate that it is effective to identify the dipping of seismogenic faults using the spatial distribution pattern of coseismic landslides. [ABSTRACT FROM AUTHOR]

On 5 September 2022, an Mw 6.6 earthquake struck Luding county, Sichuan province, China, which triggered a large number of geohazards such as landslides and collapses, leading to serious casualties and economic losses. Rapid access to landslide susceptibility and the actual distribution maps of coseismic landslides are critical for disaster management. In this study, we combined optical remote sensing image interpretation with the Generic Atmospheric Correction Online Service (GACOS) assisted InSAR stacking technique to map pre-event landslides, and then used the detected landslide dataset to generate a landslide susceptibility map in the earthquake-affected area using a machine learning algorithm. In addition, we used multi-source optical remote sensing images to establish an inventory of 2 692 coseismic landslides and analyze its relationships with topography, seismic and geological factors. Our results showed that 70.2% of the total coseismic landslide area（47 km~²） were within the areas with moderate or higher susceptibility. These coseismic landslides were mainly distributed in elevations of 1 200-2400 m, with slopes of 35°-50°, 4-20 km from the epicenter, within 1 km of faults and with a lithology of sericites silt slate. The coseismic landslides also damaged at least 10.34 km of roads. It is believed that this research can provide data support for the assessment and prevention of geohazards in the earthquake-affected areas. [ABSTRACT FROM AUTHOR]

To date, a number of studies have been conducted to examine the relationship between seismic ground motion and coseismic landslides. However, the impacts of seismic ground motion on coseismic landslide occurrence are not fully understood owing to the poor spatial resolution of seismic ground motion data. Recently, seismic observation research has expanded with the use of satellite technology, as crustal deformation can be observed using pairs of SAR (synthetic aperture radar) satellite data. With this technique, obtaining information regarding the ground surface displacement induced by earthquakes is possible at a high spatial resolution, without the need for interpolation or extrapolation. In this study, we focus specifically on the interrelated impacts of seismic ground motion and topography on coseismic landslide occurrence, which has previously been difficult to detect. Using high-resolution ground surface displacement from SAR data, we examine these interrelated impacts in detail and assess coseismic landslide occurrence based on seismic ground motion and topography. Results show that the developed formula accurately reproduces coseismic landslide occurrence and that the impact behaviors of the two factors on landslide occurrence are different. Finally, based on the new formula, we suggest two different trends for the attenuation of seismic ground motion and topography related to coseismic landslide occurrence. [ABSTRACT FROM AUTHOR]

Coseismic landslides are common secondary earthquake geohazards in meizoseismal areas. Newmark sliding block permanent displacement method has been widely adopted to develop regional coseismic landslide hazard maps. However, uncertainties from the slope parameters (e.g., cohesion, pore water pressure, and block thickness) are not commonly considered in the ground displacement predictions. This study proposes a novel framework that consists of two uncertainty assessment methods of Monte Carlo and logic tree simulations (MCS and LTS) with seven different displacement regression functions to predict the regional coseismic landslides' permanent displacement. Compared with the existing methods, the proposed framework is argument-driven, avoiding huge number of repetitive simulations. The Jiuzhaigou earthquake, in China, is considered as an illustrative example to compare the performance of the framework with considered regression functions. The corresponding results show that using LTS, with a certain regression function, leads to better predictions compared to using MCS. It is demonstrated that the proposed framework can provide a meaningful measure for making informed decisions to diminish the potential risk of earthquake induced landslides, and/or generating emergency strategies to mitigate post-earthquake consequences. It should be noted that the application of the proposed method for deposits where the soil strength parameter values do not fit the normal distribution may be limited as only normal distribution for soil strengths is considered in this study. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Engineering Geology / Gongcheng Dizhi Xuebao is the property of Journal of Engineering Geology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

At 3:08 on 6 September 2018 (UTC +9), massive landslides were triggered by an earthquake of Mw 6.6 that occurred in Hokkaido, Japan. In this paper, a coseismic landslide inventory that covers 388 km2 of the earthquake-impacted area and includes 5828 coseismic landslides with a total landslide area of 23.66 km2 was compiled by using visual interpretations of various high-resolution satellite images. To analyze the spatial distribution and characteristics of coseismic landslides, five factors were considered: the peak ground acceleration (PGA), elevation, slope gradient, slope aspect, and lithology. Results show more than 87% of the landslides occurred at 100 to 200 m elevations. Slopes in the range of 10~20°are the most susceptible to failure. The landslide density of the places with peak ground acceleration (PGA) greater than 0.16 g is obviously larger than those with PGA less than 0.02 g. Compared with the number and scale of coseismic landslides caused by other strong earthquakes and the mobility of the coseismic landslides caused by the Haiyan and Wenchuan earthquakes, it was found that the distribution of coseismic landslides was extremely dense and that the mobility of the Hokkaido earthquake was greater than that of the Wenchuan earthquake and weaker than that of the Haiyuan earthquake, and is described by the following relationship: L = 18.454 ∗ H0.612. Comparative analysis of coseismic landslides with similar magnitude has important guiding significance for disaster prevention and reduction and reconstruction planning of landslides in affected areas. [ABSTRACT FROM AUTHOR]

Yogyakarta is one of the large cities in Central Java, located on Java Island, Indonesia. The city, and the Pleret sub-district, where the study has taken place, is prone to earthquake hazards, because it is close to several seismically active zones, such as the Sunda Megathrust and the active fault known as the Opak Fault. Since a devastating earthquake of 2006, the population of the Pleret sub-district has increased significantly. Thus, the housing demand has increased, and so is the pace of low-cost housing that does not meet earthquake-safety requirements, and furthermore are often located on unstable slopes. The local alluvial material covering a jigsaw of unstable blocks and complex slope is conditions that can amplify the negative impacts of earthquakes. Within this context, this study is aiming to assess the multi-hazards and risks of earthquakes and related secondary hazards such as ground liquefaction, and coseismic landslides. To achieve this, we used geographic information systems and remote sensing methods supplemented with outcrop study and existing seismic data to derive shear-strain parameters. The results have revealed the presence of numerous uncharted active faults with movements visible from imagery and outcrops. show that the middle part of the study area has a complex geological structure, indicated by many unchartered faults in the outcrops. Using this newly mapped blocks combined with shear strain data, we reassessed the collapse probability of buildings that reach level >0.75 near the Opak River, in central Pleret sub-district. Classifying the buildings and from population distribution, we could determine that the highest risk was during nighttime as the buildings susceptible to fall are predominantly housing buildings. The secondary hazards follow a slightly different distribution with a concentration of risks in the West.

Non-landslide samples play a crucial role in landslide susceptibility mapping (LSM), although unsuitable sampling methods may degrade the performance of the prediction model. The primary objectives of this study are to explore the influence of the traditional buffer-controlled sampling method on model performance and to propose a Newmark-based sampling approach for coseismic landslides. The Jiuzhaigou meizoseismal region of China is selected as the region of study. Six sample datasets are constructed for three machine learning models, namely a deep neural network (DNN), logistic regression (LR), and a support vector machine (SVM). The samples cover two scenarios: scenario-BZ, a set of samples created using different buffer distances, and scenario-LD, a set of non-landslide samples created by the Newmark-based method. Intriguingly, the results indicate that the area under the curve (AUC) is positively correlated with the buffer distance in scenario-BZ (DNN: 0.894–0.979, SVM: 0.894–0.981, LR: 0.797–0.889), but gentle valleys in the buffer zone are assigned over-conservative susceptibility values while the probability of landslides in steep mountains outside the buffer zone is underestimated. In contrast, all models assign more reasonable susceptibility values in scenario-LD (AUC values of 0.969, 0.969, and 0.931 for the DNN, LR, and SVM models, respectively). These results suggest that the landslide susceptibility obtained by the traditional buffer-controlled method may be inaccurate, despite the prediction model achieving excellent performance. The proposed approach can therefore provide insights into coseismic landslide susceptibility in other earthquake regions. [ABSTRACT FROM AUTHOR]

Copyright of Geography & Geographic Information Science is the property of Geography & Geo-Information Science Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The rapid and accurate extraction of wide-area coseismic landslide locations is critical in earthquake emergencies. At present, the extraction of coseismic landslides is mainly based on post-earthquake site investigation or the interpretation of human–computer interactions based on remote sensing images. However, the identification efficiency is low, which seriously delays the earthquake emergency response. On the basis of the available multisource and multiscale remote sensing data, numerous studies have been carried out on the methods of coseismic landslide extraction, such as pixel analysis, object-oriented analysis, change detection, and machine learning. However, the effectiveness of coseismic landslide extraction was low in wide areas with complex topographic and geomorphic backgrounds. Therefore, this paper offers a comprehensive study of the factors influencing coseismic landslides and researches rapid and accurate wide-area coseismic landslide extraction methods with multisource remote sensing and geoscience technology. These techniques include digital elevation modeling (DEM) and its derived slopes and aspects. An embedded multichannel spectral–topographic feature fusion model for coseismic landslide extraction based on DeepLab V3+ is proposed, and a knowledge-enhanced deep learning information extraction method integrating geological knowledge is formed. Using the Jiuzhaigou Ms7.0 earthquake (seismic intensity VIII) in Sichuan Province, China, a comparison of landslide extraction models and strategies is carried out. The results show that the model proposed in this paper achieves the best balance in the accuracy and efficiency of wide-area extractions. Using multiple feature data of coseismic landslides, the problem of mixed pixels is solved. The rate of the misidentification of landslides as clouds, snow, buildings, and roads is significantly lower than in other methods. The identified landslide boundaries are smoother and more accurate, and the connectivity is better. Compared with other methods, ours can more accurately eliminate landslides not triggered by the Jiuzhaigou earthquake. While using the image block strategy to ensure extraction efficiency, it also improves the extraction accuracy of wide-area coseismic landslides in complex backgrounds.

The coseismic landslide is one of the important hazard phenomena in the hilly and seismically active mountainous region. It is, therefore, essential to map the areas susceptible to coseismic landslides, especially for the seismically active region. In the present work, the probabilistic assessment of coseismic landslides has been carried out for Goriganga valley located in the Kumaun Himalaya, India, which lies in the highest seismically active zone of the seismic zoning map of India. Several studies suggest that this region is prone to a great future earthquake of Mw ≥8.0. In this context, mapping of the coseismic landslide has been made for the future scenario earthquakes of 7.0, 8.0, and 8.6 Mw using modified Newmark's analysis. The modified Newmark's analysis provides the permanent displacement of the potential landslide, by integrating (1) joint strength of rock mass, (2) critical acceleration of the slope, and (3) peak ground acceleration of the region. Newmark permanent displacement has been estimated, which provides the distribution of predicted slope failure in the area. It has been observed that 41% of the area exhibits ˃40 cm Newmark's permanent displacement corresponding to Mw 8.6 earthquake and thus susceptible to failure, followed by 8.0 and 7.0 Mw earthquake with 36 and 14% of the area susceptible to the coseismic landslide, respectively. Further, the maximum permanent displacements for the simulated earthquakes of Mw 7.0, 8.0, and 8.6 are 76, 279, and 502 cm, respectively. [ABSTRACT FROM AUTHOR]

At 3:08 on 6 September 2018 (UTC +9), massive landslides were triggered by an earthquake of Mw 6.6 that occurred in Hokkaido, Japan. In this paper, a coseismic landslide inventory that covers 388 km2 of the earthquake-impacted area and includes 5828 coseismic landslides with a total landslide area of 23.66 km2 was compiled by using visual interpretations of various high-resolution satellite images. To analyze the spatial distribution and characteristics of coseismic landslides, five factors were considered: the peak ground acceleration (PGA), elevation, slope gradient, slope aspect, and lithology. Results show more than 87% of the landslides occurred at 100 to 200 m elevations. Slopes in the range of 10~20°are the most susceptible to failure. The landslide density of the places with peak ground acceleration (PGA) greater than 0.16 g is obviously larger than those with PGA less than 0.02 g. Compared with the number and scale of coseismic landslides caused by other strong earthquakes and the mobility of the coseismic landslides caused by the Haiyan and Wenchuan earthquakes, it was found that the distribution of coseismic landslides was extremely dense and that the mobility of the Hokkaido earthquake was greater than that of the Wenchuan earthquake and weaker than that of the Haiyuan earthquake, and is described by the following relationship: L = 18.454 ∗ H0.612. Comparative analysis of coseismic landslides with similar magnitude has important guiding significance for disaster prevention and reduction and reconstruction planning of landslides in affected areas.

Yogyakarta is one of the large cities in Central Java, located on Java Island, Indonesia. The city, and the Pleret sub-district, where the study has taken place, is prone to earthquake hazards, because it is close to several seismically active zones, such as the Sunda Megathrust and the active fault known as the Opak Fault. Since a devastating earthquake of 2006, the population of the Pleret sub-district has increased significantly. Thus, the housing demand has increased, and so is the pace of low-cost housing that does not meet earthquake-safety requirements, and furthermore are often located on unstable slopes. The local alluvial material covering a jigsaw of unstable blocks and complex slope is conditions that can amplify the negative impacts of earthquakes. Within this context, this study is aiming to assess the multi-hazards and risks of earthquakes and related secondary hazards such as ground liquefaction, and coseismic landslides. To achieve this, we used geographic information systems and remote sensing methods supplemented with outcrop study and existing seismic data to derive shear-strain parameters. The results have revealed the presence of numerous uncharted active faults with movements visible from imagery and outcrops. show that the middle part of the study area has a complex geological structure, indicated by many unchartered faults in the outcrops. Using this newly mapped blocks combined with shear strain data, we reassessed the collapse probability of buildings that reach level >0.75 near the Opak River, in central Pleret sub-district. Classifying the buildings and from population distribution, we could determine that the highest risk was during nighttime as the buildings susceptible to fall are predominantly housing buildings. The secondary hazards follow a slightly different distribution with a concentration of risks in the West. [ABSTRACT FROM AUTHOR]

Earthquake-triggered landslides frequently occur in active mountain areas, which poses great threats to the safety of human lives and public infrastructures. Fast and accurate mapping of coseismic landslides is important for earthquake disaster emergency rescue and landslide risk analysis. Machine learning methods provide automatic solutions for landslide detection, which are more efficient than manual landslide mapping. Deep learning technologies are attracting increasing interest in automatic landslide detection. CNN is one of the most widely used deep learning frameworks for landslide detection. However, in practice, the performance of the existing CNN-based landslide detection models is still far from practical application. Recently, Transformer has achieved better performance in many computer vision tasks, which provides a great opportunity for improving the accuracy of landslide detection. To fill this gap, we explore whether Transformer can outperform CNNs in the landslide detection task. Specifically, we build a new dataset for identifying coseismic landslides. The Transformer-based semantic segmentation model SegFormer is employed to identify coseismic landslides. SegFormer leverages Transformer to obtain a large receptive field, which is much larger than CNN. SegFormer introduces overlapped patch embedding to capture the interaction of adjacent image patches. SegFormer also introduces a simple MLP decoder and sequence reduction to improve its efficiency. The semantic segmentation results of SegFormer are further improved by leveraging image processing operations to distinguish different landslide instances and remove invalid holes. Extensive experiments have been conducted to compare Transformer-based model SegFormer with other popular CNN-based models, including HRNet, DeepLabV3, Attention-UNet, U2Net and FastSCNN. SegFormer improves the accuracy, mIoU, IoU and F1 score of landslide detectuin by 2.2%, 5% and 3%, respectively. SegFormer also reduces the pixel-wise classification error rate by 14%. Both quantitative evaluation and visualization results show that Transformer is capable of outperforming CNNs in landslide detection.

[Display omitted] • Coseismic landslide susceptibility is predicted via 3.43% of coseismic landslides. • CGBoost achieves high prediction accuracy and outperforms 8 state-of-the-art methods. • The performance of the proposed Crossgat is superior to the popular GNNs. The accurate prediction of landslide susceptibility shortly after a violent earthquake is quite vital to the emergency rescue in the 72-h "golden window". However, the limited quantity of interpreted landslides shortly after a massive earthquake makes landslide susceptibility prediction become a challenge. To address this gap, this work suggests an integrated method of Crossing Graph attention network and xgBoost (CGBoost). This method contains three branches, which extract the interrelations among pixels within a slope unit, the interrelations among various slope units, and the relevance between influencing factors and landslide probability, respectively, and obtain rich and discriminative features by an adaptive fusion mechanism. Thus, the difficulty of susceptibility modeling under a small number of coseismic landslides can be reduced. As a basic module of CGBoost, the proposed Crossing graph attention network (Crossgat) could characterize the spatial heterogeneity within and among slope units to reduce the false alarm in the susceptibility results. Moreover, the rainfall dynamic factors are utilized as prediction indices to improve the susceptibility performance, and the prediction index set is established by terrain, geology, human activity, environment, meteorology, and earthquake factors. CGBoost is applied to predict landslide susceptibility in the Gorkha meizoseismal area. 3.43% of coseismic landslides are randomly selected, of which 70% are used for training, and the others for testing. In the testing set, the values of Overall Accuracy, Precision, Recall, F1-score, and Kappa coefficient of CGBoost attain 0.9800, 0.9577, 0.9999, 0.9784, and 0.9598, respectively. Validated by all the coseismic landslides, CGBoost outperforms the current major landslide susceptibility assessment methods. The suggested CGBoost can be also applied to landslide susceptibility prediction in new earthquakes in the future. [ABSTRACT FROM AUTHOR]

On August 8, 2017, a magnitude Mw6.5 (Ms7.0) earthquake occurred in Jiuzhaigou County, Aba Prefecture, in the northern part of Sichuan Province, China, with a focal depth of 20 km and an epicenter located at (33.2°N, 103.8°E). Due to the significant magnitude of the earthquake, a large number of coseismic landslides were triggered. Despite previous research conducted by experts on the landslides caused by the Jiuzhaigou earthquake, the actual number of landslides has been severely underestimated in the previously published papers. Through field surveys and visual interpretation of high-resolution remote sensing images before and after the mainshock, we have established a detailed inventory of earthquake-induced landslides. The results indicate that the event caused a minimum of 9428 landslides covering a total area of 18.82 km 2 . These landslides are mainly distributed in the IX intensity area of the earthquake. The landslides mainly consist of medium-sized landslides and debris flows. They predominantly occur in areas with an altitude ranging from 2600 m to 3600 m, with slopes greater than 30° and facing east and southeast. The Lower Carboniferous and Middle Carboniferous formations are more prone to triggering landslides, and landslides are more concentrated within 1 km of roads and in forested areas. Additionally, as the distance from roads and the epicenter increases, the values of LAP and LND decrease, indicating a positive correlation between the two. There are more landslides within 2 km from the fault and within a range of 6 km-9 km from the epicenter. In conclusion, this study provides a comprehensive landslide inventory with broader coverage and increased accuracy. It also conducts a comprehensive analysis of the spatial distribution patterns of landslides. This contributes to a deeper understanding of the causes of coseismic landslides and further research on the impact of landslides in affected areas., Competing Interests: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (© 2024 The Authors.)

Earthquake-triggered landslides significantly contribute to worsening the impact of seismic events; thus, comprehensive landslide inventories are essential for improving seismic hazard assessment. During complex seismic sequences, landslides are triggered by more than one event and the final inventory reflects the spatial and temporal evolution of the sequence. Here, I analyze the landslides triggered by the 2018 Lombok (Indonesia) seismic sequence. I use high-resolution satellite imagery to map 4823 landslides triggered after the 05/08/2018 event (Mw 6.9) and 9319 landslides after the 19/08/2018 event (Mw 6.9). I analyze the distribution and evolution over time of landslide density and landslide area percentage. Despite the significant increase in number and cumulative area of the landslides, the 05/08 and 19/08 events share the maximum dimension of individual landslides; this suggests that the maximum intensity is equal for the two events, i.e., X on the Environmental Scale Intensity scale. I compare the distribution of landslides with macroseismic information provided by eyewitnesses through online questionnaires. Finally, I investigate the role of earthquake environmental effects within seismic sequences, showing that effects on the natural environment provide complementary information with respect to traditional intensity and felt reports. [ABSTRACT FROM AUTHOR]

An accurate and detailed seismic landslide inventory is essential to better understand the landslide mechanism and susceptibility. The 8th August 2017 Mw 6.5 Jiuzhaigou Earthquake of China initiated a large number of coseismic landslides. The results of the post-seismic survey show the actual landslide number might be underestimated in previous publications. Coupled with field investigation and visual interpretation on high-resolution remote sensing images before and after the main shock, we established a detailed inventory of landslides triggered by the earthquake. Results show that this event caused at least 4 834 individual landslides with a total area of 9.64 km2. They are concentrated in an elliptical area of 434 km2, dominated by medium- and small-scale rock falls and debris slides. Statistics indicate that, except for slope aspect that seems not significantly correlated with the landsliding, these landslides are most common in the places with following features: elevation of 2 800-3 400 m, slope angle greater than 30°, slope positions of upper, middle and flat slopes, and Carboniferous limestone and dolomite. Besides, the landslide area percentage (LAP) and landslide number density (LND) values decrease with the increasing distance to river channels and roads, implying a positive correlation. Instead of centering around the epicenter, most of these coseismic landslides are distributed along the inferred seismogenic fault, which means that the seismogenic structure played a more important role than the location of the epicenter. Remarkable differences in landslide densities along the fault indicate the varied landslide susceptibility which may be attributed to other varied controls along the fault such as the rock mass strength. In sum, this study presents a more detailed inventory of the landslides triggered by the 2017 Mw 6.5 Jiuzhaigou Earthquake, describes their distribution pattern and analyzes its control factors, which would be helpful to understand the genesis of the coseismic landslides and further study their long-term impact on the environment of the affected area. [ABSTRACT FROM AUTHOR]

The three‐dimensional discrete element method has been employed to analyze the dynamic fragmentation and lubrication mechanisms of coseismic Tangjiashan landslide induced by the 2008 Ms 8.0 Wenchuan earthquake. The numerical results show that the internal rock damage occurs and propagates gradually along the basal failure plane due to seismic shaking. At the peak seismic shaking, a sudden increase of tensile and shear stresses can lead to the complete breakage of basal bonds. This is associated with a sudden relief of overburden stress and rapid decrease of basal stress ratio. Thus, the slope fails as a whole and moves quickly downslope. In this process, several large transversal cracks develop at the middle and upper rear regions, disintegrating the slope mass into several large blocks. During landslide propagation, the thickness of basal fragmented layer increases progressively due to intense shearing, and the basal stress ratio reduces accordingly from 0.68 to 0.28. The reduction of landslide basal stress ratio occurs when the strong basal resistance is overcome by seismic‐ and gravity‐induced shear forces together with intense particle rearrangements. It can be quantified by vibrational and rotational granular temperatures of the basal shear layer, with the peak values of 35.2 and 11.6 m2/s2, respectively. The widespread internal slope fragmentation and subsequent lubrication have been identified as the key mechanisms governing landslide motion, which appear to be the intrinsic features of landslide irrespective of its triggering mechanism. The seismic shaking is more relevant to the detachment than to the spreading of landslide mass. Key Points: The dynamic fragmentation and lubrication mechanisms of coseismic landslides have been analyzed by discrete element methodThe landslide is triggered by the seismic shaking‐induced rock fragmentation and reduction of basal stress ratioThe widespread slope fragmentation and lubrication are the intrinsic features of landslide irrespective of its triggering mechanism [ABSTRACT FROM AUTHOR]

Abstract: Landslides in granular soils can be highly hazardous when exhibiting flow‐like behavior. The extensive mass wasting associated with the 2008 Mw = 7.9 Wenchuan earthquake (China) left several cubic kilometers of loose granular material deposited along steep slopes and in low‐order channels. Rainfall‐triggered remobilization of these deposits evolved often into catastrophic flow‐like landslides. Ten years after the earthquake, most of the deposits are still in place but landslide rates have decreased significantly. Internal erosion‐induced grain coarsening is one possible process producing this decrease. Through experiments on loose artificial slopes we demonstrate the major role of the internally erodible small grains in triggering failure and fluidization and producing grain coarsening. Under the same hydraulic boundary, if the erodible fraction is removed or reduced, the loose deposits remain stable or fail without fluidizing. Our results provide an experimental evidence to the patterns of sediment export and debris flows observed in nature after a strong earthquake. [ABSTRACT FROM AUTHOR]

Geometric parameters are useful for characterizing earthquake-triggered landslides. This paper presents a detailed statistical analysis on this issue using the landslide inventory of the 2013, Minxian, China Mw 5.9 earthquake. Based on GIS software and a 5-m resolution DEM, geometric parameters of 635 coseismic landslides (with areas larger than 500 m) were obtained, including height, length, width, reach angle (arc tangent of the height-length ratio), and aspect ratio (length-width ratio). The fitting relationship of height and length from these data is H = 0.6164 L + 0.4589, with an average reach angle of 31.65°. The landslide aspect ratios concentrate in the range of 1.4∼2.6, with an average of 2.11. According to the plane geometric shapes and aspect ratios, the landslides are classified into four categories: transverse landslide (LA1, L/ W ≤ 0.8), isometric landslide (LA2, 0.8 < L/ W ≤ 1.2), longitudinal landslide (LA3, 1.2 < L/ W ≤ 3), and elongated landslide (LA4, L/ W > 3). Statistics of these four types of landslides versus ten classified control factors (elevation, slope angle, slope aspect, curvature, slope position, distance to drainages, lithology, seismic intensity, peak ground acceleration, and distance to seismogenic fault) are used to examine their possible correlations and the landslide-prone areas, which would be helpful to the landslide disaster mitigation in the affected area. [ABSTRACT FROM AUTHOR]

The 25 April 2015 M w 7.8 Gorkha earthquake produced strong ground motions across an approximately 250 km by 100 km swath in central Nepal. To assist disaster response activities, we modified an existing earthquake-triggered landslide model based on a Newmark sliding block analysis to estimate the extent and intensity of landsliding and landslide dam hazard. Landslide hazard maps were produced using Shuttle Radar Topography Mission (SRTM) digital topography, peak ground acceleration (PGA) information from the U.S. Geological Survey (USGS) ShakeMap program, and assumptions about the regional rock strength based on end-member values from previous studies. The instrumental record of seismicity in Nepal is poor, so PGA estimates were based on empirical Ground Motion Prediction Equations (GMPEs) constrained by teleseismic data and felt reports. We demonstrate a non-linear dependence of modeled landsliding on aggregate rock strength, where the number of landslides decreases exponentially with increasing rock strength. Model estimates are less sensitive to PGA at steep slopes (> 60°) compared to moderate slopes (30–60°). We compare forward model results to an inventory of landslides triggered by the Gorkha earthquake. We show that moderate rock strength inputs over estimate landsliding in regions beyond the main slip patch, which may in part be related to poorly constrained PGA estimates for this event at far distances from the source area. Directly above the main slip patch, however, the moderate strength model accurately estimates the total number of landslides within the resolution of the model (landslides ≥ 0.0162 km 2 ; observed n = 2214, modeled n = 2987), but the pattern of landsliding differs from observations. This discrepancy is likely due to the unaccounted for effects of variable material strength and local topographic amplification of strong ground motion, as well as other simplifying assumptions about source characteristics and their relationship to landsliding. [ABSTRACT FROM AUTHOR]

The 2016 Mw 7.0 Kumamoto earthquake triggered numerous landslides of fallout tephra deposits around Aso Volcano in southwestern Japan. Although rainfall-induced landslides have been repeatedly observed on the tephra-mantled slopes of this region, the spatial patterns and frequency of coseismic landslides remain unclear because of the scarce landslide evidence associated with palaeoseismicity. To examine the distribution of landslides induced by the 2016 Kumamoto earthquake and possibly by prior large earthquakes, we conducted geomorphological analyses using airborne light detection and ranging (LiDAR) data, together with geological investigations of tephra deposits in the Takanoobane lava dome (TLD), a post-caldera cone of Aso Volcano. Repeated LiDAR surveys performed before and after the earthquake allowed us to quantify the volume and geometry of five coseismic landslides (K1–K5), as well as three component blocks of the K1 landslide (K1a–c). The surveys also allowed identification of 12 older landslides (O1–O12) fully covered by vegetation. Source areas of recent coseismic landslides ranged from 1050 to 14,920 m2; those of older landslides were estimated to be 1110–15,730 m2. The H / L ratios of the K1 landslide and its component blocks K1a–c ranged from 0.139 to 0.165, despite their volumes ranging from 104 to 105 m3. The K1 landslide occurred on the southwest-facing slope, where older landslides were not recognised before the earthquake. Its slip surface was formed at the base of the 30 ka pumice layer. The largest one of the older landslides (O1) is located on the south-facing slope, adjacent to the K1 landslide. Stratigraphic investigations revealed that most of the tephra sequence, including the 30 ka pumice layer, has already been eroded at the centre of the O1 landslide. Radiocarbon dating further constrained the landslide age to ca. 7.0 ka. Because its topographic and stratigraphic features are similar to the features of the K1 landslide, the O1 landslide may have been triggered by a palaeoearthquake approximately 7000 years ago, which was previously unknown around the Aso caldera region. • Repeated LiDAR surveys allowed quantification of recent coseismic landslides and identification of older landslides. • Five coseismic landslides occurred without overlapping 12 older landslides. • Coseismic landslides showed high mobility, despite their small volume (≤ 105 m3). • Tephra deposit unconformity indicated that an older landslide slipped at the base of the 30 ka pumice layer. • A palaeoearthquake at 7.0 ka may have induced this older landslide. [ABSTRACT FROM AUTHOR]