Dr. Alec van Herwijnen

For a slab avalanche to release, a crack in a weak snow layer beneath a cohesive snow slab has to initiate and propagate. Information on crack propagation is essential for assessing avalanche triggering potential. In the field, this information can be gathered with the propagation saw test (PST), a field test that provides valuable data on crack propagation propensity. The first PSTs were performed about 20 years ago and standards have since been established. However, there are still differences in how the PST is performed. Standards in North America require the column ends to be cut vertically, whereas in Europe they are typically cut normal to the slope. In this study, we investigate the effect of these different column geometries on the critical cut length. To this end, we conducted 27 pairs of PST experiments, each pair consisting of one PST with slope-normal cut ends and one PST with vertical-cut ends. Our experiments showed that PSTs with normal cut ends have up to 50 % shorter critical cut lengths, and the difference predominantly depends on the slope angle and slab thickness. We developed two load-based models to convert critical cut lengths between the test geometries: (i) a uniform slab model that treats the slab as one uniform layer and (ii) a layered model that accounts for stratification. For validation, we compare these models with a modern fracture mechanical model. For the rather uniform slabs of our experiments, both load-based models were in excellent agreement with measured data. For slabs with an artificial layering, the uniform load-model predictions reveal deviations from the fracture mechanical model, whereas the layered model was still in excellent agreement. This study reveals the influence that the geometry of field tests and the slope angle of the field site have on test results. It also shows that only accurately prepared field tests can be reliable and therefore meaningful. However, we provide models to correct for imprecise field test geometry effects on the critical cut length.

Glide-snow avalanches release at the soil-snow interface due to a loss friction which is suspected to be linked to interfacial water. Presently, the formation and distribution of the interfacial water are not well understood and glide-snow avalanches are considered unpredictable. We investigated the source, quantity, and spatial distribution of interfacial water before and during avalanche release through spatio-temporal field monitoring. The measurement setup consists of a sensor grid covering a slope with frequent glide-snow avalanche activity. The 24 grid sensors measured the soil temperature and liquid water content (LWC) throughout seasons 2021/22 to 2023/24. Snow and interfacial temperature and LWC were monitored locally with a vertical sensor profile ranging from the soil into the snow. Seven glide-snow avalanches released over the sensor grid and their investigation showed the following: (i) interfacial water originated from geothermal heat, rain, and meltwater percolation; (ii) the quantity of snow LWC was lower for glide-snow avalanches that released in early winter than in spring; (iii) soil temperatures in the release area were higher than in the remaining slope if interfacial water originated from geothermal heat; (iv) if interfacial water originated from rain and/or melt, we observed (locally) higher soil LWC in the release area; and (v) for the majority of observed avalanches the spatial variability in soil LWC across the slope reached a local minimum at the time of avalanche release. In the future, with continued monitoring, the spatio-temporal investigation of the soil LWC and temperature will help to quantify the drivers of glide-snow avalanche release at the slope scale. This will contribute to improved glide-snow avalanche forecasting and mitigation.

Capillary suction across the soil-snow interface is a possible mechanism for the formation of wet basal snow layers, which are necessary for snow gliding and glide-snow avalanches. However, little is known about the conditions under which this process occurs. We investigated capillary suction across the soil-snow interface considering realistic snow and soil properties. Snow properties were determined from snow profiles and soil properties were determined from field measurements of liquid water content, matric potential, soil texture, and bulk density for 40 alpine soils in Davos, Switzerland, as well as a field site in the region (Seewer Berg) with glide-snow avalanche activity. For the alpine soils investigated here, the results show that capillary flow from the soil to the snow is possible for realistic snow properties, but requires a soil saturation of approximately 90% or higher at the soil surface. When comparing the 90% saturation threshold to field measurements, the results suggest that capillary suction across the soil-snow interface is unlikely to contribute significantly to the formation of wet basal layers on Seewer Berg. These results are also relevant for soil and snow hydrology, where water transport across the soil-snow interface is important and understudied.

During winter, public avalanche forecasts provide crucial information for professional decision-makers as well as recreational backcountry users. While avalanche forecasting has traditionally relied exclusively on human expertise, avalanche warning services increasingly integrate data-driven models to support the forecasting process. This study assesses a random-forest classifier trained with weather data and physical snow-cover simulations as input for predicting dry-snow avalanche danger levels during the initial live testing in the winter season of 2020–2021 in Switzerland. The model achieved ∼ 70 % agreement with published danger levels, performing equally well in nowcast and forecast mode. Using model-predicted probabilities, continuous expected danger values were computed, showing a high correlation with the sub-levels as published in the Swiss forecast. The model effectively captured temporal dynamics and variations across different slope aspects and elevations but showed lower performance during periods with persistent weak layers in the snowpack. SHapley Additive exPlanations (SHAP) were employed to make the model's decision process more transparent, reducing its "black-box" nature. Beyond increasing the explainability of model predictions, the model encapsulates 20 years of forecasters' experience in aligning weather and snowpack conditions with danger levels. Therefore, the presented approach and visualization could also be employed as a training tool for new forecasters, highlighting relevant parameters and thresholds. In summary, machine-learning models like the danger-level model, often considered black-box models, can provide reliable, high-resolution, and comparably transparent “second opinions” that complement human forecasters' danger assessments.

When porous materials are subjected to compressive loads, localized failure chains, commonly termed anticracks, can occur and cause large-scale structural failure. Similar to tensile and shear cracks, the resistance to anticrack growth is governed by fracture toughness. Yet, nothing is known about the mixed-mode fracture toughness for highly porous materials subjected to shear and compression. We present fracture mechanical field experiments tailored for weak layers in a natural snowpack. Using a mechanical model for interpretation, we calculate the fracture toughness for anticrack growth for the full range of mode interactions, from pure shear to pure collapse. The measurements show that fracture toughness values are significantly larger in shear than in collapse, and suggest a power-law interaction between the anticrack propagation modes. Our results offer insights into the fracture characteristics of anticracks in highly porous materials and provide important benchmarks for computational modeling.

Assessing avalanche triggering potential requires detailed information on failure initiation and crack propagation. The propagation saw test (PST) is a field test that offers valuable data on crack propagation propensity, but its results, specifically the critical cut length, are complex to interpret due to their dependence on snowpack properties, test geometry, and slope angle. The slope angle significantly affects measured cut lengths, even when snow cover properties remain relatively constant. To address this, we developed a  echanical model that accounts for the influence of slope angle on the critical cut length. In this model, the unsupported section of the slab above the saw cut is treated as a cantilever beam exerting force on the weak layer, with variations in slope angle affecting the mass and orientation of the cantilever. The model incorporates easily measurable field quantities and shows good agreement with experimental data, significantly reducing slope-induced scatter. This model enables the transfer of PST data between flat and steep terrains with reduced uncertainty and provides a foundation for developing a PST-based index for crack propagation propensity in snowpacks.

A snow slab avalanche releases after failure initiation and crack propagation in a highly porous weak snow layer buried below a cohesive slab. While our knowledge of crack propagation during avalanche formation has greatly improved over the last decades, it still remains unclear how snow mechanical properties affect the dynamics of crack propagation. This is partly due to a lack of non-invasive measurement methods to investigate the micro-mechanical aspects of the process. Using a DEM model, we therefore analyzed the influence of snow cover properties on the dynamics of crack propagation in weak snowpack layers. By focusing on the steady-state crack speed, our results showed two distinct fracture process regimes that depend on slope angle, leading to very different crack propagation speeds. For long experiments on level terrain, weak layer fracture is mainly driven by compressive stresses. Steady-state crack speed mainly depends on slab and weak layer elastic moduli as well as weak layer strength. We suggest a semi-empirical model to predict crack speed, which can be up to 0.6 times the slab shear wave speed. For long experiments on steep slopes, a supershear regime appeared, where the crack propagation speed reached approximately 1.6 times the slab shear wave speed. A detailed micro-mechanical analysis of stresses revealed a fracture principally driven by shear. Overall, our findings provide new insight into the micro-mechanics of dynamic crack propagation in snow, and how these are linked to snow cover properties.

Glide-snow avalanches involve significant volumes of moving snow and present a major challenge to infrastructure in snow-covered mountain regions. Their initiation consists of a sequence of processes that remains poorly understood. It is conceptualized as starting with a loss of friction along the snow-ground interface due to the presence of liquid water. To shed more light on the effect of liquid water at the snow-ground interface on friction, we tested snow samples on a temperature-controlled substratum in a cold laboratory for different conditions of roughness and snow liquid water content (LWC). We performed 51 interface shear tests, analyzed the stress-displacement behavior, and studied the effect of interfacial water on the friction angles (static and kinetic). A sort of stick-slip behavior was observed on lubricated glass. On a dry and rough surface, the snow failed across the sample and no sliding occurred at the interface. Under dry conditions on a smooth surface or lubricated conditions on a rough surface, a strain-softening behavior with an interfacial strength was observed. For the tests where the frictional strength was exceeded, the friction angle, both static and kinetic, showed no dependence on the liquid water content. In the future, we plan to refine and extend the experimental setup to investigate snow compaction and analyze the effect of snowpack properties on gliding behavior.

In the rapidly evolving mountain cryosphere, snow avalanches threaten livelihoods, settlements and infrastructure. In this Review, we analyse past and projected impacts of climate change on avalanche activity and the associated risks. The limited availability of comprehensive datasets, the potential confounding factors and the limitations of statistical approaches can make it difficult to identify trends in avalanche activity. However, available data indicate a general decrease in avalanche number, size, seasonality and active paths at low elevations, and an increase in the proportion of wet avalanches relative to dry avalanches. Increased snowfall at high elevations can lead to peaks in avalanche activity and an increase in the number of wet and slush-like avalanches. Activity patterns gradually shift from low to high elevations under continued warming. These changes affect avalanche risk; however, risk is also influenced by factors such as land use and the growth or decline of human settlements. The impact of these factors varies across diverse mountain environments, making it challenging to predict how risk will evolve under a changing climate. Therefore, future research should aim to couple an improved systemic understanding of the impacts of these factors with slope-scale projections of avalanche hazards and risks to support sustainable mountain development and adaptation strategies.

In a recent review of the available scientific literature, impacts of climate change on avalanche activity and the associated risks were analyzed. The work done includes an evaluation of existing data and approaches, a synthesis of past and future changes in avalanche hazards and risks in various mountain environments worldwide with focus on their relation to climate change, elaboration of a conceptual model of trends in avalanche activity and related risks and the formulation of research priorities to address critical knowledge gaps. This paper synthetizes the main outcomes of this review.

Glide-snow avalanches release due to a loss of friction at the snow-ground interface, which can result in large avalanches that endanger infrastructure in alpine regions. The underlying processes are still relatively poorly understood, in part due to the limited data available on glide processes. Here, we introduce a pixel-based algorithm to detect glide cracks in time-lapse photographs under changing illumination and shadow conditions. The algorithm was applied to 14 years of time-lapse images at Dorfberg (Davos, Switzerland). We analysed 947 glide-snow events at a high-spatial (0.5 m) and temporal (2-15 min) resolution. Avalanche activity and glide-crack opening dynamics were investigated across timescales ranging from seasonally to hourly. Events were separated into surface (meltwater percolation) and interface events (no meltwater percolation). The results show that glide activity is highly variable between and within seasons. Most avalanches released without crack formation or within 24 h after crack opening, and release was favoured in the afternoon hours. Glide rates often showed a stick-slip behaviour. The acceleration of glide rates and non-constant increases in glide crack aspect ratio were indicators for avalanche release. This comprehensive dataset provides the basis for further investigations into glide-snow avalanche drivers.

Glide-snow avalanches release at the ground–snow interface due to a loss in basal friction. They pose a threat to infrastructure because of the combination of unreliable mitigation measures, limited forecasting capabilities, and a lack of understanding of the release process. The aim of this study was to investigate the influence of spatial variability in basal friction and snowpack properties on the avalanche release area distribution and the release location. We developed a pseudo-3D, mechanical, threshold-based model that consists of many interacting snow columns on a uniform slope. Parameterizations in the model are based on our current understanding of glide-snow avalanche release. The model can reproduce the power law glide-snow avalanche release area distribution as observed on Dorfberg (Davos, Switzerland). A sensitivity analysis of the input parameters showed that the avalanche release area distribution was mostly influenced by the homogeneity (correlation length and variance) of the basal friction and whether the basal friction was reduced suddenly or in small increments. Larger release areas were modeled for a sudden decrease and a more homogeneous basal friction. The spatial variability of the snowpack parameters had little influence on the release area distribution. Extending the model to a real-world slope showed that the modeled location of avalanche releases qualitatively matched the observed locations. The model can help narrow down the length scales and timescales for field investigations. Simultaneously, it can grow in complexity with our increasing understanding of glide-snow avalanche release processes. Input parameters such as the basal friction or snowpack parameters could potentially all be connected to the liquid water content. This would allow for the use of meteorological measurements to drive the model. The model has the potential to help identify potentially dangerous conditions for large or numerous avalanches which would help improve glide-snow avalanche forecasting.

Glide-snow avalanche release at the ground-snow interface and are currently considered unpredictable due to a lack of process understanding. It is generally accepted that the release of glide-snow avalanches is due to a loss of friction at the ground-snow interface, which is suspected to be linked to interfacial water. However, the source, quantity, and spatial distribution of the interfacial water are currently unknown. We installed a grid of sensors for spatio-temporal measurements within a slope with frequent glide-snow avalanche activity. The 24 grid sensors measured the temperature and liquid water content of the soil across the slope during seasons 2021/22 to 2023/24. During this time period, seven glide-snow avalanches released over the sensor grid. Investigation of three events showed that: (i) the interfacial water originated from geothermal heat, rain, and meltwater percolation, (ii) the liquid water content of the snow was lower for glide-snow avalanches that released in early winter than in spring, (iii) if the interfacial water originated from geothermal heat, we observed (locally) higher soil temperatures below the release area, (iv) if interfacial water originated from rain/melt, we observed (locally) higher soil LWC below the release area. In the future, with continued monitoring, the spatio-temporal investigation of the soil will help to quantify the drivers of glide-snow avalanche release which likely depend on the source of interfacial water.

Knowing snowpack stratigraphy is paramount for avalanche forecasting. In many forecasting operations the stratigraphy information comes from manual observations and traditional ramsonde  easurements. However, manual assessments are inherently subjective. Ram profiles can help to estimate the amount of erodible snow or identify large weak basal snow layers but lack the resolution necessary for  dentifying thin or soft layers. To overcome these challenges, digital cone penetrometers such as the SnowMicroPen (SMP) offer accurate stratigraphy data but are often limited by cost and primarily used for research purposes. In contrast, the snow SCOPE developed by PropagationLabs is cheaper, seems user-friendly, swiftly recording resistance profiles, and thus presents a promising alternative. The goal of this study is to objectively evaluate the SCOPE profiles with co-located SMP and standard measurements measured throughout the Alps. Our analysis, which is based on a dedicated matching algorithm, enables us to quantify the agreement with SMP measurements regarding force and vertical positioning.

Anthropogenic climate change is rapidly altering high mountain environments, including changing the frequency, dynamic behavior, location, and magnitude of alpine mass movements. Here, we review three decades of scientific literature (1995 to early 2024) to assess to what degree observational records from the European Alps – as the region with the most comprehensive records – reveal these changes. We do this for the processes that are most common in this region, namely rockfall, rock avalanches, debris flows, ice avalanches, and snow avalanches. The systematic literature search and review yielded 335 publications, of which we omitted publications that did not focus primarily on observational records. The remaining 103 publications used observations from over 100 sites and 30 inventories to investigate the connection between climate change and mass movements. About one third of the relevant studies found a measurable impact of climate change on the investigated alpine mass movement processes (with the exception of large rock avalanches). The clearest climate-controlled trends are (i) increased rockfall frequency in high-alpine areas due to higher temperatures, (ii) fewer and smaller snow avalanches due to scarcer snow conditions at low and mid elevations, and (iii) a shift towards avalanches with more wet snow and fewer powder clouds. While there is (iv) despite a clear increase in debris-flow triggering precipitation, debris-flow activity has not been found to uniformly increase, though there is some evidence for increasing activity above treeline and at locations without historical precedence. The trends for (v) ice avalanches are spatially very variable with no clear direction. Ice temperatures are measurably increasing, but – despite a theoretical expectation – this has not impacted ice avalanche activity to date. The reviewed literature also reveals that quantifying the impact of climate change on these mass movements remains difficult in part due to the complexities of the natural system, but also because of limitations in the available datasets, confounding effects, and existing statistical processing techniques. Better assessments could be achieved if we would more broadly support the compilation and maintenance of large standardized data catalogs, bring together various dispersed datasets (in said catalogs), including from social and citizen science projects, invest in long-term natural observatories, and develop suitable processing techniques. Better observations will additionally support the development and performance of process-based models. If we can advance natural hazard research on these fronts, more quantitative predictions of future change are well within our reach.

Liquid water at the ground–snow interface is thought to play a crucial role in the release of glide-snow avalanches, which can be massive and threaten infrastructure in alpine regions. Several mechanisms have been postulated to explain the formation of this interfacial water. However, these mechanisms remain poorly understood, in part because suitable measurement techniques are lacking. Here, we demonstrate the use of neutron radiography for imaging water transport in soil–snow systems. Columns of sand, gravel and snow were used to simulate the capillary forces of the soil–vegetation–snow layering found in nature. The columns were connected to a water reservoir to maintain a constant-pressure boundary condition and placed in a climatic chamber within the neutron beam. We show that neutron radiography is capable of measuring changes in the optical density distribution (related to liquid water content) within all three layers of the model system. Results suggest that a porous interface between the sand and snow may induce the formation of a water layer in the basal snowpack. Improved understanding of the water transport in soil–snow systems should lead to better prediction of glide-snow avalanche release and could also benefit other fields such as snow hydrology.

The cryosphere in mountain regions is rapidly transforming due to climate change, yet the impact of these changes on snow avalanche activity remains uncertain. Here, we use a snow cover model driven by downscaled climate projections to evaluate future alterations in dry- and wet-snow avalanche occurrences throughout the 21st century in the Swiss Alps. We assess avalanche activity by employing machine learning models trained with observed records of avalanches. Our findings indicate an overall decline in the occurrence of dry-snow avalanches during the months of December to May that is partially compensated for by an increase in wet-snow avalanche activity. Depending on elevation and the emission scenario considered, we anticipate a net reduction in the total avalanche activity ranging from under 10 % to as much as 60 % by the end of the century compared to 45–75 avalanche days per year at the beginning of the century. Projections further reveal a shift of wet-snow avalanche activity to earlier winter months. Analysis of changes in prominent snow grain types offers a coherent explanation of projected changes beyond a mere decrease in snow depth and snow cover duration. Overall, our study quantifies for the first time the significant influence of climate change on snow avalanche activity in the Swiss Alps and may serve as a benchmark for further mountain regions with similar avalanche climates.

The destructive power of avalanches demands fast detection and early warning systems to mitigate avalanche risk in mountain areas. Since avalanches are extended moving sources of seismic energy, array systems based on seismic sensors can be used to monitor avalanches over wide areas, regardless of weather and visibility. Operationally, one of the main challenges of these detection systems is the automatic discrimination of avalanche signals from other sources. Various machine learning techniques have, therefore, been used to automatically differentiate avalanche events within seismic recordings. So far, these models mainly relied on expert knowledge and a time-consuming selection process to extract signal attributes from the time and frequency domains. In this study, we therefore investigate the detection performance of models trained with expert-derived features and fully automatic extracted features using autoencoders. Autoencoders are neural networks that efficiently compress (encode) input data, such as seismic signals, down to its  essential features to reconstruct (decode) the original input. Hence, meaningful features can be automatically learnt during this process. To develop the models, we use a catalogue of labelled events recorded by a seismic array system deployed above Davos (Switzerland) since the winter season 2020-2021. We assess the performance of different classifiers developed with data from individual sensors by aggregating the predictions from all array components through ensemble predictions. Our results show that models trained with autoencoder-extracted features performed as well as those relying on expert features and spectrogram images. When tested on a continuous seismic dataset, the autoencoder model based on the frequency content of the signals achieved the best balance between avalanche detection accuracy and false alarm rate. Looking forward, we aim to integrate the best approach into an operational, near-real-time avalanche detection system.

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.

Fracture toughness, a critical material parameter for assessing resistance to crack growth, has been  xtensively studied in weak snow layers for mode I (slope normal) and mode II (shear direction). However, mode III (out-of-plane) fracture toughness remains unexplored but is crucial for describing cross-slope crack propagation. This study introduces an adaptation of the propagation saw test (PST) to investigate mode III fracture toughness by tilting the experimental setup sideways, allowing for the capture of out-of-plane  ontributions. This novel method, combined with a closed-form analytical model, successfully isolates and quantifies fracture modes, marking the first significant capture of mode III fracture toughness. Experimental validation supports the method’s efficacy, providing new insights into crack propagation characteristics in the cross-slope direction.
These findings enhance our understanding of slab avalanche release, offering a broader framework for modeling fracture properties in cross-slope directions.