Landslide Susceptibility Mapping Using RBFN-Based Ensemble Machine Learning Models

This study was aimed to prepare landslide susceptibility maps for the Pithoragarh district in Uttarakhand,India,using advanced ensemble models that combined Radial Basis Function Networks(RBFN)with three ensemble learning techniques:DAGGING(DG),MULTIBOOST(MB),and ADABOOST(AB).This combination resulted in three distinct ensemble models:DG-RBFN,MB-RBFN,and AB-RBFN.Additionally,a traditional weighted method,Information Value(IV),and a benchmark machine learning(ML)model,Multilayer Perceptron Neural Network(MLP),were employed for comparison and validation.The models were developed using ten landslide conditioning factors,which included slope,aspect,elevation,curvature,land cover,geomorphology,overburden depth,lithology,distance to rivers and distance to roads.These factors were instrumental in predicting the output variable,which was the probability of landslide occurrence.Statistical analysis of the models’performance indicated that the DG-RBFN model,with an Area Under ROC Curve(AUC)of 0.931,outperformed the other models.The AB-RBFN model achieved an AUC of 0.929,the MB-RBFN model had an AUC of 0.913,and the MLP model recorded an AUC of 0.926.These results suggest that the advanced ensemble ML model DG-RBFN was more accurate than traditional statistical model,single MLP model,and other ensemble models in preparing trustworthy landslide susceptibility maps,thereby enhancing land use planning and decision-making.

Empirical-Statistical Models Based on Remote Sensing for Estimating the Volume of Landslides Induced by the Wenchuan Earthquake

The Wenchuan Ms 8.0 earthquake on May 12, 2008 induced a huge number of landslides. The distribution and volume of the landslides are very important for assessing risks and understanding the landslide - debris flow - barrier lake - bursts flood disaster chain. The number and the area of landslides in a wide region can be easily obtained by remote sensing technique, while the volume is relatively difficult to obtain because it requires some detailed geometric information of slope failure surface and sub-surface. Different empirical models for estimating landslide volume were discussed based on the data of 107 landslides in the earthquake-stricken area. The volume data of these landslides were collected by field survey. Their areas were obtained by interpreting remote sensing images while their apparent friction coefficients and height were extracted from the images unifying DEM (digital elevation model). By analyzing the relationships between the volume and the area, apparent friction coefficients, and the height, two models were established, one for the adaptation of a magnitude scale landslide events in a wide range of region, another for the adaptation in a small scope. The correlation coefficients (R2) are 0.7977 and 0.8913, respectively. The results estimated by the two models agree well with the measurement data.

Exploring landslide erosion volume-area scaling relationships by slip depth using changes in DTMs for basin sediment volume estimation

Landslides produce large quantities of sediment deposits and reduce reservoir life. This study investigated landslides at the Shihmen Reservoir basin in Taiwan that were induced by Typhoon Sinlaku and Typhoon Jangmi in 2008. We formulate scaling relationships between landslide erosion volume and area and conclude that sediment budget can be estimated based on the easier-todetermine landslide erosion area. The methodologies applied for the investigation were geomorphological analysis through 5 m × 5 m digital terrain models(DTMs) of the basin created before and after the landslide events and spatial analysis through a geographic information system. The erosion area and volume of landslides were measured through the subtraction of DTMs produced before and after the events. Statistical analysis revealed that the landslide erosion frequency–magnitude distribution exhibited power-law behaviors with a scaling exponent of 2.15 for the frequency–area distribution and 1.66 for the frequency–volume distribution. This paper proposes different scaling relationships for different moving depths, and landslide erosion volumes were estimated on the basis of depth; thus, landslides of different scales can be distinguished to avoid errors in volume estimation. Two different scaling exponents are proposed: 1.21 for landslide erosions with depths of less than 2 m and 1.01 for landslide erosions with depths of more than 2 m. The proposed scaling relationships are practical for landslide erosion volume estimation by different depths according to the landslide area, and they can provide preliminary results for sediment budget planning in a reservoir basin.

Volume expansion rates of seismic landslides and influencing factors: A case study of the 2008 Wenchuan earthquake

When a big landside occurs, source material can change into loose deposit during its runout, causing the increase of the total landslide volume to some extent. Such changes can influence the quantification of seismic landslides. The objective of this paper was to study the volume expansion rate of landslides based on the data of 1417 co-seismic landslides triggered by the 2008 Wenchuan, China Mw 7.9 earthquake. We also analyzed the correlations between this rate and landslide geometric parameters(volume, height(H), length-width ratio(L/W), length-height ratio(L/H)), and environmental factors(peak ground acceleration(PGA), lithology, slope angle and aspect). The results show that the total source volume of the 1417 landslides is 1248 million m3, while the total volume of the deposit is 1501 million m3, which means the total volume expanding rate(Et) is 20.3% with the average volume expansion rate(Ea) 22.6%. The analysis indicates that volume expansion rate generally decreases with the increasing volume and height of landslides, while becoming larger with increasing L/H and L/W. Besides, the volume expansion rate is closely related to the landslide type and the volume scale of landslides. This study analyses volume change of co-seismic landslides deeply, permitting to help the correct quantification of the source volume and deposit volume of seismic landslide and a useful reference for the correct quantification of landslide volume.

The Volume Calculation Method of Rock Collapses and Loess Landslides Triggered by the 1556 AD Huaxian M81/2 Earthquake in Shaanxi Province, China

Accurate volume calculation of each individual landslide triggered by strong historical earthquakes can help understand the characteristics of the typical earthquake-induced landslides,thus providing significant information for the modification of the focal parameters of historical earthquakes.In this study,we select one rock fall and three loess landslides triggered by the 1556 AD Huaxian M8⅟􀄟earthquake,compute their volumes using the low-altitude high-precision Unmanned Aerial Vehicle(UAV)photogrammetry and landslide profile restoration methods.The results show that:①the whole influencing area of the Huangjiagou Rock Fall is approximately 3.03×105 m2 and the area of the collapsed rock accumulated at the slope foot is 3.33×104 m2,accounting for approximately 10%of the entire influencing range.However,the estimated volume of the collapsed rock is only 0.699×106 m3,indicating a rock fall with large influencing range but limited collapsed rock;②the geological form of thethree loess landslides are preserved intactly,with volumes of 0.283×108 m3,0.074×108 m3,and 0.377×108 m3.These important geological hazard relics reflect the strong vibrations and severe casualties in the meizoseismal area;③loess landslides are the key reason of the serious death toll in the hilly-gully loess area.Our new method can be used to estimate the influencing area and the actual volume of each individual landslide,and rationally evaluate the role of earthquake landslides in the disaster.In addition,quantitative research on secondary disasters triggered by strong historical earthquakes is beneficial for understanding the surface process and focal parameters of the earthquakes.

The Calculation Method for the Volume of Landslides and Dammed Lake Sediments Triggered by a Strong Historical Earthquake in the Loess Plateau:A Case Study of Qiuzigou, Gansu Province, Northwest China

The quantitative calculation of the volume of large earthquake-triggered landslides and related dammed lake sediments is of great significance in the study of secondary disasters and focal parameters of strong historical earthquakes.In this study,the dammed lake induced by Qishan M7 earthquake(Lingtai County,Gansu Province,Northwest China)is selected as the research object.Based on the information collected from the 4 boreholes in the dammed lake area,we further take advantage of the lowlevel Unmanned Aerial Vehicle(UAV)photogrammetry and the morphology recovery method,to calculate the volume of the dammed lake and landslides,respectively.Finally,major conclusions are obtained as follows:①the AMS-14C age at the bottom of the Qiuzigou Dammed Lake sediments is 2890±30 BP,which coincides with the 780 BC Qishan earthquake;furthermore,the Qiuzigou Landslides seem to have been triggered by the earthquake,forming an enclosed dammed lake deposition environment after the upstream sediments accumulate;②the Qiuzigou landslides are opposite-sliding landslides that have blocked the river valley;in detail,landslide volumes at the right and left banks are 235×104 m3 and 229×104 m3,respectively.The length of the dammed lake is 2.6 km,with a thickness of approximately 43 m near the landslides,and the total sedimentary volume is 573×104 m3;③the erosion rate of Qiuzigou Landslide Dammed Lake is 0.44 mm/a,the accumulation rate is 15.05 mm/a,and the soil erosion modulus is 593 t/(km2/a),characterized as slight erosion.Quantitative research on the formation of landslides and dammed lakes from strong historical earthquakes is vital for increasing our understanding of the vibrational characteristics and surface action processes of these types of earthquakes.

Thermo-hydro-poro-mechanical responses of a reservoir-induced landslide tracked by high-resolution fiber optic sensing nerves

Thermo-poro-mechanical responses along sliding zone/surface have been extensively studied.However,it has not been recognized that the potential contribution of other crucial engineering geological interfaces beyond the slip surface to progressive failure.Here,we aim to investigate the subsurface multiphysics of reservoir landslides under two extreme hydrologic conditions(i.e.wet and dry),particularly within sliding masses.Based on ultra-weak fiber Bragg grating(UWFBG)technology,we employ specialpurpose fiber optic sensing cables that can be implanted into boreholes as“nerves of the Earth”to collect data on soil temperature,water content,pore water pressure,and strain.The Xinpu landslide in the middle reach of the Three Gorges Reservoir Area in China was selected as a case study to establish a paradigm for in situ thermo-hydro-poro-mechanical monitoring.These UWFBG-based sensing cables were vertically buried in a 31 m-deep borehole at the foot of the landslide,with a resolution of 1 m except for the pressure sensor.We reported field measurements covering the period 2021 and 2022 and produced the spatiotemporal profiles throughout the borehole.Results show that wet years are more likely to motivate landslide motions than dry years.The annual thermally active layer of the landslide has a critical depth of roughly 9 m and might move downward in warmer years.The dynamic groundwater table is located at depths of 9e15 m,where the peaked strain undergoes a periodical response of leap and withdrawal to annual hydrometeorological cycles.These interface behaviors may support the interpretation of the contribution of reservoir regulation to slope stability,allowing us to correlate them to local damage events and potential global destabilization.This paper also offers a natural framework for interpreting thermo-hydro-poro-mechanical signatures from creeping reservoir bank slopes,which may form the basis for a landslide monitoring and early warning system.

Fiber optic monitoring of an anti-slide pile in a retrogressive landslide

Anti-slide piles are one of the most important reinforcement structures against landslides,and evalu-ating the working conditions is of great significance for landslide mitigation.The widely adopted analytical methods of pile internal forces include cantilever beam method and elastic foundation beam method.However,due to many assumptions involved in calculation,the analytical models cannot be fully applicable to complex site situations,e.g.landslides with multi-sliding surfaces and pile-soil interface separation as discussed herein.In view of this,the combination of distributed fiber optic sensing(DFOS)and strain-internal force conversion methods was proposed to evaluate the working conditions of an anti-sliding pile in a typical retrogressive landslide in the Three Gorges reservoir area,China.Brillouin optical time domain reflectometry(BOTDR)was utilized to monitor the strain distri-bution along the pile.Next,by analyzing the relative deformation between the pile and its adjacent inclinometer,the pile-soil interface separation was profiled.Finally,the internal forces of the anti-slide pile were derived based on the strain-internal force conversion method.According to the ratio of calculated internal forces to the design values,the working conditions of the anti-slide pile could be evaluated.The results demonstrated that the proposed method could reveal the deformation pattern of the anti-slide pile system,and can quantitatively evaluate its working conditions.

Uncertainties in landslide susceptibility prediction:Influence rule of different levels of errors in landslide spatial position

The accuracy of landslide susceptibility prediction(LSP)mainly depends on the precision of the landslide spatial position.However,the spatial position error of landslide survey is inevitable,resulting in considerable uncertainties in LSP modeling.To overcome this drawback,this study explores the influence of positional errors of landslide spatial position on LSP uncertainties,and then innovatively proposes a semi-supervised machine learning model to reduce the landslide spatial position error.This paper collected 16 environmental factors and 337 landslides with accurate spatial positions taking Shangyou County of China as an example.The 30e110 m error-based multilayer perceptron(MLP)and random forest(RF)models for LSP are established by randomly offsetting the original landslide by 30,50,70,90 and 110 m.The LSP uncertainties are analyzed by the LSP accuracy and distribution characteristics.Finally,a semi-supervised model is proposed to relieve the LSP uncertainties.Results show that:(1)The LSP accuracies of error-based RF/MLP models decrease with the increase of landslide position errors,and are lower than those of original data-based models;(2)70 m error-based models can still reflect the overall distribution characteristics of landslide susceptibility indices,thus original landslides with certain position errors are acceptable for LSP;(3)Semi-supervised machine learning model can efficiently reduce the landslide position errors and thus improve the LSP accuracies.

How do the landslide and non-landslide sampling strategies impact landslide susceptibility assessment? d A catchment-scale case study from China

The aim of this study is to investigate the impacts of the sampling strategy of landslide and non-landslide on the performance of landslide susceptibility assessment(LSA).The study area is the Feiyun catchment in Wenzhou City,Southeast China.Two types of landslides samples,combined with seven non-landslide sampling strategies,resulted in a total of 14 scenarios.The corresponding landslide susceptibility map(LSM)for each scenario was generated using the random forest model.The receiver operating characteristic(ROC)curve and statistical indicators were calculated and used to assess the impact of the dataset sampling strategy.The results showed that higher accuracies were achieved when using the landslide core as positive samples,combined with non-landslide sampling from the very low zone or buffer zone.The results reveal the influence of landslide and non-landslide sampling strategies on the accuracy of LSA,which provides a reference for subsequent researchers aiming to obtain a more reasonable LSM.

Spatiotemporal deformation characteristics of Outang landslide and identification of triggering factors using data mining

Since the impoundment of Three Gorges Reservoir(TGR)in 2003,numerous slopes have experienced noticeable movement or destabilization owing to reservoir level changes and seasonal rainfall.One case is the Outang landslide,a large-scale and active landslide,on the south bank of the Yangtze River.The latest monitoring data and site investigations available are analyzed to establish spatial and temporal landslide deformation characteristics.Data mining technology,including the two-step clustering and Apriori algorithm,is then used to identify the dominant triggers of landslide movement.In the data mining process,the two-step clustering method clusters the candidate triggers and displacement rate into several groups,and the Apriori algorithm generates correlation criteria for the cause-and-effect.The analysis considers multiple locations of the landslide and incorporates two types of time scales:longterm deformation on a monthly basis and short-term deformation on a daily basis.This analysis shows that the deformations of the Outang landslide are driven by both rainfall and reservoir water while its deformation varies spatiotemporally mainly due to the difference in local responses to hydrological factors.The data mining results reveal different dominant triggering factors depending on the monitoring frequency:the monthly and bi-monthly cumulative rainfall control the monthly deformation,and the 10-d cumulative rainfall and the 5-d cumulative drop of water level in the reservoir dominate the daily deformation of the landslide.It is concluded that the spatiotemporal deformation pattern and data mining rules associated with precipitation and reservoir water level have the potential to be broadly implemented for improving landslide prevention and control in the dam reservoirs and other landslideprone areas.

Uncertainties of landslide susceptibility prediction: Influences of random errors in landslide conditioning factors and errors reduction by low pass filter method

In the existing landslide susceptibility prediction(LSP)models,the influences of random errors in landslide conditioning factors on LSP are not considered,instead the original conditioning factors are directly taken as the model inputs,which brings uncertainties to LSP results.This study aims to reveal the influence rules of the different proportional random errors in conditioning factors on the LSP un-certainties,and further explore a method which can effectively reduce the random errors in conditioning factors.The original conditioning factors are firstly used to construct original factors-based LSP models,and then different random errors of 5%,10%,15% and 20%are added to these original factors for con-structing relevant errors-based LSP models.Secondly,low-pass filter-based LSP models are constructed by eliminating the random errors using low-pass filter method.Thirdly,the Ruijin County of China with 370 landslides and 16 conditioning factors are used as study case.Three typical machine learning models,i.e.multilayer perceptron(MLP),support vector machine(SVM)and random forest(RF),are selected as LSP models.Finally,the LSP uncertainties are discussed and results show that:(1)The low-pass filter can effectively reduce the random errors in conditioning factors to decrease the LSP uncertainties.(2)With the proportions of random errors increasing from 5%to 20%,the LSP uncertainty increases continuously.(3)The original factors-based models are feasible for LSP in the absence of more accurate conditioning factors.(4)The influence degrees of two uncertainty issues,machine learning models and different proportions of random errors,on the LSP modeling are large and basically the same.(5)The Shapley values effectively explain the internal mechanism of machine learning model predicting landslide sus-ceptibility.In conclusion,greater proportion of random errors in conditioning factors results in higher LSP uncertainty,and low-pass filter can effectively reduce these random errors.

Spatial distribution of shallow landslides caused by Typhoon Lekima in 2019 in Zhejiang Province, China

In recent years, the coastal region of Southeast China has witnessed a significant increase in the frequency and intensity of extreme rainfall events associated with landfalling typhoons. The hilly and mountainous terrain of this area, combined with rapid rainfall accumulation, has led to a surge in flash floods and severe geological hazards. On August 10, 2019, Typhoon Lekima made landfall in Zhejiang Province, China, and its torrential rainfall triggered extensive landslides, resulting in substantial damage and economic losses. Utilizing high-resolution satellite images, we compiled a landslide inventory of the affected area, which comprises a total of 2,774 rainfallinduced landslides over an area of 2965 km2. The majority of these landslides were small to mediumsized and exhibited elongated, clustered patterns. Some landslides displayed characteristics of high-level initiation, obstructing or partially blocking rivers, leading to the formation of debris dams. We used the inventory to analyze the distribution pattern of the landslides and their relationship with topographical, geological, and hydrological factors. The results showed that landslide abundance was closely related to elevation, slope angle, faults, and road density. The landslides were predominantly located in hilly and low mountainous areas, with elevations ranging from 150 to 300 m, slopes of 20 to 30 degrees, and a NE-SE aspect. Notably, we observed the highest Landslide Number Density(LND) and Landslide Area Percentage(LAP) in the rhyolite region. Landslides were concentrated within approximately 4 km on either side of fault zones, with their size and frequency negatively correlated with distances to faults, roads, and river systems. Furthermore, under the influence of typhoons, regions with denser vegetation cover exhibited higher landslide density, reaching maximum values in shrubland areas. In areas experiencing significantly increased concentrated rainfall, landslide density also showed a corresponding rise. In terms of spatial distribution, the rainfall-triggered landslides primarily occurred in the northeastern part of the study area, particularly in regions characterized by complex topography such as Shanzao Village in Yantan Town, Xixia Township, and Shangzhang Township. The research findings offer crucial data on the rainfallinduced landslides triggered by Typhoon Lekima, shedding light on their spatial distribution patterns. These findings provide valuable references for mitigating risks and planning reconstruction in typhoon-affected area.

Time series prediction of reservoir bank landslide failure probability considering the spatial variability of soil properties

Historically,landslides have been the primary type of geological disaster worldwide.Generally,the stability of reservoir banks is primarily affected by rainfall and reservoir water level fluctuations.Moreover,the stability of reservoir banks changes with the long-term dynamics of external disastercausing factors.Thus,assessing the time-varying reliability of reservoir landslides remains a challenge.In this paper,a machine learning(ML)based approach is proposed to analyze the long-term reliability of reservoir bank landslides in spatially variable soils through time series prediction.This study systematically investigated the prediction performances of three ML algorithms,i.e.multilayer perceptron(MLP),convolutional neural network(CNN),and long short-term memory(LSTM).Additionally,the effects of the data quantity and data ratio on the predictive power of deep learning models are considered.The results show that all three ML models can accurately depict the changes in the time-varying failure probability of reservoir landslides.The CNN model outperforms both the MLP and LSTM models in predicting the failure probability.Furthermore,selecting the right data ratio can improve the prediction accuracy of the failure probability obtained by ML models.

A local rainfall-triggered giant landslide occurred in a region along a high-speed railway on the Qinghai‒Tibetan Plateau

An increasing number of geological hazards along high-speed railways on the Qinghai‒Tibetan Plateau have occurred and have resulted in a profound influence on old infrastructure,which has attracted increasing attention.The landslide event that occurred on September 15,2022,in Jiujiawan village,Xining city,Qinghai Province,is a typical case.Based on field investigations and remote sensing interpretations,a comprehensive analysis was conducted on the landslide.Additionally,the potential secondary failure of the current Jiujiawan landslide was assessed using Fast Lagrangian Analysis of Continua in Three Dimensions(FLAC3D).Based on the application of the small baseline subset-interferometric synthetic aperture radar(SBAS-InSAR)technique to SAR images from February 24,2017 to September 14,2022,a significant westward horizontal deformation was found to have been formed prior to the occurrence of the landslide.The maximum annual average deformation rate in the line of sight(LOS)direction reached-45 mm/yr,with a maximum cumulative deformation of-178 mm.This value was consistent with the continual increase in annual precipitation(2.51 mm/yr)prior to the occurrence of the landslide.The accumulated precipitation before the landslide was 279.8 mm,accounting for 54.2%of the total annual precipitation,with a particularly notable surge in monthly precipitation observed during August(250.3 mm).Additionally,the occurrence of a seismic event with a magnitude of Ms 6.9 in Menyuan County,80 km away from Xining,could be a potential triggering factor to the landslide,as evidenced by an abrupt subsidence alteration observed prior to and following the earthquake.The maximum subsidence in the line of sight(LOS)direction exceeded 11 mm,exhibiting a highly consistent spatial distribution with the occurrence range of landslides.These results suggest that the Jiujiawan landslide was likely induced by earthquake events in the early stage and heavy rainfall in the later stage.The FLAC3D numerical simulation show that after the landslide,the slope remained marginally stable under natural conditions;however,it is susceptible to reactivation with heavy rainfall.

Numerical Simulation of Rainfall-induced Xianchi Reservoir Landslide in Yunyang,Chongqing,China

A calamitous landslide happened at 22:00 on September 1,2014 in the Yunyang area of Chongqing City,southwest China,enforcing the evacuation of 508 people and damaging 23 buildings.The landslide volume comprised 1.44 million m^(3) of material in the source area and 0.4 million m^(3) of shoveled material.The debris flow runout extended 400 m vertically and 1600 m horizontally.The Xianchi reservoir landslide event has been investigated as follows:(1)samples collected from the main body of landslide were carried out using GCTS ring shear apparatus;(2)the parameters of shear and pore water pressure have been measured;and(3)the post-failure characteristics of landslide have been analyzed using the numerical simulation method.The excess pore-water pressure and erosion in the motion path are considered to be the key reasons for the long-runout motion and the scale-up of landslides,such as that at Xianchi,were caused by the heavy rainfall.The aim of this paper is to acquired numerical parameters and the basic resistance model,which is beneficial to improve simulation accuracy for hazard assessment for similar to potentially dangerous hillslopes in China and elsewhere.

Landslide hazard susceptibility evaluation based on SBAS-InSAR technology and SSA-BP neural network algorithm:A case study of Baihetan Reservoir Area

Landslide hazard susceptibility evaluation takes on critical significance in early warning and disaster prevention and reduction.In order to solve the problems of poor effectiveness of landslide data and complex calculation of weights for multiple evaluation factors in the existing landslide susceptibility evaluation models,in this study,a method of landslide hazard susceptibility evaluation is proposed by combining SBAS-InSAR(Small Baseline Subsets-Interferometric Synthetic Aperture Radar)and SSA-BP(Sparrow Search Algorithm-Back Propagation)neural network algorithm.The SBAS-InSAR technology is adopted to identify potential landslide hazards in the study area,update the cataloging data of landslide hazards,and 11 evaluation factors are chosen for constructing the SSA-BP model for training and validation.Baihetan Reservoir area is selected as a case study for validation.As indicated by the results,the application of SBAS-InSAR technology,combined with both ascending and descending orbit data,effectively addresses the incomplete identification of landslide hazards caused by geometric distortion of single orbit SAR data(e.g.,shadow,overlay,and perspective contraction)in deep canyon areas,thereby enabling the acquisition of up-to-date landslide hazard data.Moreover,in comparison to the conventional BP(Back Propagation)algorithm,the accuracy of the model constructed by the SSA-BP algorithm exhibits a significant increase,with mean squared error and mean absolute error reduced by 0.0142 and 0.0607,respectively.Additionally,during the process of susceptibility evaluation,the SSA-BP model effectively circumvents the issue of considerable manual interventions in calculating the weight of evaluation factors.The area under the curve of this model reaches 0.909,surpassing BP(0.835),random forest(0.792),and the information value method(0.699).The risk of landslide occurrence in the Baihetan Reservoir area is positively correlated with slope,surface temperature,and deformation rate,while it is negatively correlated with fault distance and normalized difference vegetation index.Geological lithology exerts minimal influence on the occurrence of landslides,with the risk being low in forest land and high in grassland.The method proposed in this study provides a useful reference for disaster prevention and mitigation departments to perform landslide hazard susceptibility evaluations in deep canyon areas under complex geological conditions.

Probabilistic back-analysis of rainfall-induced landslides for slope reliability prediction with multi-source information

Probabilistic back-analysis is an important means to infer the statistics of uncertain soil parameters,making the slope reliability assessment closer to the engineering reality.However,multi-source information(including test data,monitored data,field observation and slope survival records)is rarely used in current probabilistic back-analysis.Conducting the probabilistic back-analysis of spatially varying soil parameters and slope reliability prediction under rainfalls by integrating multi-source information is a challenging task since thousands of random variables and high-dimensional likelihood function are usually involved.In this paper,a framework by integrating a modified Bayesian Updating with Subset simulation(mBUS)method with adaptive Conditional Sampling(aCS)algorithm is established for the probabilistic back-analysis of spatially varying soil parameters and slope reliability prediction.Within this framework,the high-dimensional probabilistic back-analysis problem can be easily tackled,and the multi-source information(e.g.monitored pressure heads and slope survival records)can be fully used in the back-analysis.A real Taoyuan landslide case in Taiwan,China is investigated to illustrate the effectiveness and performance of the established framework.The findings show that the posterior knowledge of soil parameters obtained from the established framework is in good agreement with the field observations.Furthermore,the updated knowledge of soil parameters can be utilized to reliably predict the occurrence probability of a landslide caused by the heavy rainfall event on September 12,2004 or forecast the potential landslides under future rainfalls in the Fuhsing District of Taoyuan City,Taiwan,China.

Multistate transition and coupled solid-liquid modeling of motion process of long-runout landslide

The recognition,repetition and prediction of the post-failure motion process of long-runout landslides are key scientific problems in the prevention and mitigation of geological disasters.In this study,a new numerical method involving LPF3D based on a multialgorithm and multiconstitutive model was proposed to simulate long-runout landslides with high precision and efficiency.The following results were obtained:(a)The motion process of landslides showed a steric effect with mobility,including gradual disintegration and spreading.The sliding mass can be divided into three states(dense,dilute and ultradilute)in the motion process,which can be solved by three dynamic regimes(friction,collision,and inertial);(b)Coupling simulation between the solid grain and liquid phases was achieved,focusing on drag force influences;(c)Different algorithms and constitutive models were employed in phase-state simulations.The volume fraction is an important indicator to distinguish different state types and solid‒liquid ratios.The flume experimental results were favorably validated against long-runout landslide case data;and(d)In this method,matched dynamic numerical modeling was developed to better capture the realistic motion process of long-runout landslides,and the advantages of continuum media and discrete media were combined to improve the computational accuracy and efficiency.This new method can reflect the realistic physical and mechanical processes in long-runout landslide motion and provide a suitable method for risk assessment and pre-failure prediction.

Deformation,structure and potential hazard of a landslide based on InSAR in Banbar county,Xizang(Tibet)

The Tibetan Plateau is characterized by complex geological conditions and a relatively fragile ecological environment.In recent years,there has been continuous development and increased human activity in the Tibetan Plateau region,leading to a rising risk of landslides.The landslide in Banbar County,Xizang(Tibet),have been perturbed by ongoing disturbances from human engineering activities,making it susceptible to instability and displaying distinct features.In this study,small baseline subset synthetic aperture radar interferometry(SBAS-InSAR)technology is used to obtain the Line of Sight(LOS)deformation velocity field in the study area,and then the slope-orientation deformation field of the landslide is obtained according to the spatial geometric relationship between the satellite’s LOS direction and the landslide.Subsequently,the landslide thickness is inverted by applying the mass conservation criterion.The results show that the movement area of the landslide is about 6.57×10^(4)m^(2),and the landslide volume is about 1.45×10^(6)m^(3).The maximum estimated thickness and average thickness of the landslide are 39 m and 22 m,respectively.The thickness estimation results align with the findings from on-site investigation,indicating the applicability of this method to large-scale earth slides.The deformation rate of the landslide exhibits a notable correlation with temperature variations,with rainfall playing a supportive role in the deformation process and displaying a certain lag.Human activities exert the most substantial influence on the spatial heterogeneity of landslide deformation,leading to the direct impact of several prominent deformation areas due to human interventions.Simultaneously,utilizing the long short-term memory(LSTM)model to predict landslide displacement,and the forecast results demonstrate the effectiveness of the LSTM model in predicting landslides that are in a continuous development and movement phase.The landslide is still active,and based on the spatial heterogeneity of landslide deformation,new recommendations have been proposed for the future management of the landslide in order to mitigate potential hazards associated with landslide instability.