Dr. Betty Sovilla

Powder snow avalanches (PSAs) radiate infrasound energy, yet the source mechanism remains unclear, limiting hazard monitoring and mitigation with infrasound-based technologies. Here, we analyze a unique data set from a large PSA to improve the understanding of the source mechanism. Through comparison of cluster activity within the airborne layers of the PSA and the recorded infrasound signal in the frequency domain, we demonstrate that infrasound is mainly generated from particle clusters suspended by turbulent eddies or ejected from the denser basal layer. Further correlating infrasound amplitudes with radar-derived spatial distributions of these clusters, we reveal a distributed source extending hundreds of meters behind the avalanche front. Additionally, we establish a relationship between infrasound and kinetic energy of suspended particles. These findings deepen our understanding of the complex dynamics of infrasound generation, offering valuable insights for avalanche detection and early warning strategies, and fundamental comprehension of PSA dynamics.

Powder snow avalanches (PSAs) pose significant threats to human lives and infrastructure in mountain regions. Measuring inside these natural flows is challenging because of their destructive power, which limits our understanding of these events. Previous research revealed a complex layering, characterized by a superposition of three distinct layers. The dense basal layer, where particle-particle interactions are  ominant, is situated below two aerial layers: the suspension and transition layer, both of which are particle laden turbulent flows. In contrast to the suspension layer, the transition layer hosts a higher amount of suspended mass and intense clustering processes, resulting in high local density fluctuations. Up to now technological limitations have prevented measurements within the dilute flows, especially the inability to capture turbulence at large scales. This lack of direct observations of dilute layers complicates model development and inhibits accurate predictions of PSAs’ destructive impacts.
To address this gap, we developed a non-invasive measurement technique consisting of a high-speed camera array. We installed this array on a vertical structure at the Vallée de la Sionne test site in Switzerland,  robing the transition and suspension layers of fully developed PSAs. The new array consists of three high-speed cameras and captures images at mm resolution from 5 m up to 11 m above ground.
On December 2nd, 2023, we measured the first PSA using the new setup. The collected images show individual suspended snow particles inside the transition and suspension layers, revealing complex air-particle dynamics. With image processing algorithms, including particle image velocimetry, we extracted essential flow characteristics such as the vertical velocity profiles.
This research paves the way for understanding the destructive potential of PSAs, but also lays the basis for improving prediction models and mitigation strategies. Additionally, it offers intriguing images from the inside of a PSA.

In mountainous regions, snow avalanches pose significant threats to both populations and infrastructure. As the flow progresses, it can accumulate additional mass by entraining bed material along its path, thereby amplifying its destructive potential. Volume increases of over tenfold the initial volume have been documented from avalanche observation and measurements. Entrainment is usually accounted for in depth-averaged models through a source term in the mass conservation equation and simplified entrainment criteria. However, quantitative assessment of erosion/entrainment rates with various flow types and bed material properties has seldomly been explored under well-controlled conditions. In this work, we investigate avalanche dynamics as well as erosion/entrainment mechanisms on the basis of depth-resolved particle-based simulations. Through such a modeling approach, processes like frontal ploughing, basal abrasion as well as erosion-deposition waves naturally emerge. Our findings highlight that, the interaction of the flow with the initially static bed material may have multiple and diverse effects on the flow behavior. First, the bed material can be eroded, but only partially entrained in the flow. Second, while the flow mobility is generally reduced by the presence of the bed, it may be enhanced in some cases due to the mechanical weakening of the bed. Finally, we propose a relationship linking the entrainment rate to a dimensionless parameter that compares the flow-induced stress to the bed shear resistance. This work contributes to a better understanding of the dynamics of snow avalanches and entrainment process and can lead to a refinement of depth-averaged formulations for practical purposes. Future research should consider factors such as rate-dependent snow rheology, air pore pressure and bed fluidization mechanisms to further refine our understanding and predictive capabilities in avalanche modeling.

In densely populated mountain regions, snow avalanches are among the main natural hazards and have long been a threat to settlements and critical infrastructure. Yet over the past decades, avalanche fatalities on roads and in settlements have been reduced through extensive avalanche hazard assessment and mitigation measures. In the protection scheme, fast dry-snow avalanches have long been considered the most destructive and thus the determining avalanches. In this case, impact pressure depends primarily on flow velocity, which is commonly obtained from depth-averaged avalanche simulations in engineering practice. With a warming climate, however, slower wet-snow avalanches are becoming more frequent. The impact pressure of these avalanches depends more on flow height and rheological properties of snow, such as friction, cohesion, and compressibility. In the midst of these changing conditions, current impact pressure calculation procedures involving numerical models based on depth-averaged mass and momentum balance equations struggle to capture the complex rheological behavior of snow and three-dimensional avalanche-obstacle interaction effects. We argue that a paradigm shift is needed if we are to meet the challenges posed by wet-snow avalanches. To that end, we propose a highly efficient 3D model based on the Material Point Method (MPM). Through its continuum formulation, it allows for the incorporation of established elastoplastic constitutive models and naturally handles collisions and fractures in a large deformation framework without mesh distortion problems. We demonstrate the relevance and the potential of this modeling approach by simulating key avalanche processes observed at the Vall ´ee de la Sionne avalanche test site. Furthermore, we quantitatively compare calculated impact pressure magnitude and distribution to measurements to show the accuracy of our approach and address the consequences of common modeling assumptions in avalanche engineering, such as neglecting snow compressibility.
This work aims to highlight the complex rheological composition of avalanche impact pressure and demonstrates the importance of a more fundamental and general impact pressure assessment in light of climate change.

With advancements in computational power and recent developments in particle-based numerical modeling, it is now possible to perform fully three-dimensional (3D) and real-scale snow avalanche simulations using methods such as the Material Point Method (MPM). These 3D MPM simulations explicitly model flow processes and velocity components in the slope-normal direction, allowing for accurate capture of phenomena such as flow detachment from the terrain in response to abrupt topography variations. In a recent case study where we applied MPM to simulate a snow avalanche at slope scale in 3D from release to run-out, we observed significant slope-normal velocity components, particularly in regions with pronounced variations in terrain curvature. This observation raised the question of the practical significance of slope-normal flow velocity components, especially in terms of their interaction with objects susceptible to uprooting, protection galleries, or other structures where slope-normal velocities might exert enough force to dislodge or damage roofs. To investigate this further, we use the flow conditions from a slope-scale MPM simulation of the 2019 “Salez” avalanche in Davos, Switzerland, which overflows a road gallery, as boundary conditions for Discrete Element Method (DEM) simulations. In these DEM simulations, we focused on the avalanche flow on the gallery roof, analyzing the resulting normal and shear forces as well as the velocity components normal and parallel to the roof. A comparison between the stresses obtained from DEM simulations and the calculations based on the Swiss Guidelines reveals that the guidelines provide conservative yet relatively accurate stress estimates. By distinguishing between the normal and parallel velocity components relative to the gallery roof, we find that the normal component is closely correlated with the resulting stresses, highlighting the potential significance of normal velocity components in determining the load on galleries.

Avalanche modeling is an essential tool to assess snow avalanche hazard. Today, most popular numerical approaches adopt depth-Averaged equations. These methods are computationally efficient but limited in cap turing processes occurring in the flow depth direction, e.g., erosion or deposi tion, which are often considered using ad hoc parameterizations or neglected completely. However, processes such as snow erosion, can crucially influence the flow dynamics and run-out and are often not negligible. We address these issues by using a new three-dimensional model, based on the Material Point Method (MPM) and finite strain elastoplasticity. To assess the possibilities and challenges associated with these highly detailed but computationally ex pensive calculations, we simulated the "Salezer" snow avalanche that released in Davos, Switzerland, in 2019. To reproduce the event in our simulations, we use the release areas mapped in a photogrammetric drone survey and es timate the snow conditions on the day of the event. We compare macroscopic features, such as flow outline and snow deposition, of the simulated avalanche to field observations. An in-depth analysis of transient 3D flow structures at the avalanche head demonstrates the degree of physical detail in the model, but also highlight challenges which still need to be addressed.

Waves are omnipresent in avalanches on Earth and other planets. The dynamic nature of waves makes them dangerous in geological hazards such as debris flows, turbidity currents, lava flows, and snow avalanches. Extensive research on granular waves has been carried out by using theoretical and numerical approaches with idealized assumptions. However, the mechanism of waves in realistic complex situations remains intangible, as it is notoriously difficult to capture complex granular waves on real terrain. Here, we leverage a recently developed hybrid Eulerian-Lagrangian numerical scheme and an elastoplastic constitutive model to investigate the processes involved in waves of snow avalanches, including erosion, deposition, and flow instability induced by terrain irregularity. This enables us to naturally simulate roll-waves, erosion-deposition waves, and their transitions in a single large-scale snow avalanche on real terrain. Simulated wave features show satisfactory consistency with field data obtained with different radar technologies. Based on a dimensionless analysis, the wave mechanics is not only controlled by the Froude number and local topography but also by the mass of the wave which governs the entrainment propensity. This study offers new insights into wave mechanisms of snow avalanches and provides a novel and promising pathway for exploring transient waves in granular mass movements.

Gravitational mass movements may erode and/or entrain a significant amount of bed material that can strongly affect the flow dynamics until the moving mass eventually deposits and comes to rest. Snow avalanches generally release on slopes covered by a metastable and thus potentially erodible snow cover that can have a wide range of strength – or cohesion – depending on the type of snow and its physical properties. As the avalanche flows, the snow cover is fully or partially entrained at the front and at the base of the flow, increasing the mass of the avalanche. Conversely, at the tail, snow may be deposited along the track, reducing the overall flowing mass. The balance between entrainment and deposition therefore determines the growing or decaying of the avalanche in terms of mass. To date, it remains unclear how cohesion influences these processes and what consequences it has for avalanche dynamics and run-out. Here, we perform simulations based on the Discrete Element Method (DEM) to analyze the influence of cohesion and slope angle on the erosion, entrainment, mixing and deposition processes. This method makes it possible to follow the dynamics of the particles within the flow very precisely, something that cannot be done in real experiments. In the model, the cohesion is represented as the combined effects of a fragmentation potential associated with the strength of the bonds, and an aggregation potential associated with the stickiness of the particles. For various combinations of input parameters and material properties, we release a heap of particles over an erodible bed and simulate the entrainment and deposition mechanisms. Our results show on the one hand that a low strength (< 3 kPa) promotes a ploughing entrainment mechanism and a large entrainment velocity, up to 3 m/s. On the other hand a high strength (> 3 kPa) favors basal abrasion as the flow front is not able to destabilize the erodible bed at once. In this case, the entrainment velocity decreases typically below 1 m/s. This has important consequences on mixing: the front in granular flows with low strength and adhesion is typically made of freshly entrained material coming from the whole depth of the bed, while remains of the released material can be found at the front of highly cohesive avalanches. Finally, the deposition process is analyzed by evaluating the relationship between the deposit thickness h_stop and slope angle θ which extends the framework of the model of h_stopθ from cohesionless to cohesive granular flows. We find that a higher bond strength of the flowing material increases the deposition height. Our work improves our understanding of the mechanics of cohesive granular flows and may contribute to improving parameterizations in depth-averaged models used to simulate geophysical mass flows such as rock, ice, snow avalanches, debris flows and landslides.

Snow avalanches are the deadliest natural hazards in Switzerland and cause significant economic losses annually. Therefore, avalanche detection and prediction are crucial in ensuring safety and mobility in the Swiss Alps. Operationally, avalanche forecasts are issued daily during the winter season to inform the public and local authorities about the avalanche hazard. Whereas traditional avalanche forecasting has been a human-expert decision-making process entirely, machine-learning models have become increasingly used in recent years. In this context, the Swiss avalanche warning service has been at the forefront of using these innovative data-driven tools to supplement current forecasting practices. These data-driven approaches have shown the potential to provide objective decision tools with higher spatiotemporal resolution than human-based forecasts. Nevertheless, the predictions generated by these machine learning models are opaque, usually referred to as 'black boxes', making it challenging for avalanche forecasters to interpret how variables are combined to make predictions. To address this, we employed an explanation method to interpret predictions from a machine learning model developed to predict danger levels for dry snow conditions in Switzerland. Moreover, precise avalanche data collected from an automated seismic detection system at the Vallée de la Sionne test site (Switzerland) are used to validate these danger level forecasts during the winter season 2020-2021. A novelty in future model implementations would be integrating these timely avalanche detections into forecasting models.

Powder snow avalanches (PSAs) with high-energy airborne layers pose a significant threat to buildings and facilities intended to be protected by infrastructure like dams. However, understanding the factors driving the development of these hazardous suspension layers remains challenging due to limited  xperimental data and understanding of PSA generation mechanisms. Recent advances in infrasound research have revealed that infrasound is primarily generated from particle clusters suspended in the airborne layer of PSAs by turbulent eddies or ejected from the denser basal layer. In addition, the infrasound signal is correlated with the kinetic energy of all active particles within these layers, offering a promising avenue to address this challenge. In this study, we analyze 27 infrasound signals collected at the Vall ´ee de la Sionne test site in Switzerland, covering various PSA dimensions and degrees of powder cloud development. We systematically quantify amplitudes, cumulative intensities, and energy distribution extracted from the infrasound  easurements. These findings are compared with reference data from a high-resolution GEODAR radar to assess the evolution of the powder cloud in both temporal and spatial dimensions. Subsequently, we correlate this information with boundary conditions such as snow cover, and previous avalanche activity to understand their influence on the development of high-energy airborne layers in PSAs. Conversely, we explore how these factors contribute to the decay of the powder cloud, thereby enhancing our ability to assess and mitigate avalanche hazard.

Skiing or snowboarding down a mountain covered with fluffy new snow can provide a unique experience,  eminiscent of floating above clouds. This sensation is often intuitively attributed to the supporting effect of the interstitial air loaded by the skis or snowboard, creating a feeling of weightlessness. Interestingly, this same principle may also account for the rapid descent and extensive runout distance of much less pleasant phenomena, such as powerful mixed snow avalanches. These avalanches exhibit a complex structure, characterized by a dense basal regime overlaid and preceded by an intermittent layer of coherent snow clusters, along with a lighter suspension layer of fine particles mixed with air. Despite their significance, the  processes driving the formation of these structures and the reasons for the remarkable mobility of mixed snow avalanches remain incompletely understood. One prevailing theory proposes that this mobility could result from the fluidization of the porous snow cover, which is transiently weakened by the development of excess pore air pressures upon undrained loading by the avalanche. We test this hypothesis, by means of two-phase simulations. The solid ice grains are simulated using either the Discrete Element Method (DEM) or as a continuum through the Material Point Method (MPM). Compared to previous works, we simulate explicitly the air both within the pores and in the ambient, using CFD (specifically the Finite Volume Method) which is coupled to the conservation equations of the solid phase. Simple preliminary simulations are presented in this work. We simulate a snowboarder dropping over a fresh snow cover. The simulations show that, if the permeability of the snow cover is low enough, the pressurized pore air may lead to fluidization of the snow, favoring its mechanical weakening and suspension as a vertical jet. Instead, if the permeability of the snow is too high, pore air pressure quickly diffuses, which cannot generate significant suspension of the particles. Similar modelling approaches will be applied in future to model mixed snow avalanches interacting with an erodible snow layer. Ultimately, we expect this research to contribute to improved entrainment and fluidization models in depth-averaged numerical methods.

The Material Point Method (MPM) is an Eulerian-Lagrangian numerical technique used to solve partial differential equations in continuum mechanics. It was initially introduced by Sulsky and co-authors in 1994 (Sulsky et al., 1994) as an extension of the Particle-In-Cell method, and has since then mainly been used and further developed in the geomechanics and graphics communities. MPM can deal with problems involving large deformations, collisions, fractures and can naturally handle freesurface flows, concepts which require specific treatments in mesh-based methods like the finite volume or finite element method. In recent years, MPM has received a growing attention in snow and avalanche science. On the one hand, it was used to animate snow in a mesmerizing manner in the Disney movie Frozen, and on the other hand it has been used to simulate avalanche release and flow at the slope scale, in 3D with unprecedented physical details. Since then, the model has been continuously improved and used to simulate and better understand various important processes related to snow and avalanche mechanics and we review here these applications and discuss practical implications. These processes include 1) snow microstructure deformation and failure; 2) avalanche release; 3) avalanche dynamics over complex topography; 4) interaction between avalanches and obstacles, forests or lakes; 5) erosion and entrainment. Although it has mostly been used in research so far, we believe that MPM has a strong potential for engineering applications, especially related to impact problems and the design of sustainable artificial or natural mitigation measures. Finally, because capturing specific physical processes such as snow entrainment require significant computational resources, we developed a novel approach based on a depth-averaged version of MPM (DAMPM) for efficient avalanche release and flow simulations.

Avalanches and other hazardous mass movements pose a danger to the population and critical infrastructure in alpine areas. Hence, understanding and continuously monitoring mass movements are crucial to mitigate their risk. We propose to use Distributed Acoustic Sensing (DAS) to measure strain rate along a fiber-optic cable to characterize ground deformation induced by avalanches. We recorded 12 snow avalanches of various dimensions at the Vallée de la Sionne test site in Switzerland, utilizing existing fiber-optic infrastructure and a DAS interrogation unit during the winter 2020/2021. By training a Bayesian Gaussian Mixture Model, we automatically characterize and classify avalanche-induced ground deformations using physical properties extracted from the frequency-wavenumber and frequency-velocity domain of the DAS recordings. The resulting model can estimate the probability of avalanches in the DAS data and is able to differentiate between the avalanche-generated seismic near-field, the seismo-acoustic far-field, and the mass movement propagating on top of the fiber. By analyzing the mass-movement propagation signals, we are able to identify group velocity packages within an avalanche that propagate faster than the phase velocity of the avalanche front, indicating complex internal structures. Importantly, we show that the seismo-acoustic far-field can be detected before the avalanche reaches the fiber-optic array, highlighting DAS as a potential research and early warning tool for hazardous mass movements.

In avalanche engineering and hazard mapping, computing impact pressures exerted by avalanches on rigid structures has long been a difficult task that requires combining empirical equations, rules of thumb, engineering judgment and experience. Until the 1990s, well-documented avalanches were seldom, and the main source of information included back-analysis of damage to structures and scarce field measurements. By the 1990s, several field sites were equipped across Europe, and since then they have provided new insights into the physics of impact. The main problem has been the difficulty in interpreting and generalizing the results to propose sound methods for estimating impact pressure. Testing a wide range of flow conditions has also been difficult in the field. To go a step forward in the elaboration of new guidelines for computing avalanche forces, we developed a numerical code based on the Discrete Element Method (DEM), which made it possible to simulate how an avalanche interacts with a rigid obstacle and to study how impact pressure depends on obstacle shape and size, as well as the avalanche flow regime. We extracted pressure and velocity data from the Vallée de la Sionne database to validate the DEM code, calibrate the model parameters, and elaborate avalanche scenarios. We studied four avalanches scenarios related to distinct flow regimes of the avalanche's dense core. In these scenarios, snow cohesion and velocity were imposed at the upstream boundary of the computational domain. Building on earlier work, we generalized an empirical equation for computing impact pressure as a function of snow cohesion, velocity, flow regime, and structure shape and size. Various coefficients were defined and calibrated from our DEM data. Within the range of tested values, we found good agreement between estimated pressure and field data.

Abstract: The calculation of the impact pressure on obstacles in granular flows is a fundamental issue of practical relevance, e.g. for snow avalanches impacting obstacles. Previous research shows that the load on the obstacle builds up, due to the formation of force chains originating from the obstacle and extending into the granular material. This leads to the formation of a mobilized domain, wherein the flow is influenced by the presence of the obstacle. To identify the link between the physical mobilized domain properties and the pressure exerted on obstacles, we simulate subcritical cohesionless and cohesive avalanches of soft particles past obstacles with circular, rectangular or triangular cross-section using the Discrete Element Method. Our results show that the impact pressure decreases non-linearly with increasing obstacle width, regardless of the obstacle’s cross-section. While the mobilized domain size is proportional to the obstacle width, the pressure decrease with increasing width originates from the jammed material inside the mobilized domain. We provide evidence that the compression inside the mobilized domain governs the pressure build-up for cohesionless subcritical granular flows. In the cohesive case, the stress transmission in the compressed mobilized domain is further enhanced, causing a pressure increase compared with the cohesionless case. Considering a kinetic and a gravitational contribution, we are able to calculate the impact pressure based on the properties of the mobilized domain. The equations used for the pressure calculation in this article may be useful in future predictive pressure calculations based on mobilized domain properties.

Erosion significantly affects the dynamics of gravity-driven mass flows. In snow avalanches, the snow cover can be substantially eroded but only partially entrained, however, there are very limited investigations to substantiate this difference. Here, we study various erosion and entrainment behaviors in snow avalanches using the material point method (MPM), finite strain elastoplasticity and critical state soil mechanics. With varied snow properties, distinct erosion patterns are obtained and analyzed with the mass change rate. When there is significant eroded and entrained mass, properties of released snow and erodible bed snow have clear correlations with the eroded mass, but not with the entrained mass, disclosing the difference in erosion and entrainment. Both enhanced and inhibited avalanche mobilities due to erosion and entrainment are captured under different conditions of snow properties and lengths of release and erodible zones. These new insights may stimulate the development of advanced erosion and entrainment models for large-scale avalanches.

With ongoing global warming, snow avalanche dynamics may change as snow cohesion and friction strongly depend on temperature. In the field, a diversity of avalanche flow regimes has been reported including fast, sheared flows and slow plugs. While the significant role of cohesion and friction has been recognized, it is unclear how these mechanical properties affect avalanche flow regimes. Here, we model granular avalanches on a periodic inclined plane, using the distinct element method to better understand and quantify how inter-particle cohesion and ground friction influences avalanche velocity profiles. The cohesion between particles is modeled through bonds that can subsequently break and form, thus representing fragmentation and aggregation potentials, respectively. The implemented model shows a good ability to reproduce the various flow regimes and transitions as observed in nature: for low cohesion, highly sheared and fast flows are obtained while slow plugs form above a critical cohesion value and for lower ground frictions. Simulated velocity profiles are successfully compared to experimental measurements from the real-scale test site of Vallée de la Sionne in Switzerland. Even though the model represents a strong simplification of the reality, it offers a solid basis for further investigation of relevant processes happening in snow avalanches, such as segregation, erosion and entrainment, with strong impacts on avalanche dynamics research, especially in a climate change context.

Pyroclastic surges are lethal hazards from volcanoes that exhibit enormous destructiveness through dynamic pressures of 10⁰–10² kPa inside flows capable of obliterating reinforced buildings. However, to date, there are no measurements inside these currents to quantify the dynamics of this important hazard process. Here we show, through large-scale experiments and the first field measurement of pressure inside pyroclastic surges, that dynamic pressure energy is mostly carried by large-scale coherent turbulent structures and gravity waves. These perpetuate as low-frequency high-pressure pulses downcurrent, form maxima in the flow energy spectra and drive a turbulent energy cascade. The pressure maxima exceed mean values, which are traditionally estimated for hazard assessments, manifold. The frequency of the most energetic coherent turbulent structures is bounded by a critical Strouhal number of ~0.3, allowing quantitative predictions. This explains the destructiveness of real-world flows through the development of c. 1–20 successive high-pressure pulses per minute. This discovery, which is also applicable to powder snow avalanches, necessitates a re-evaluation of hazard models that aim to forecast and mitigate volcanic hazard impacts globally.

Tall structures play an important role in the economic well-being of mountainous regions. They serve to distribute hydroelectric energy (power line transmission towers), generate tourist income (cableway pylons) and protect important infrastructure from natural hazards (trees). It is not uncommon for these structures to be destroyed by the air-blast of a powder avalanche. In this paper we investigate the shock response spectra of an asymmetric steel structure subjected to powder avalanche impact. We perform a spectral analysis of measured accelerations in the in-plane (avalanche flow) and cross-plane (normal to avalanche flow) directions. The structural eigenfrequencies in each direction are likewise calculated using a simple finite element model consisting of three-dimensional beam elements. Mode superposition is then used to reconstruct the measured accelerations, velocities and displacements in order to identify the magnitude and duration of the powder avalanche loading. From the analysis we find that the powder avalanche must contain high-frequency, short-duration blasts (f > 2 Hz) of moderate magnitude (p ≈ 1-10 kPa). High frequency blasts are also present in the cross-plane, transverse flow direction. This result, however, does not exclude the existence of longer duration blasts, which, because of the frequency mismatch, cannot excite the measurement pylon. It appears the high-frequency content of the cloud attenuates to endanger low-frequency structures (f ≈ 1 Hz) such as trees at the lateral edges of the flow, or hanging cables located well above the core. The shock response of a specific structure therefore depends on its location relative to the avalanche core, depending on how the blast-structures dissipate in time and space. Because the powder cloud contains a wide range of blast frequencies, tall structures are vulnerable to the action of the powder cloud, even if they are located well outside the avalanche core.

Snow avalanches cause fatalities and economic loss worldwide and are one of the most dangerous gravitational hazards in mountainous regions. Various flow behaviors have been reported in snow avalanches, making them challenging to be thoroughly understood and mitigated. Existing popular numerical approaches for modeling snow avalanches predominantly adopt depth-averaged models, which are computationally efficient but fail to capture important features along the flow depth direction such as densification and granulation. This study applies a three-dimensional (3D) material point method (MPM) to explore snow avalanches in different regimes on a complex real terrain. Flow features of the snow avalanches from release to deposition are comprehensively characterized for identification of the different regimes. In particular, brittle and ductile fractures are identified in the different modeled avalanches shortly after their release. During the flow, the analysis of local snow density variation reveals that snow granulation requires an appropriate combination of snow fracture and compaction. In contrast, cohesionless granular flows and plug flows are mainly governed by expansion and compaction hardening, respectively. Distinct textures of avalanche deposits are characterized, including a smooth surface, rough surfaces with snow granules, as well as a surface showing compacting shear planes often reported in wet snow avalanche deposits. Finally, the MPM modeling is verified with a real snow avalanche that occurred at Vallée de la Sionne, Switzerland. The MPM framework has been proven as a promising numerical tool for exploring complex behavior of a wide range of snow avalanches in different regimes to better understand avalanche dynamics. In the future, this framework can be extended to study other types of gravitational mass movements such as rock/glacier avalanches and debris flows with implementation of modified constitutive laws.