At about 21:30 pm on Sunday, 1 June 2025, a catastrophic landslide destroyed a temporary settlement of seasonal Cordyceps (medicinal fungus) collectors in Muta Town, Dingqing County, Tibet. About 2.0 million m3 of rock detached from the crest of the snow-covered mountains area, which resulted in 3 deaths and 7 people missing. This landslide and the severity of its effects were promoted by three concurrent phenomena: development conditions, time, and location. This event raised great concern both in China and beyond, and it exposed our lack of understanding of the landslide types induced by frost heave. This study uses extensive field investigation, satellite remote sensing, UAV aerial photography, rock mechanics testing, and numerical simulations to determine the deformation history, failure modes, and dynamic process of the landslide. The preliminary results suggest that the landslide source area experienced long-term deformation with the development of significant transverse cracks distributed along bedding planes. An exposed reverse slope was formed by the denudation of the strata of a Z-shaped fold, which had undergone frequent freeze-thaw cycles damage and finally triggered the landslide. The maximum landslide velocity was 30 m/s, the maximum accumulation thickness reached 24 m, and the landslide essentially ceased movement after 120 s. This paper reveals the failure mechanism whereby frost heave in high-altitude seasonal permafrost regions triggers slope instability in degraded red strata areas. Our findings provided new insight into landslide failure caused by frost heave damage on high hillslopes in alpine mountain areas.

Different momentum mapping schemes in the Material Point Method (MPM) significantly affect simulation results, yet their impacts in geomechanics remain underexplored. This paper compares the numerical performance of four common momentum mapping methods—PIC, FLIP, mixed PIC/FLIP, and APIC—in simulating large deformations, focusing on predicting slope failure modes. Four representative problems are analyzed: Rings Collision, aluminum rod collapse, progressive failure of sensitive soil, a large-scale Daguangbao slope collapse, and Yigong debris flow. We evaluate each method regarding kinetic energy evolution, energy conservation, numerical stability, and failure mode prediction, while examining the effect of grid spacing and time step on the results. Our findings indicate that while the PIC method is simple and relatively stable, its energy dissipation impedes accurate failure representation. The FLIP method preserves kinetic energy better and captures larger sliding distances, but it suffers from notable numerical noise and demands very small-time steps for stability. The mixed PIC/FLIP approach partially alleviates oscillatory behavior; however, it dissipates more energy than FLIP and is sensitive to both the mixing parameter and discretization conditions. In contrast, the APIC method exhibits excellent energy conservation with lower numerical damping. Notably, we reveal for the first time that the APIC method’s predicted failure mode in slope collapse simulations is strongly influenced by grid size and time step: smaller grids or larger time steps yield clearer local shear bands, while extremely small time steps or larger grids lead to anomalously irregular plastic shear regions. This study provides crucial theoretical insights and practical recommendations for choosing and tuning discretization parameters in geotechnical engineering. Overall, our results offer an essential reference for improving simulation accuracy and reliability in geotechnical applications.

Although coupled models integrating statistical analysis and machine learning are widely used in landslide susceptibility mapping, coupling strategy selection and model interpretability require in-depth investigation. Herein, a fundamental dataset pertaining to Liangshan Yi Autonomous Prefecture was constructed based on 2,681 landslide points and 14 environmental factors. Following factor screening, four statistical models [frequency ratio, information value, weights of evidence (WoE), and certainty factor models] were employed to quantify the environmental factors. The results were used as input for five machine learning models [nonlinear generalized additive, gated recurrent unit network, random forest (RF), support vector machine, and extreme gradient boosting models] to establish 20 coupled evaluation models. Model performance was assessed using the area under the receiver operating characteristic curve, F1 score, and landslide ratio. The Shapley additive explanations (SHAP) algorithm was used to interpret factor contribution mechanisms. Among the factor linkage methods, the WoE model achieved the highest prediction accuracy (mean F1 score = 0.935) and stability (standard deviation = 0.005). From the perspective of prediction models, the RF model yielded the optimal accuracy (mean F1 score = 0.942) and notable stability (standard deviation = 0.002). The WoE–RF coupled model was identified as the optimal solution for the study area. The SHAP algorithm effectively revealed factor operating mechanisms at global and local scales, successfully linking prediction results with landslide causation. Thus, this work provides a theoretical basis for constructing coupled models and optimizing the factor system.

The stress-strain behavior of remolded moraine soils is highly sensitive to freeze-thaw (F-T) cycles. However, studies on the evolution of the stress-strain curve and microstructure in remolded moraine soils under F-T cycles remain limited. To address this gap, a comprehensive experimental program was conducted that included F-T cycle tests, triaxial shear tests (TST), and micro-CT scanning to examine the influence of F-T cycles on a remolded moraine soil from Linzhi City, Xizang. In addition, a novel Duncan-Chang-Jiang-Lu (DC-JL) model has been proposed in this study to fit and predict stress-strain curves. The results show that F-T cycles cause a marked reduction in shear strength (TSS) and the initial average elastic modulus (E 0) following softening and a hyperbolic bulge failure mode. TSS decreases by 12%similar to 30%, and E 0 decreases by 30%similar to 45%. The DC-JL model reliably predicts stress-strain curves for any F-T cycle within the experimental range (N = 0-20) using results from N = 0, 2, 5, 10, and 20 cycles. Furthermore, the residual strength provided by the DC-JL model closely matches the experimental data. F-T cycles significantly impact mechanical behavior by deteriorating the internal microstructure, increasing micro-porosity (from 14.25% to 18.69%), and leading to the formation of numerous cracks. The findings of this study are not only valuable for a better understanding of freeze-thaw degradation in moraine soils but also for assessing long-term stability using the proposed new model.

The Wumeng Mountain Area is located in the transitional zone between the Sichuan Basin and the Yungui Plateau and between the Yunnan–Guizhou Plateau and the Qianzhong Mountainous Region. Widely distributed geological structural masses, characterized by gently inclined, interbedded sandstones, mudstones, and shales, are highly susceptible to geohazards. Rock avalanches triggered by the coupled effects of rainfall, earthquakes, and anthropogenic activities frequently occur. On February 8, 2025, a rock avalanche occurred in Jinping village, Junlian County, resulting in the tragic loss of 29 lives. To elucidate the disaster mechanisms and movement processes, field investigations, LiDAR mapping, and geological model analysis were conducted. The results indicate that the region is characterized by frequent moderate- to low-intensity earthquakes, strong weathering and unloading, historical mining activities, steep slopes with fractured rock structures and surface cracks; the development of loose deposits formed by terraced ancient landslides, rockfalls, alluvial deposits, and artificial accumulations at the base of the slope; and continuous rainfall triggering landslides and resulting in long-distance movement and deposition. Rock avalanches can be categorized into three types of disaster-related movement processes: slope rock mass instability and failure, cascading-scale amplification of scraping and plowing deposition bodies, and topography-constrained upheaval deposition. These findings provide reference data for analyzing the formation mechanisms and predicting the hazards of red-bedrock avalanches.

Areas of serpentine in ophiolitic melange zones often trigger large rock avalanches and exhibit strong movement. However, how the mechanism under which they post-failure hypermobility and long runout are unclear. Here, we identify and analyze a representative prehistoric rock avalanche, the Basu rock avalanche, with a large volume and a high mobility, which developed in the Nu River ophiolitic melange zone of the Tibetan Plateau. Based on field investigations, experimental, and Numerical simulation analyses we determined its development background and thus explained why it was hypermobile. This rock avalanche, with a volume of approximately 3.15 x 10(9) m(3), occurred around similar to 187 ka before present (B.P.). It developed on a marble nappe, with serpentine soft rock exposed locally at its base. It may have ultimately been triggered under seismic action, resulting in intense movement. The lubrication effect of fine-grained serpentine particles within the slip zone facilitated the hypermobility of the rock avalanche, resulting in both a large volume and an extended runout distance. This demonstrates that serpentine soft fine particles widely distributed in the suture zone are a typical lubricating material. The hypermobility of this large rock avalanche are striking and emphasizes the need to determine where, how and when these rare but high-magnitude rock avalanche events may occur. We proposed a new perspective on the triggering mechanisms of the rock avalanches and further verified the hypothesis of powder lubrication control effects.

Land surface temperature (LST) is crucial for understanding winter landslide evolution. This study combines Unmanned Aerial Vehicle (UAV) photogrammetry and infrared thermography (IRT) to monitor winter landslides in China’s Wumeng Mountain region. Using the Yangjiazhai landslide—induced by underground coal mining—as a case study, we demonstrate significant correlations between IRT-detected LST anomalies and surface cracks: (1) cracks with elevated temperatures are likely connected to subsurface goaf zones; (2) excessively widened cracks show no thermal anomalies due to enhanced air convection. The research reveals that key landslide components have distinct LST signatures, governed by differential soil–rock moisture and crack networks. For accurate high-altitude winter LST acquisition, UAV thermal surveys should be conducted under overcast, fog-free conditions to reduce solar interference. This validates UAV visible–infrared fusion for extracting landslide boundaries, cracks, slumping zones, bedrock patterns, and moisture distribution. The methodology establishes a new pathway for investigating winter landslide deformation and instability, confirming IRT’s operational viability in high-altitude alpine regions.

Substrate materials are recognized as exerting a critical and pervasive influence on the hypermobility and final morphology of rock avalanches. However, the fundamental mechanisms governing the interaction between avalanche masses and different substrates and the resultant effects on emplacement behaviours have yet to be fully elucidated, necessitating further empirical constraints. Based on satellite imagery, field investigations and statistical analyses, a comprehensive investigation was conducted into the geomorphology and sedimentology of the large Dzarang Tso Rock Avalanche (DTRA) in west-central Tibetan Plateau, China. The DTRA deposit, comprising similar to 105 Mm(3) leucocratic granitoid clasts, travelled 4.4 km on an unconfined flat environment, allowing rock avalanche processes to be ideally expressed. This deposit exhibits distinctive morphologies along its travel path, comprising lateral levees, transverse ridges, flowbands, a raised distal edge, and hummocks, which are attributable to the variable emplacement behaviours controlled by the substrate nature. The most remarkable morphology of longitudinal flowbands, which are commonly well-developed in glacier environments, was interpreted as a response to unusual rheology and spontaneous vibrations during the avalanche propagation under an ice-free condition, challenging the necessity of an ice surface in its formation. The contrasting morphological variation from flowbands to hummocks in the distal area is attributed to an abrupt change in substrate from the gravel-dominated piedmont alluvial deposits to the soft saturated lacustrine sediments. The soft saturated substrate contributes to lateral spreading to generate hummocky morphology and producing progressive bulldozing deformation that is expressed as a series of ductile and brittle soft sediment deformation structures, including convolute laminations, diapiric structures, water-escape structures, faults and reworked inclined stratifications in the frontal area of the avalanche deposit. Our research emphasizes that the travel path substrate plays a crucial role in shaping the final morphologies and in influencing the overall emplacement behaviour of unconfined rock avalanches. Compared to the gravel-dominated strong substrate, a soft saturated substrate can considerably enhance the runout of rock avalanches by providing a lubricated layer with low friction resistance. This finding is not only helpful in interpreting similar morphologies in other terrestrial counterparts but also provides a key geological prototype for guiding, applying, and testing further experimental, numerical, and theoretical models.

Calcareous cementation in moraine soils remains insufficiently studied, despite its critical role in influencing mechanical behavior and slope stability. To address this gap, a series of tests were conducted, including field surveys, X-ray diffraction (XRD) tests, Scanning Electron Microscope (SEM) analysis, energy-dispersive X-ray spectroscopy (EDS) tests, grain sieve tests, and dilute hydrochloric acid (HCl) pretreatment tests. XRD results revealed widespread carbonate cementation in the study area, primarily composed of calcite and dolomite (5.09–30.26

Research on cementation in natural moraine soils remains limited. To address this gap, a series of tests, including field surveys, X-ray diffraction (XRD), energy-dispersive X-ray spectroscopy (EDS), grain sieve analysis, and a pretreatment using hydrochloric acid (HCl) and hydrogen peroxide (H₂O₂), were conducted. The results reveal that clay-carbonate cementation is widespread in moraine soils, which are primarily composed of clay minerals and calcite (CaCO₃). XRD analysis revealed a clay mineral content ranging from 6

The Qinghai–Tibet Plateau has some of the most complex geological structures, the most intense tectonic activity, and the most severe geological disasters in the world. Under the control of multidynamic and cross-scale coupling effects such as plate interaction, plateau uplifting, climate change, and human engineering activities, the mechanism of major slope failure disasters is complex. There have been multiple occurrences of giant landslides and subsequent formation of giant barrier lakes in the Yigong Zangbo River Basin in the eastern Himalayan tectonic zone. After barrier-lake outburst, a basin-wide disaster chain occurred, seriously endangering the hydropower projects along the river and the safety of human life and property. This research explores the disaster environments, sliding structures, and failure mechanisms of the Benduo mountain slope in the Yigong Zangbo basin of the Qinghai–Tibet Plateau via field surveys, high-resolution unmanned aerial vehicle (UAV) surveys, adit explorations, and interferometric synthetic aperture radar (InSAR) monitoring. The results show that the Benduo mountain slope has a high-steep slope, good free-face conditions, a strong rock-mass unloading effect, and developed faults. Under the influence of faults, freeze–thaw cycles, earthquakes, and rock-mass unloading, multiple sets of discontinuous structural planes are developed in the shallow surfaces and deep rock mass. Owing to the structural planes, the rock mass exhibits various textures and clear zoning characteristics. Under these disaster environment conditions, the deformation and failure phenomena of the Benduo mountain slope are obvious. The slope is divided into three zones based on the shallow-surface deformation and failure phenomena. The most intense deformation areas are located at the rear edge and the leading edge of the Benduo mountain slope, and their maximum annual deformation rates are 85 and 80 mm/yr, respectively. Under the control of the structural planes, the deformation and failure modes of the Benduo mountain slope primarily include toppling, sliding–fracturing, and locking. The findings of this study provide a scientific basis for the deformation and failure mechanism of high-steep mountains and the risk prevention and control of basin disaster chains in the Qinghai–Tibet Plateau.

Debris flow is a common geological disaster in mountainous areas, characterized by its sudden onset, frequent occurrence, and high destructive power. Retaining dams are one of the most commonly used measures for debris flow prevention and are widely applied in debris flow management projects. This study investigates the impact resistance of retaining dams in high-altitude cold regions by establishing a three-dimensional numerical model of the retaining dam. The results show that the impact depth, resultant impact force, and acceleration of the prestressed reinforced concrete retaining dam with embedded prestressed reinforcement are significantly lower than those of the concrete retaining dam. The prestressed reinforced concrete retaining dam with embedded prestressed reinforcement can improve its impact resistance, effectively mitigating the impact of debris flow block collisions. The impact depth and resultant impact force of the prestressed reinforced concrete retaining dam both increase with the steel ball's impact speed, impact angle, and impact mass, while they decrease with an increase in the shape coefficient of the steel ball. The effects of different parameters of the steel ball on the impact depth and resultant impact force of the barrier vary. The research findings provide a scientific basis for the design of barriers in the prevention and control of debris flows in high-altitude cold regions.

The generation of post-fire debris flows has been shown to significantly differ from that of non-fire related debris flows, particularly in terms of erosion patterns and response to rainfall, necessitating further research on the complete hazard generation process. To explore this phenomenon, a post-fire debris flow event at Ren’e Yong gully in China was analyzed using simulations conducted with OpenLisem. This approach enabled the consideration of multiple factors within the simulation, including rainfall interception, soil infiltration, surface runoff, erosion, channel incision, bank slope erosion, and subsequent landslides. The results showed as follows: i)Overland flow initiated more rapidly and intensely in burned areas compared to unburned ones, and it also diminishes more quickly as rainfall decreases; ii)Surface erosion increases with the severity of the burn, leading to greater channel erosion in areas with larger burned extents; iii)The erosion phase of post-fire debris flow can be categorized into four stages: initial rainfall splattering, surface erosion and channel initiation, enhanced channel erosion during the debris flow process, and channel bank slides. This simulation successfully replicates the entire process of post-fire debris flow generation, demonstrating how increased surface runoff and erosion in burned areas contribute to the formation of debris flows.

Landslides pose a persistent and widespread hazard in mountainous regions, especially in areas with a history of ancient landslides that are prone to reactivation. The landslide at Paifang Pedestrian Street in Daguan County, Yunnan Province, China, exemplifies this risk due to its location atop an ancient landslide deposit, making it highly vulnerable during extreme weather events. This study employs the finite element method to simulate the coupled seepage-stress process, considering the effects of extreme rainfall, urban development, and geological conditions on slope stability. The findings indicate that increased rainfall intensity and duration significantly reduce the Factor of Safety (FOS). Under a peak rainfall of 140 mm/day, the FOS decreases from 1.24 to 1.14 after 10 days. Furthermore, top loading and bottom unloading exacerbate slope instability, with the FOS dropping to 1.02 within four days under a loading pressure of 10,000 Pa. After five days of rainfall, maximum unloading results in a displacement of 41.8 mm and a critical FOS of 1.03. The presence of ancient sliding surfaces further compromises stability, with zones containing multiple sliding bands being particularly susceptible to reactivation. This study emphasizes the importance of incorporating both environmental and anthropogenic factors in landslide risk assessments and advocates for more comprehensive governance strategies that consider their combined effects.

Spatial heterogeneity has a significant effect on the efficiency of geohazard susceptibility mapping. Taking Xide County in Sichuan Province as an example, a semi-quantitative zoning scheme based on the hazard-pregnant environment is proposed. The study area is divided into four zones based on the geohazard’s density and the hazard-pregnant environment’s spatial characteristics. The eco-detector and risk detector of the Geo-detector were designed to evaluate the effects of zoning. Accordingly, the evaluation factors in different zones were selected and weighed using the data fluctuation-weighted method, and the geohazard susceptibility of each zone was mapped using ArcGIS software. The AUC value, precision rate, recall rate, and coefficient of variation were used as evaluation metrics to examine the efficacy of the susceptibility mapping before and after zoning. The results demonstrate that the accuracy and dependability of susceptibility mapping can be significantly increased by resolving the problem of spatial heterogeneity. The coefficient of variation is not only an essential metric for evaluating susceptibility mapping, but it may also be essential for developing the evaluating factor system. The work promotes the refined development of geohazard susceptibility mapping and proposes a method that is greatly generalizable and reliable.

Acquiring the interior pore characteristics is essential to the utilisation of glacial till which is the material of permeability barrier to many engineering projects in cold regions. In this work, the pore space of glacial till samples with different cementation degrees and freeze-thaw cycles was investigated by CT technique. Representative elementary volume method (REV) was utilised to ensure a suitable scale of CT volume data, and a novel calibration method was proposed to detect the accuracy of CT segmentation. Subsequently, the effect of cyclic freeze-thaw action on the pore structure of glacial till samples was investigated. Quantitative pore structure analysis shows that cyclic freeze-thaw action resulted in a decrease in porosity of both well cemented and poorly cemented glacial till. The accessible porosity, fractal dimension, and coordination number of glacial till pores also decrease with the increment of freeze-thaw cycles. It can be observed that blocking of the channels of macropores is the main morphological evolution characteristic of glacial till with cyclic freezing and thawing, which can provide reference in designing permeability barrier in glacial till layers.

Large amounts of co-seismic landslides provide abundant debris sources following a strong earthquake and are prone to initiate and generate debris flow under heavy rainfall. Assessing the susceptibility they pose and what drives their movement in the years following the mainshock has not yet been attempted, primarily because multitemporal debris flow inventories are lacking. This study conducted statistical analyses using the multitemporal debris flow inventory (2008–2021) following the 2008 Wenchuan earthquake and a set of conditional factors (debris source, terrain, and hydrology). The dynamic susceptibility evaluation model was created using logistic regression. The temporal evolution of these factors affecting debris flow runout was explained. Our findings suggest that topography variables grew rapidly after the earthquake, while debris source factors dominated for seven years after the disaster before declining progressively to zero. These findings may have significant consequences for traditional susceptibility assessment models in regions where co-seismic landslides are the dominant debris flow source.

On August 20, 2019, a catastrophic debris flow occurred in the Cutou catchment, leading to the formation of a dam in the Min River. This debris barrier triggered a chain of multiple hazards, resulting in widespread flooding of houses upstream and downstream. To understand and assess the impact of this hazard chain, we conducted a comprehensive study using numerical modeling, field survey, the solid material reserve, and the probabilistic method. We began by conducting a detailed investigation of the 2019 event and calibrated the model parameters using survey data. Subsequently, we simulated and analyzed the entire process of hazard chain event, which was divided into four stages by the characteristics of water level. Finally, we developed a comprehensive and efficient approach for hazard assessment using the solid material reserve method and the probabilistic method. This approach enabled us to generate a hazard distribution map for the debris flow chain, providing valuable insights for effective hazard prevention in the Wenchuan area.

Moraine soils exposed to the moraine landscape surface due to glaciers retreating and global warming have been the subject of interest for researchers. The particle size distribution (PSD) of moraine soils is essential to understanding their engineering properties and soil characteristics. However, the wide aggregation phenomena observed in these soils introduce difficulties in particle size analysis. To overcome this, an experimental group test (EGT) and control groups test (CGT) were designed, which included a series of pretreatment methods to disperse particles, such as grain sieving test, laser diffraction test, microscopic observation, and scanning electron microscope. The tests were performed to record the aggregation phenomena and perform particle size test analysis. The results revealed widespread evidence of aggregation phenomena in each grain group (ranging from < 0.075 mm to 60 mm ~ 40 mm), encapsulated by three basic conceptual models (Core-coating, Core(s)-coat, and Coat-coring) for moraine soil. A new pretreatment method was developed, which involved the addition of 0.5% polycarboxylate in optimal ultrasonic dispersion to disperse aggregation phenomena before the sieving and laser diffraction tests. This method proved to be effective in removing the aggregation phenomena, which significantly impacted on the particle size distribution of moraine soils. The aggregation phenomena caused an overestimation of the content of the gravel group (about 7.91%) and an underestimation of the content of the silt-clay group (about 5.07%). After effectively removing the aggregation phenomena, many samples collected from poorly graded gravelly soils (GP) became silty sand soil (SM) and silty gravel soil (GM). Additionally, the average particle size was reduced by approximately 40%.

The Jiaju landslide is a large soil‒rock palaeolandslide in the Danba section of the Upper Dadu River in southwestern Sichuan Province, China. In this work, geological investigations, long-term comprehensive monitoring, experiments, and numerical simulations were used to determine the formation mechanism and evolution of this landslide. A complex armchair-shaped terrain with a substantial height difference between the rear and the foot of the slope, multiple structural defects within the landslide-accumulation region, and a coupling effect among multiple factors controlling river erosion and human activity were identified. The dynamic landslide deformation process was recorded by the integration of global positioning system (GPS) and inclinometer data, and the kinematic behaviour of the landslide was simulated using a Fast Lagrangian Analysis of Continua in 3 Dimensions (Flac3D) model in three stages. (1) During the last rapid uplift of the Tibetan Plateau and strong river erosion, the Jiaju bank slope was deformed and failed, forming a massive, thick palaeolandslide with a volume of approximately 2.76 × 107 m3. (2) The structure of the landslide materials deteriorated, and the weak structural plane was softened by groundwater and surface loading. The landslide was reactivated, with deformation and failure occurring through sliding and fracturing. (3) The landslide exhibited creep deformation with multistage and multilayer sliding surfaces via front-slope toe excavation and river-level fluctuations.