Showing papers in "Landslides in 2020"

Improved landslide assessment using support vector machine with bagging, boosting, and stacking ensemble machine learning framework in a mountainous watershed, Japan

Abstract: Heavy rainfall in mountainous terrain can trigger numerous landslides in hill slopes. These landslides can be deadly to the community living downslope with their fast pace, turning failures into catastrophic debris flows and avalanches. Active tectonics coupled with rugged topography in a complex geoenvironment multiplies this likelihood. The available hazard maps are usually helpful in mitigating disasters. However, fool-proof predicting landslide susceptibility identification remains a challenge in landslide discipline. Recently, ensemble machine learning (ML) techniques have proved the potential to provide a more accurate and efficient solution in spatial modeling. The main purposes of the current study are to examine and evaluate the predictive capability of support vector machine hybrid ensemble ML algorithms, i.e., the bagging, boosting, and stacking for modeling the catastrophic rainfall-induced landslide occurrences in the Northern parts of Kyushu Island, at the watershed scale in Japan. In this study, a landslide inventory map containing 265 landslide polygons was first interpreted from the aerial photographs and fieldwork after the September 2017 rainfall event. The raw data were randomly separated into two parts using a 70/30 sampling strategy for training and validating the landslide models. Then, 13 predisposing factors were prepared as predictors and dependent variable. The landslide susceptibility maps (LSM) were validated by the area under the receiver operating characteristic curve (AUC). The results of validation showed that the AUC values of the four models (SVM-Stacking, SVM, SVM-Bagging, and SVM-Boosting) varied from 0.74 to 0.91. The SVM-boosting model outperformed the other models, while SVM-stacking model has found to be the lowest performance. The outcome suggests that an ensemble ML model does not necessarily mean good performance. It is always preferable to select an appropriate model, such as the one proposed the hybrid novel ensemble SVM-boosting model, which could significantly improve the accuracies of LSM. Also, from Information Gain Ratio (IGR) we found that the rainfall factor mainly affects the results, that agrees with the analogy of present study.

A deep learning algorithm using a fully connected sparse autoencoder neural network for landslide susceptibility prediction

Abstract: The environmental factors of landslide susceptibility are generally uncorrelated or non-linearly correlated, resulting in the limited prediction performances of conventional machine learning methods for landslide susceptibility prediction (LSP). Deep learning methods can exploit low-level features and high-level representations of information from environmental factors. In this paper, a novel deep learning–based algorithm, the fully connected spare autoencoder (FC-SAE), is proposed for LSP. The FC-SAE consists of four steps: raw feature dropout in input layers, a sparse feature encoder in hidden layers, sparse feature extraction in output layers, and classification and prediction. The Sinan County of Guizhou Province in China, with a total of 23,195 landslide grid cells (306 recorded landslides) and 23,195 randomly selected non-landslide grid cells, was used as study case. The frequency ratio values of 27 environmental factors were taken as the input variables of FC-SAE. All 46,390 landslide and non-landslide grid cells were randomly divided into a training dataset (70%) and a test dataset (30%). By analyzing real landslide/non-landslide data, the performances of the FC-SAE and two other conventional machine learning methods, support vector machine (SVM) and back-propagation neural network (BPNN), were compared. The results show that the prediction rate and total accuracies of the FC-SAE are 0.854 and 85.2% which are higher than those of the SVM-only (0.827 and 81.56%) and BPNN (0.819 and 80.86%), respectively. In conclusion, the asymmetric and unsupervised FC-SAE can extract optimal non-linear features from environmental factors successfully, outperforms some conventional machine learning methods, and is promising for LSP.

Landslide susceptibility prediction based on a semi-supervised multiple-layer perceptron model

Abstract: Conventional supervised and unsupervised machine learning models used for landslide susceptibility prediction (LSP) have many drawbacks, such as an insufficient number of recorded landslide samples, and the subjective and random selection of non-landslide samples. To overcome these drawbacks, a semi-supervised multiple-layer perceptron (SSMLP) is innovatively proposed with several processes: (1) an initial landslide susceptibility map (LSM) is produced using the multiple-layer perceptron (MLP) based on the original recorded landslide samples and related environmental factors; (2) the initial LSM is respectively classified into five areas with very high, high, moderate, low and very low susceptible levels; (3) some reasonable grid units from the areas with very high susceptible level are selected as new landslide samples to expand the original landslide samples; (4) reasonable non-landslide samples are selected from the areas with very low susceptible level; and (5) the expanded landslide samples, reasonable selected non-landslide samples and related environmental factors are put into the MLP once again to predict the final LSM. The Xunwu County of Jiangxi Province in China is selected as the study area. Conventional supervised machine learning (i.e. MLP) and unsupervised machine learning (i.e. K-means clustering model) are selected for comparisons. The comparative results indicate that the SSMLP model has a considerably higher LSP performance than the MLP and K-means clustering in Xunwu County. The SSMLP model successfully addresses the drawbacks existed in the conventional machine learning for LSP.

Landslide detection from an open satellite imagery and digital elevation model dataset using attention boosted convolutional neural networks

Abstract: Convolution neural network (CNN) is an effective and popular deep learning method which automatically learns complicated non-linear mapping from original inputs to given labels or ground truth through a series of convolutional layers. This study focuses on detecting landslides from high-resolution optical satellite images using CNN-based methods, providing opportunities for recognizing latent landslides and updating large-scale landslide inventory with high accuracy and time efficiency. Considering the variety of landslides and complicated backgrounds, attention mechanisms originated from the human visual system are developed for boosting the CNN to extract more distinctive feature representations of landslides from backgrounds. As deep learning needs a large number of labeled data to train a learning model, we manually prepared a landslide dataset which is located in the Bijie city, China. In the dataset, 770 landslides, including rock falls, rock slides, and a few debris slides, were interpreted by geologists from the satellite images and digital elevation model (DEM) data and further checked by fieldwork. The landslide data was separated into a training set that trains the attention boosted CNN model and a testing set that evaluates the performance of the model with a ratio of 2:1. The experimental results showed that the best F1-score of landslide detection reached 96.62%. The results also proved that the performance of our spatial-channel attention mechanism was fairly over other recent attention mechanisms. Additionally, the effectiveness of predicting new potential landslides with high efficiency based on our dataset is demonstrated.

Machine learning ensemble modelling as a tool to improve landslide susceptibility mapping reliability

Abstract: Statistical landslide susceptibility mapping is a topic in complete and constant evolution, especially since the introduction of machine learning (ML) methods. A new methodological approach is here presented, based on the ensemble of artificial neural network, generalized boosting model and maximum entropy ML algorithms. Such approach has been tested in the Monterosso al Mare area, Cinque Terre National Park (Northern Italy), severely hit by landslides in October 2011, following an extraordinary precipitation event, which caused extensive damage at this World Heritage site. Thirteen predisposing factors were selected and assessed according to the main characteristics of the territory and through variance inflation factor, whilst a database made of 260 landslides was adopted. Four different Ensemble techniques were applied, after the averaging of 300 stand-alone methods, each one providing validation scores such as ROC (receiver operating characteristics)/AUC (area under curve) and true skill statistics (TSS). A further model performance evaluation was achieved by assessing the uncertainty through the computation of the coefficient of variation (CV). Ensemble modelling thus showed improved reliability, testified by the higher scores, by the low values of CV and finally by a general consistency between the four Ensemble models adopted. Therefore, the improved reliability of Ensemble modelling confirms the efficacy and suitability of the proposed approach for decision-makers in land management at local and regional scales.

Prediction of a multi-hazard chain by an integrated numerical simulation approach : the Baige landslide, Jinsha River, China

Abstract: Successive major landslides during October and November 2018 in Baige village, eastern Tibet, dammed the Jinsha River on two occasions, and the subsequent dam breaches instigated a multi-hazard chain that flooded many towns downstream. Analysis of high-resolution aerial images and field investigations unveiled three potentially unstable rock mass clusters in the source area of the landslides, suggesting possible future failures with potential for river-damming and flooding. In order to evaluate and understand the disaster chain effect linked to the potentially unstable rock mass, we systematically studied the multi-hazard scenarios through an integrated numerical modelling approach. Our model begins with an evaluation of the probability of landslide failure, including runout and river damming, and then addresses the dam breach and resultant flood—hence simulating and visualising an entire disaster chain. The model parameters were calibrated using empirical data from the two Baige landslides. Then, we predict the future cascading hazards via seven scenarios according to all possible combinations of potential rock mass failure. For each scenario, the landslide runouts, dam-breaching, and flooding are numerically simulated with full consideration of uncertainties among the model input parameters. The maximum dam breach flood extent, depth, velocity, and peak arrival time are predicted at sequential sites downstream. As a first attempt to simulate the full spectrum of a landslide-induced multi-hazard chain, our study provides insights and substantiates the value provided by multi-hazard modelling. The integrated approach described here can be applied to similar landslide-induced chains of hazards in other regions.

Assessing the potential of soil moisture measurements for regional landslide early warning

Abstract: In mountainous terrain, rainfall-induced landslides pose a serious risk to people and infrastructure. Regional landslide early warning systems (LEWS) have proven to be a cost-efficient tool to inform the public about the imminent landslide danger. While most operational LEWS are based on rainfall exceedance thresholds only, recent studies have demonstrated an improvement of the forecast quality after the inclusion of soil hydrological information. In this study, the potential of in situ soil moisture measurements for regional landslide early warning is assessed. For the first time, a comprehensive soil moisture measurement database was compiled for Switzerland and compared with a national landslide database (Swiss flood and landslide damage database, WSL). The time series were homogenized and normalized to represent saturation values. From ensembles of sensors, the mean and standard deviation saturation were calculated and infiltration events were delimited, characterized, and classified as landslide-triggering or non-triggering based on the occurrence of landslides within a specified forecast distance. A logistic regression function was applied to model the landslide activity based on the infiltration event characteristics and several models were analysed and compared with receiver operating characteristics (ROC). A strong distance dependence becomes apparent showing a forecast goodness decrease with increasing distance between water content measurement site and landslide, and a better forecast goodness for long-lasting as opposed to short-duration precipitation events. While most variability can be explained by the two event properties antecedent saturation and change of saturation during an infiltration event, event properties that describe antecedent conditions are more important for long-lasting as opposed to short-duration precipitation events that can be better explained by properties describing event dynamics. Overall, the analysis demonstrated that in situ soil moisture data effectively contains specific information useful for landslide early warning.

Landslide susceptibility assessment in complex geological settings: sensitivity to geological information and insights on its parameterization

Abstract: The literature about landslide susceptibility mapping is rich of works focusing on improving or comparing the algorithms used for the modeling, but to our knowledge, a sensitivity analysis on the use of geological information has never been performed, and a standard method to input geological maps into susceptibility assessments has never been established. This point is crucial, especially when working on wide and complex areas, in which a detailed geological map needs to be reclassified according to more general criteria. In a study area in Italy, we tested different configurations of a random forest–based landslide susceptibility model, accounting for geological information with the use of lithologic, chronologic, structural, paleogeographic, and genetic units. Different susceptibility maps were obtained, and a validation procedure based on AUC (area under receiver-operator characteristic curve) and OOBE (out of bag error) allowed us to get to some conclusions that could be of help for in future landslide susceptibility assessments. Different parameters can be derived from a detailed geological map by aggregating the mapped elements into broader units, and the results of the susceptibility assessment are very sensitive to these geology-derived parameters; thus, it is of paramount importance to understand properly the nature and the meaning of the information provided by geology-related maps before using them in susceptibility assessment. Regarding the model configurations making use of only one parameter, the best results were obtained using the genetic approach, while lithology, which is commonly used in the current literature, was ranked only second. However, in our case study, the best prediction was obtained when all the geological parameters were used together. Geological maps provide a very complex and multifaceted information; in wide and complex area, this information cannot be represented by a single parameter: more geology-based parameters can perform better than one, because each of them can account for specific features connected to landslide predisposition.

Landslides after wildfire: initiation, magnitude, and mobility

Abstract: In the semiarid Southwestern USA, wildfires are commonly followed by runoff-generated debris flows because wildfires remove vegetation and ground cover, which reduces soil infiltration capacity and increases soil erodibility. At a study site in Southern California, we initially observed runoff-generated debris flows in the first year following fire. However, at the same site three years after the fire, the mass-wasting response to a long-duration rainstorm with high rainfall intensity peaks was shallow landsliding rather than runoff-generated debris flows. Moreover, the same storm caused landslides on unburned hillslopes as well as on slopes burned 5 years prior to the storm and areas burned by successive wildfires, 10 years and 3 years before the rainstorm. The landslide density was the highest on the hillslopes that had burned 3 years beforehand, and the hillslopes burned 5 years prior to the storm had low landslide densities, similar to unburned areas. We also found that reburning (i.e., two wildfires within the past 10 years) had little influence on landslide density. Our results indicate that landscape susceptibility to shallow landslides might return to that of unburned conditions after as little as 5 years of vegetation recovery. Moreover, most of the landslide activity was on steep, equatorial-facing slopes that receive higher solar radiation and had slower rates of vegetation regrowth, which further implicates vegetation as a controlling factor on post-fire landslide susceptibility. Finally, the total volume of sediment mobilized by the year 3 landslides was much smaller than the year 1 runoff-generated debris flows, and the landslides were orders of magnitude less mobile than the runoff-generated debris flows.

A step beyond landslide susceptibility maps: a simple method to investigate and explain the different outcomes obtained by different approaches

Abstract: Landslide susceptibility assessment is vital for landslide risk management and urban planning, and the scientific community is continuously proposing new approaches to map landslide susceptibility, especially by hybridizing state-of-the-art models and by proposing new ones. A common practice in landslide susceptibility studies is to compare (two or more) different models in terms of AUC (area under ROC curve) to assess which one has the best predictive performance. The objective of this paper is to show that the classical scheme of comparison between susceptibility models can be expanded and enriched with substantial geomorphological insights by focusing the comparison on the mapped susceptibility values and investigating the geomorphological reasons of the differences encountered. To this aim, we used four susceptibility maps of the Wanzhou County (China) obtained with four different classification methods (namely, random forest, index of entropy, frequency ratio, and certainty factor). A quantitative comparison of the susceptibility values was carried out on a pixel-by-pixel basis, to reveal systematic spatial patterns in the differences among susceptibility maps; then, those patterns were put in relation with all the explanatory variables used in the susceptibility assessments. The lithological and morphological features of the study area that are typically associated to underestimations and overestimations of susceptibility were identified. The results shed a new light on the susceptibility models, identifying systematic errors that could be probably associated either to shortcomings of the models or to distinctive morphological features of the test site, such as nearly flat low altitude areas near the main rivers, and some lithological units.

Longevity analysis of landslide dams

Abstract: Landslide dams are extremely dangerous because dammed rivers can inundate upstream areas with rising water levels and flood downstream areas after dam breaching. The longevity of landslide dams, which is uncertain, is of great significance for dam failure prevention and mitigation since it determines the time available to take mitigation measures. In this study, the full longevity of landslide dams is divided into three stages (infilling, overflowing and breaching) for better estimation. The influences of dam characteristic parameters (triggers, dam materials and geometric/hydrological parameters) on the full longevity of landslide dams (the period from landslide dam formation to the end of dam failure) as well as on each of the three stages are analysed based on the database. Based on eight dimensionless variables, regression models for estimating the full longevity of landslide dams are developed with a R2 value of 0.781, and regression models for the three-stage longevity (the longevity as the sum of the periods of the three stages) by considering infilling, overflowing and breaching are established with a R2 value of 0.938. It is found that the landslide dam longevity cannot be predicted by one or two influencing factors since it is affected by multiple factors. The relative importance of each control variable is evaluated based on sensitivity analysis: the trigger is the most significant variable in the breaching stage since it affects the size of dam particles, the water content and the inflow rate (e.g. the rainfall trigger results in a larger inflow rate); the lake volume coefficient is more significant in the overflowing stage because it indicates the potential volume of water eroding the dam; and the average annual discharge coefficient is the most important factor in the infilling stage because it controls the time to impound water. The longevity predicted by different models are compared. The models developed in this paper show better accuracy due to the consideration of more parameters based on more cases. In particular, the three-stage longevity regression model shows better accuracy than that of other models because it considers the particular influencing factors for each stage. Three case studies (the “10·10” Baige, Hsiaolin and Tangjiashan landslide dams) are presented to show the application of the regression models developed in this paper. The dam longevity can be predicted more precisely if the timely inflow rate can be estimated by site monitoring or multi-temporal remote sensing images and pre-event digital elevation model (DEM).

Soil moisture information can improve shallow landslide forecasting using the hydrometeorological threshold approach

Abstract: Empirical thresholds indicating the meteorological conditions leading to shallow landslide triggering are one of the most important components of landslide early warning systems (LEWS). Thresholds have been determined for many parts of the globe and present significant margins of improvement, especially for the high number of false alarms they produce. The use of soil moisture information to define hydro-meteorological thresholds is a potential way of improvement. Such information is becoming increasingly available from remote sensing and sensor networks, but to date, there is a lack of studies that quantify the possible improvement of the performance of LEWS. In this study, we investigate this issue by modelling the response of slopes to precipitations, introducing also the possible influence of uncertainty in soil moisture provided by either field sensors or remote sensing, and investigating various soil depths at which the information may be available. Results show that soil moisture information introduced within hydro-meteorological thresholds can significantly reduce the false alarm ratio of LEWS, while keeping at least unvaried the number of missed alarms. The degree of improvement is particularly significant in the case of soils with small water storage capacity.

Numerical simulation of landslide-generated waves during the 11 October 2018 Baige landslide at the Jinsha River

Abstract: Landslides at river embankments can block watercourses, imperiling the safety of vessels and downstream hydropower stations. The Baige landslide, which occurred on 11th October 2018, is taken as an example to study the landslide motion and landslide-generated wave evolutions. The elasto-viscoplastic and renormalization group (RNG) turbulence models are employed in the FLOW3D software, treating the motion of the Baige landslide as a viscous flow. Numerical results show that the maximum velocity of the slide was approximately 75 m/s when entering the Jinsha River. Further, the waves triggered by massive debris avalanches at three different locations are investigated. The maximum velocity of the landslide-generated wave and the maximum run-up in the Jinsha River reached 45 m/s and 53.9 m, respectively, on the slide axis. The maximum run-up terrain elevation of the wave was 3039.7 m. The simulation results are basically consistent with the actual field observations and fit well with high-speed flow-like landslides. In this case, the displaced water was dominant due to the significant volume of the failure mass and the shallow watercourse of the Jinsha River. The run-down waves located on the source region axis contribute to the rise of water level downstream and upstream. The results from this case study serve as a practical inspiration for research on disaster processes.

Long-term InSAR, borehole inclinometer, and rainfall records provide insight into the mechanism and activity patterns of an extremely slow urbanized landslide

Abstract: New radar satellites provide global coverage and the possibility of long-term, regular frequency (days-weeks) surface displacement measurements through the application of high precision multi-temporal InSAR (Synthetic Aperture Radar Interferometry) techniques. This represents an excellent opportunity to investigate and improve our understanding of the behavior of extremely slow landslides, as well as of the long- to short-term controls of their activity. In urban settings, such landslides deserve special attention, as their cumulative movements can cause significant socio-economic damage. Here, we re-examine the case of a long-lived, deep-seated landslide in the Apennine Mountains (Italy) which was urbanized between the late 1970s and early 2000s. The case provides a rare opportunity to highlight the benefits of the integrated analysis of long-term (several years) borehole inclinometer measurements with 15 years of multi-temporal InSAR displacement data. We present evidence of the landslide composite nature and asymmetry, and draw attention to the recent period of accelerated movement that coincided with the foot failure event. This helps constraining the interpretation of the borehole and InSAR data and demonstrating the predominantly rotational landslide mechanism. We show how a detailed analysis of sparse inclinometer and more spatially continuous InSAR measurements, when combined with local rainfall records, can reveal long- to short-term patterns of temporal variability in landslide motions and allow anticipating the consequences of future landslide activity.

Modeling and predicting reservoir landslide displacement with deep belief network and EWMA control charts: a case study in Three Gorges Reservoir

Abstract: The accurate modeling and predicting of landslide deformation is crucial to the prevention of landslide hazard. This paper presents a pioneering study of modeling and predicting the reservoir landslide displacement with deep learning algorithm. A data-driven framework using deep belief network and control chart has been introduced to explore the temporal patterns of displacement and potential of identifying seasonal faster displacement. First, the continuous wavelet analysis has been applied to decompose the time-series precipitation, reservoir water level, and displacement into seasonal and residual components. Second, the deep belief network has been constructed to predict the future displacement. Third, it utilizes the exponentially weighted moving average (EWMA) control chart to derive the boundaries as alarm conditions of seasonal faster displacement. A group of tests are conducted to compare the performance of the deep belief network with other state-of-the-art machine learning algorithms. Computational results demonstrated the effectiveness of the deep belief network in extracting highly non-linear data features. In addition, the advantage of utilizing control charts has been further validated by the accuracy of examining the seasonal faster displacement based on the case study in Baishuihe landslide in Three Gorges Reservoir, China.

Dynamic modeling of large deformation slope failure using smoothed particle finite element method

Abstract: In this paper, a novel node-based explicit smoothed particle finite element method (SPFEM), on the basis of the particle finite element method (PFEM) framework, is utilized to evaluate the stability of slopes and to simulate the post-failure behavior of soil. The main advantage of SPFEM in slope stability analysis lies in its capabilities to consider the whole dynamic failure process of slope and to simulate large deformation and post-failure of soils. For the stability analysis of a cohesive soil slope, the shear strength reduction technique with a kinetic energy-based criterion for distinguishing slope failure is adopted to obtain the factor of safety (FOS) of a slope, and the FOS is compared with that obtained by the classical FEM and LEM approaches for further validation. Then, the dynamic failure process of a non-cohesive granular material slope is simulated using Drucker-Prager constitutive model. The influence of friction resistance of granular material, as well as the repose angle of slope after failure, is discussed. Finally, the progressive failure behavior of a long clayey slope is modeled using SPFEM in conjunction with a strain-softening Tresca constitutive model. The retrogressive failure behavior of a long clayey slope is analyzed.

Coastal landslides in Palu Bay during 2018 Sulawesi earthquake and tsunami

Abstract: In this paper, we investigate the sources for Palu Bay tsunamis, which occurred in September 2018 after a Mw 7.5 earthquake struck on a strike-slip fault. Previous land-based field studies have identified numerous coastal landslides as possible sources for tsunami generation. Here, we present new post-earthquake bathymetry survey data in the bay. We compare the new very detailed nearshore bathymetry with the pre-earthquake data so as to quantify the magnitudes and configurations of coastal landslides and the characteristics of landslide-generated waves. We perform numerical simulations of tsunami propagation, isolating the coastal landslides as the only sources. The results show that, although the landslide-generated waves characteristically have shorter periods than the observed tsunami waves, the combination of the waves is able to produce long waves comparable with those observed, due to the ringing effects of the trapped waves inside the bay. Therefore, we conclude that landslide-generated waves played a significant role in the tsunamis in Palu Bay. Nevertheless, it is likely that the tsunamis were caused by a combination of tectonic and landslide sources.

On the landslide tsunami uncertainty and hazard

Abstract: Landslides are the second most frequent tsunami source worldwide. However, their complex and diverse nature of origin combined with their infrequent event records make prognostic modelling challenging. In this paper, we present a probabilistic framework for analysing uncertainties emerging from the landslide source process. This probabilistic framework employs event trees and is used to conduct tsunami uncertainty analysis as well as probabilistic tsunami hazard analysis (PTHA). An example study is presented for the Lyngen fjord in Norway. This application uses a mix of empirical landslide data combined with expert judgement to come up with probability maps for tsunami inundation. Based on this study, it is concluded that the present landslide tsunami hazard analysis is largely driven by epistemic uncertainties. These epistemic uncertainties can be incorporated in the probabilistic framework. Conducting a literature analysis, we further show examples of how landslide and tsunami data can be used to better constrain landslide uncertainties, combined with statistical and numerical analysis methods. We discuss how these methods, combined with the probabilistic framework, can be used to improve landslide tsunami hazard analysis in the future.

Reactivation of giant Jiangdingya ancient landslide in Zhouqu County, Gansu Province, China

Abstract: On 12 July 2018, the giant Jiangdingya ancient landslide reactivated and blocked the Bailongjiang River in Nanyu Town, Gansu Province, China. The ancient landslide was ca. 4.1~4.9 × 107 m3, and the 2018 reactivated landslide was ca. 4.8~5.5 × 106 m3, with the characteristics of a deep-seated retrogressive landslide. The terrain and geological structure of the landslide area are very complicated with alpine valleys and high gradient rivers, and the ancient landslide had maturely developed. The Pingding-Huama fault, which has been active since the late Quaternary, passed through the landslide region. Because of the active fault and tectonic uplift, the lithology is deeply fragmented. Based on the remote sensing interpretation and field investigation, development characteristics and reactivation mechanism of the Jiangdingya ancient landslide are described.

Numerical study of granular debris flow run-up against slit dams by discrete element method

Abstract: Run-up of granular debris flows against slit dams on slopes is a complex process that involves deceleration, deposition, and discharge. It is imperative to understand the run-up mechanism and to predict the maximum run-up height for the engineering design and hazard mitigation. However, the interaction between granular flows and slit dams, which significantly affects the run-up height, is still not well understood. In this study, an analytical model based on the momentum approach was derived to predict the run-up heights of granular debris flows. A numerical investigation of granular debris flow impacting slit dams using the discrete element method (DEM) was then conducted. The influence of the Froude number (NFr) and the relative post spacing (R) on run-up height were studied. This study illustrates that the analytical model based on the momentum approach can predict the run-up heights well within a certain range of Froude numbers. There is a critical value of relative post spacing (RC): within the critical value, the maximum run-up height is insensitive to the relative post spacing; once R exceeds the critical value, the maximum run-up height decreases rapidly as the relative post spacing increases.

Seismic signal recognition and interpretation of the 2019 “7.23” Shuicheng landslide by seismogram stations

Abstract: A systematic study of the physical and mechanical processes of landslide development and evolution is important for forecasting, early warning, and prevention of landslide hazards. In the absence of on-site monitoring data, seismic networks can be employed to continuously record ground seismicity generated during landslides. However, landslide seismic signals are relatively weak and inevitably affected by noise interference. Furthermore, systematic characterization and reconstruction of the landslide evolution process remain poorly reported. An evaluation method to recognize landslide events based on seismic signal characteristics is therefore important. This study analyzes the 2019 “7.23” Shuicheng landslide based on data from nearby seismic stations. A landslide seismic signal recognition method is developed based on short-time Fourier transform (STFT) and band-pass filter (BP-filter) analysis. Data from 14 stations near the landslide were reviewed and the landslide data from one station was selected for analysis. The landslide seismic signal was noise-attenuated by using the empirical mode decomposition (EMD) and BP-filter methods. Fast Fourier transform (FFT), STFT, and power spectral density analyses were applied to the landslide seismic signal with higher signal-to-noise ratio (SNR) to obtain the time–frequency signal characteristics of the landslide process. Finally, combined with landslide field survey data, the dynamic process of the landslide was reconstructed based on the seismic signal, and the landslide was divided into four stages: the fracture-transition stage, the accelerated initiation stage, the bifurcation-scraping stage, and the deposition stage. The dynamic characteristics of each stage of the landslide are presented. The results indicate that the initial fracture point of the landslide is located between the bottom of the sliding source area and the top of the acceleration zone, not as traditionally thought, at the top of the sliding source area; this would be difficult to determine through field survey and analysis only. These results provide theoretical guidance for the study of seismic signal extraction, identification of landslide dynamic parameters, and characterization and reconstruction of landslide processes.

Landslides across the USA: occurrence, susceptibility, and data limitations

Abstract: Detailed information about landslide occurrence is the foundation for advancing process understanding, susceptibility mapping, and risk reduction. Despite the recent revolution in digital elevation data and remote sensing technologies, landslide mapping remains resource intensive. Consequently, a modern, comprehensive map of landslide occurrence across the United States (USA) has not been compiled. As a first step toward this goal, we present a national-scale compilation of existing, publicly available landslide inventories. This geodatabase can be downloaded in its entirety or viewed through an online, searchable map, with parsimonious attributes and direct links to the contributing sources with additional details. The mapped spatial pattern and concentration of landslides are consistent with prior characterization of susceptibility within the conterminous USA, with some notable exceptions on the West Coast. Although the database is evolving and known to be incomplete in many regions, it confirms that landslides do occur across the country, thus highlighting the importance of our national-scale assessment. The map illustrates regions where high-quality mapping has occurred and, in contrast, where additional resources could improve confidence in landslide characterization. For example, borders between states and other jurisdictions are quite apparent, indicating the variation in approaches to data collection by different agencies and disparity between the resources dedicated to landslide characterization. Further investigations are needed to better assess susceptibility and to determine whether regions with high relief and steep topography, but without mapped landslides, require further landslide inventory mapping. Overall, this map provides a new resource for accessing information about known landslides across the USA.

The effect of topography on landslide kinematics: a case study of the Jichang town landslide in Guizhou, China

Abstract: On July 23, 2019, a large-scale landslide occurred in Jichang town, Shuicheng County, Liupanshui City, Guizhou Province in China. The landslide, which moved along two gullies, resulted in strong punching-shear, induced scarping on vegetation and large destruction of houses, and finally formed a deposit with a volume of 2 × 106 m3. This research aims to understand the effect of topography on landslide kinematics. To achieve this aim, a detailed field investigation was first carried out with an unmanned aerial vehicle (UAV) aerial photography survey, resident interviews, and field sampling. The rainfall analysis indicates the effective rainfall within 7 days before landslides was 70.14 mm which exceeded the rainfall threshold of 54.3 mm in this region, which finally triggered the landslide. Traditional soil mechanic tests were then performed to identify the soil properties of the source material. Combined with numerical simulation using the nonlinear shallow water equation, the whole process of landslides was divided into four stages: instability stage, acceleration stage, transformation stage, and impact and accumulation stage. The simulations results show the landslide block slid with a low velocity of 8 m/s for about 100 m. Then, Froude number of landslide increases from 2 to 3 when passing the high and steep terrain, indicating that landslide change to inertial dominated with potential same Froude behavior of classic debris flow. The rupture mass slid with the peak velocity of 23 m/s and diverged in two gullies and ran out for about 600 m. The maximum velocity is 23 m/s in east gully while only 15 m/s in west gully. Compared with deep and incised valleys in the west, shallow and straight valley in the east decreases the deposit depth and further increases the velocity of landslide material with increased runout distance. This research may provide a fast flow path of back analyzing geo-hazards on complex terrain and serve as a basis for future research on long runout landslides.

Rainfall-induced landslides and debris flows in Mengdong Town, Yunnan Province, China

Abstract: On 2 September 2018, an intense rainstorm swept Mengdong Town in Yunnan Province of China, inducing a serious landslide and debris flow disaster with 10 deaths and 11 missing. Image interpretation, field survey, and slope stability analysis were used to examine the characteristics and initiation mechanism of this hazard. A total of 1774 landslide scars were identified, the area of which occupied 8.26% of the study region. Due to the spatial inhomogeneity of precipitation, these scars mainly concentrated along valleys in the lower part of the study area. Besides, well-forested hillslopes were more prone to landslides, indicating the limited role of trees in stabilizing slopes in extreme rainfall events. The initiation of landslides is mainly attributed to the week cohesion of the saturated clayey sand beneath the root zone, where the tensile resistance of roots was absent, and the increase of positive pore water pressure. Additionally, 288 landslide scars were situated adjacent to roads, demonstrating that the road construction activity had intensified the landslide disaster. Owing to the relatively low mobility of landslides, a considerable portion of landslide debris deposited on the valley floor in smaller watersheds (< 6.03 hm2), while the remaining portion entered high-order channels. In these channels, where stream power was relatively large, the woody debris and soil carried by landslides were entrained by streamflow, and debris flows were formed. Moreover, the magnitude of debris flow was amplified by the vertical and lateral erosion of the stream channel.

Investigating the effects of clay/sand content on depositional mechanisms of submarine debris flows through physical and numerical modeling

Abstract: In this study, we carried out a series of physical experiments using a submersed flume model to investigate how sand/clay content influences the depositional mechanism of submarine debris flows. A three-dimensional biphasic numerical model, with a Herschel-Bulkley rheology, was used to back-analyze the physical experiments. The calibrated numerical model was then used in a back-calculation to investigate the effects of viscosity on the deposition process. The results show that as submarine debris flows mix with water during the deposition process, shear stress at the slurry-water interface generates a vortex that leads to a swirl-wedge front head. High-viscosity slurry flows with a swirl-wedge front head travel at a higher aspect ratio and with a greater radius of rotation. Hydroplaning was observed when the front head was lifted by water during flow. The lifting height increased with flow depth fluctuation. Higher viscosity slurry was found to lift more rapidly due to its larger vortex and the decrease in density at the front head over time, both of which promote fluctuation. Although a high-density slurry has a greater lifting height, the ratio of lifting height to front head height is lower, indicating a smaller lifting force influence. Lower density flows have higher kinetic energy as the transfer of potential energy into kinetic energy is more efficient. Kinetic energy dissipation comprises three stages: (1) gravity-dominated coherent flow and hydroplaning lead to a rapid increase in kinetic energy; (2) sharp reduction in kinetic energy as slurry mixes with water, coherence and hydroplaning are reduced, and the influence of the shear stress increases; (3) slurry mixed very well with water, turbidity current dominates the kinetic energy dissipation. High-density slurry dissipates quicker in the last two stages. In stage 3, which dominates the temporal evolution of the debris flow, the Froude number is lower than 1, the flow thins and elongates, and the deposition process of submarine debris flows is dominated by gravity, and the difference of morphology of the different cases become clear.

A case study and implication: particle finite element modelling of the 2010 Saint-Jude sensitive clay landslide

Abstract: Modelling of landslides in sensitive clays has long been recognised as a challenge. The strength reduction of sensitive clays when undergoing plastic deformation makes the failure proceed in a progressive manner such that a small slope failure may lead to a series of retrogressive failures and thus to an unexpected catastrophic landslide. The clay in the entire process may mimic both solid-like (when it is intact) and fluid-like (when fully remoulded, especially for quick clays) behaviours. Thereby, a successful numerical prediction of landslides in sensitive clays requires not only a robust numerical approach capable of handling extreme material deformation but also a sophisticated constitutive model to describe the complex clay behaviour. In this paper, the particle finite element method (PFEM) associated with an elastoviscoplastic model with strain softening is adopted for the reconstruction of the 2010 Saint-Jude landslide, Quebec, Canada, and detailed comparisons between the simulation results and available data are carried out. It is shown that the present computational framework is capable of quantitatively reproducing the multiple rotational retrogressive failure process, the final run-out distance and the retrogression distance of the Saint-Jude landslide. Furthermore, the failure mechanism and the kinematics of the Saint-Jude landslide and the influence of the clay viscosity are investigated numerically, and in addition, their implications to real landslides in sensitive clays are discussed.

Field investigation and numerical study of a siltstone slope instability induced by excavation and rainfall

Abstract: Weak rock slope instabilities are a common engineering problem during highway construction in South China. This study focused on a siltstone slope instability, which was induced during the construction of an expressway in Guangdong Province of China. Field monitoring and numerical simulation analyses were performed to examine the failure mechanism and formation processes of this landslide which is associated with construction activities and a period of prolonged rainfall. According to the characteristics of the slope deformation and the monitoring data, the slope deformation can be divided into two stages: a period of slow creep caused by excavation and an accelerated sliding period triggered by rainfall. Numerical simulation results show that during the excavation process, large horizontal displacement occurs at the front edge of the slope, and the initial plastic zone develops, resulting in a shallow landslide. During 20 days of continuous rainfall, the water content in the shallow layer of the slope increases continuously, and a transient saturated area forms at the surface of the slope. Within 7 days after the rain stops, the zero pore pressure surface of the slope gradually moves towards the interior of the slope, and the plastic zone begins to extend to the top of the slope. In addition, rainwater seeps down along the cracks to form a penetrating zone, thus accelerating the process of rock and soil mass softening, which further reduces the factor of safety of the slope. The combined effects of the excavation and rainfall ultimately lead to the failure of the siltstone slope; however, continuous rainfall is the key factor triggering deep sliding. The deformation and failure of the slope mainly undergo four stages: local collapse of the slope surface, formation of the plastic zone at the foot of the slope, bulging at the toe, and formation of tension cracks in the crown of the landslide. The failure mode of the siltstone slope belongs to be a retrogressive-type of the front edge bulging and trailing edge tension cracking. Based on the deformation characteristics and the failure mechanism of the landslide, comprehensive control measures including interim emergency mitigation measures and long-term mitigation measures are proposed.

Acceleration of thaw slump during 1997–2017 in the Qilian Mountains of the northern Qinghai-Tibetan plateau

Abstract: Climate warming increases thermokarst landscapes and thus leads to land degradation in the Circum-Arctic regions. Thermokarst landscapes were estimated to cover ~ 20% of the northern permafrost region, and their development is related to ground ice content and topographic conditions. However, changes in thaw slump distribution and development in mid- to low-latitude permafrost areas largely remain unknown. Here, we selected the Qilian Mountains of the northern Qinghai-Tibetan Plateau (QTP) as the study area. Combined with field investigation to measure the boundary of thaw slump using Real Time Kinematic (RTK), we analyzed historical changes in thaw slumps using 1969 and 1997 aerial photographs, 2009 and 2015 satellite imagery, and 2017 unmanned aerial vehicle (UAV) aerial images. The results showed that there are 15 thaw slumps covering an area of 0.03 km2, which were mainly distributed in terrain with a slope of 2–8° and elevation of 3552–3611 m. For the time periods of 1997–2009, 2009–2015, and 2015–2017, the average increase rates of thaw slump areas were approximately 61.8, 60.0, and 156.8 m2/y, and annual headwall retreat rates were approximately 1.3, 1.6 and 2.0 m/y. Based on the slopes, aspects, and altitudes from the Digital Elevation Model generated by UAV imagery, it suggested that geomorphic factors have no significant effect on the growth rates of thaw slump due to high heterogeneities of alpine environments. Our results showed that thaw slumps with polycyclic and active characteristics covered at least 0.9% of northern QTP permafrost regions and are expected to rapidly accelerate with climate warming.

Structurally controlled rock slope deformation in northern Norway

Abstract: Gravitational forcing of oversteepened rock mass leads to progressive failure, including rupture, creeping, sliding and eventual avalanching of the unstable mass. As the point of rupture initiation typically follows pre-existing structural discontinuities within the rock mass, understanding the structural setting of slopes is necessary for an accurate characterisation of the hazards and estimation of the risk to life and infrastructure. Northern Norway is an alpine region with a high frequency of large rock slope deformations. Inherited structures in the metamorphic bedrock create a recurring pattern of anisotropy, that, given certain valley orientations, causes mass instability. We review the geomorphology, structural mechanics and kinematics of nine deforming rock slopes in Troms County, with the aim of linking styles of deformation. The limits of the unstable rock mass follow either foliation planes, joint planes or inherited faults, depending on the valley aspect, slope angle, foliation dip and proximity to fault structures. We present an updated geotechnical model of the different failure mechanisms, based on the interpretations at each site of the review.

Design of active debris flow mitigation measures: a comprehensive analysis of existing impact models

Abstract: Debris flows occur in mountainous areas characterized by steep slope and occasional severe rainstorms. The massive urbanization in these areas raised the importance of studying and mitigating these phenomena. Concerning the strategy of protection, it is fundamental to evaluate both the effect of the magnitude (that concerns the definition of the hazard), in terms of mobilized volume and travel distance, and the best technical protection structures (that concerns the mitigation measures) to reduce the existing risk to an acceptable residual one. In particular, the mitigation measure design requires the evaluation of the effects of debris flow impact forces against them. In other words, once it is established that mitigation structures are required, the impacting pressure shall be evaluated and it should be verified that it does not exceed barrier resistance. In this paper, the author wants to focus on the definition and the evaluation of the impacting load of debris flows on protection structures: a critical review of main existing models and equations treated in scientific literature is here presented. Although most of these equations are based on solid physical basis, they are always affected by an empirical nature due to the presence of coefficients for fitting the numerical results with laboratory and, less frequently, field data. The predicting capability of these equations, namely the capability of fitting experimental/field data, is analysed and evaluated using ten different datasets available in scientific literature. The purpose of this paper is to provide a comprehensive analysis of the existing debris flow impact models, highlighting their strong points and limits. Moreover, this paper could have a practical aspect by helping engineers in the choice of the best technical solution and the safe design of debris flow protection structures. Existing design guidelines for debris flow protection barrier have been analysed. Finally, starting from the analysis of the hydro-static model response to fit field data and introducing some practical assumptions, an empirical formula is proposed for taking into account the dynamic effects of the phenomenon.