How to cite: Meloche, F., Simenhois, R., Sharaf, D., Greene, E., and Gaume, J.: Crack arrest in dry snow slab avalanches: Assessing the feasibility of man-made structures to stop crack propagation, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-22447, https://doi.org/10.5194/egusphere-egu26-22447, 2026.

Dry-snow slab avalanches are governed by two distinct but equally significant mechanical properties of weak snow layers: strength, which controls failure initiation, and fracture toughness, which governs whether an initial failure evolves into a self-sustaining crack. On inclined slopes, crack propagation occurs under mixed-mode loading, making avalanche release a combined multiaxial strength–fracture problem. A physically meaningful description of slab avalanche release requires both properties to be quantified together within the same weak layers.

In recent years, substantial progress has been made in characterizing these properties: multiaxial strength has been investigated extensively through laboratory experiments and numerical modeling, while mixed-mode fracture toughness has increasingly been measured directly in the field. However, these advances have largely remained disconnected. To date, strength and fracture toughness have not been measured simultaneously for the same weak snow layers, resulting in an incomplete mechanical description of the material's failure behavior.

Here, we therefore present recent experimental advances that enable direct field measurements of both failure and fracture properties of weak layers. We present a unique field-based dataset comprising multiaxial strength and mixed-mode fracture toughness measurements from two layers of buried surface hoar. Measuring both properties within the same layers allows us to construct both failure and fracture envelopes, providing a unified mechanical description of avalanche release. While both envelopes exhibit an elliptical shape, our results reveal contrasting behavior: weak layer strength is higher in compression than in shear, whereas fracture toughness is higher in shear than in compression. Relating these envelopes to weak layer microstructure, we qualitatively investigate how microstructural characteristics control failure initiation and crack propagation. Jointly constraining strength and fracture behavior with field data provides critical input for process-based avalanche release models and represents a significant step toward more physically consistent and reliable avalanche forecasting.

How to cite: Walet, M., Schöttner, J., Adam, V., Kraus, S., Rheinschmidt, F., Rosendahl, P., Weissgraeber, P., Schweizer, J., and van Herwijnen, A.: From failure to fracture: concurrent field measurements of weak layer strength and fracture toughness., EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8886, https://doi.org/10.5194/egusphere-egu26-8886, 2026.

Alpine mass movements pose a considerable risk to people and infrastructure. Snow avalanches pose a particularly prominent risk due to their widespread occurrence and potential catastrophic consequences. While significant advances have been made over the last decades to forecast avalanches, the spatio-temporal conditions that lead to avalanche release remain elusive. In fact, the problem of avalanche release is comparable to earthquake nucleation: An earthquake rupture and the fracture of a snow slab avalanche share the same underlying physics. As in the case of earthquakes, the observation of the rupture process of avalanches is limited to sensors in the farfield, such as seismic and infrasound arrays, and optical and radar methods, as well as laboratory experiments. While laboratory observations, far-field measurements on experimental test sites, and numerical simulations allow us to paint an ever more precise picture of the physics of avalanche nucleation, in situ measurements of crack propagation in the near-field of a real-world avalanche remain inaccessible so far.

How to perform such a measurement, which would not only allow us to understand the underlying physics better, but also might open new pathways to measuring precursory processes?

We designed a field experiment tackling the in situ observation of crack propagation and precursory processes. Leveraging a dense grid of seismic sensors we aim to capture the deformation in the nearfield prior, during, and after avalanche nucleation with Distributed Acoustic Sensing (DAS). In total, more than 3 km of fibreoptic cable were pulled from the top into the the steep slopes of Brämabuel near Davos (Switzerland) in autumn 2025. The cables were installed prior and during the first significant snowfall of the season, on known avalanche release slopes. Hence, they are placed centimeters below the expected weak layers, thus effectively making them an embedded strain sensor in a real-world experiment. To increase the probability of capturing the nucleation process, our DAS interrogator continuously samples at 2 m and 200 Hz, with the possibility to increase sampling rates to 1000 Hz for periods of increased avalanche risk. This continuous high spatio-temporal sampling will allow us to differentiate naturally and human-triggered slabs; in fact, skiers are easily identified in the data. In this talk, we will report on the first measurements of the 2025/2026 season and highlight the monitoring potential of our installation;

How to cite: Aichele, J., Edme, P., Andri, S., Alec, V. H., Betty, S., Pierre, H., Johan, G., Andreas, F., and Cristina, P.: Tracking avalanche nucleation in situ – A real-world experiment with embedded fibre optics, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17053, https://doi.org/10.5194/egusphere-egu26-17053, 2026.

Glide-snow avalanches occur when the entire snowpack slowly slides along the ground. Although liquid water at the snow–soil interface is known to play a role in glide-snowavalanche initiation, the mechanical behaviour at this interface remains poorly understood (Ancey & Bain, 2015) (Fees et al., 2024). These full-depth avalanches pose a significant risk to alpine infrastructure, as they are difficult to forecast and often involve large volumes.

This study investigates the influence of liquid water content, normal stress, and surface roughness on the shear strength of the snow–soil interface using controlled cold- laboratory experiments. Artificially produced snow is compacted and cut into cylindrical samples with a diameter of 8 cm, which are then sheared on two non-porous surfaces: a geotextile and a slate. Liquid water content at the interface is systematically increased through controlled heating of the basal surface, while shear force and displacement during the experiment are continuously measured, and interfacial liquid water content is quantified immediately after each test.

The experiments exhibited strain-softening behaviour under all conditions. Under dry conditions, peak shear strength increased with both idle time (the duration of surface contact before shearing) and applied normal load, while the Mohr–Coulomb friction angle remained constant for each surface. Increasing idle time resulted in a parallel upward shift of the yield surface toward higher shear strengths. Under wet conditions, the peak shear strength remains roughly stable with increasing interfacial liquid water content; shear behaviour was primarily governed by surface type and normal load.

Our findings indicate that, in addition to liquid water content, interface mechanics and surface properties play an important role in glide-snow avalanche release. The results provide new experimental insight into basal friction processes and contribute to an improved conceptual understanding of glide-snow avalanche initiation.

References

Ancey, C., & Bain, V. (2015). Dynamics of glide avalanches and snow gliding. Reviews of Geophysics.

Fees, A., Lombardo, M., van Herwijnen, A., & Schweizer, J. (2024). Glide-snow avalanches: insights from spatio-temporal soil and snow monitoring.

How to cite: Schnebli, R., Cabrera, M., Van Herwijnen, A., Gaume, J., and Bobillier, G.: Glide-snow avalanche initiation: The influence of liquid water on snow-surface friction, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20362, https://doi.org/10.5194/egusphere-egu26-20362, 2026.

How to cite: St Pierre Ostrander, T., Stitelmann, Ó., Wetter, J., Von Wartburg, J., Vincent, S., and Carrel, M.: Artificial intelligence supported extended range Doppler radar: avalanche activity measurement and mitigation verification in Evolène, Switzerland, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-3022, https://doi.org/10.5194/egusphere-egu26-3022, 2026.

Snow avalanches are among the deadliest natural hazards in mountainous regions. Yet avalanche activity is often still documented manually, and accurate avalanche release times are mostly missing. Automated monitoring systems equipped with seismic and infrasound sensors, combined with detection algorithms, could help record avalanche occurrences and provide accurate data on release time, size, and type. This comprehensive data on avalanche activity is indispensable for improving and validating avalanche forecasts and for implementing mitigation measures. At the Vallée de la Sionne (VDLS) test site in Switzerland, a combination of avalanche monitoring systems has been deployed for over two decades, including radars, cameras, seismic and infrasound stations. Additionally, avalanche researchers have manually documented and verified most avalanche events over the past 14 winter seasons to compile a unique avalanche catalogue.
To facilitate and automate avalanche detection, we aimed to implement two deep learning-based methods that scan continuous seismic and infrasound data separately in (near) real-time to detect and classify avalanche signals. Therefore, we leveraged the large volume of continuous data collected every winter at VDLS by adopting concepts from recent, powerful language models. Specifically, we pre-trained transformer networks in a self-supervised manner (i.e. without using expert labels) on a wide variety of signals mined from continuous seismic and infrasound data streams. The models receive fixed-length waveforms as input, partition them into sequential patches and compute patch-wise spectrograms. By training the models to reconstruct a portion of randomly masked patches, they learn to extract meaningful representations from the data, achieving silhouette scores of up to 0.6. This indicates good separability between avalanche and non-avalanche signals. Thus, these representations can later be used to automatically detect avalanches by fine-tuning a classifier on top. Moreover, combining predictions from the seismic and infrasound models has the potential to further improve (near) real-time avalanche detection.

How to cite: Simeon, A., van Herwijnen, A., Aichele, J., Volpi, M., Sovilla, B., Huguenin, P., Gaume, J., Fichtner, A., Edme, P., and Pérez-Guillén, C.: Self-supervised learning for automated avalanche detection from seismic and infrasound data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13917, https://doi.org/10.5194/egusphere-egu26-13917, 2026.

Snow avalanches in alpine environments are influenced by climate change. Assessing their long-term evolution remains difficult due to the scarcity of temporally consistent observation records. This lack of homogeneous data complicates the attribution of past changes to climate drivers and the development of credible future projections based on climate scenarios. To address these limitations, we develop a machine learning model that links avalanche observations with simulated meteorological and snowpack properties at a daily scale. A gradient-boosting regression model is trained to estimate daily avalanche counts, using weather and snow conditions derived from simulations. The analysis is conducted for a well-instrumented alpine catchment, the upper Haute-Maurienne valley in the French Alps, where approximately one hundred avalanche paths are monitored on a daily basis. We focus on avalanche events that reach predefined observation thresholds located at elevations of around 1800 m a.s.l. We show that particular attention to data quality and consistency has to be paid: accounting for uncertainties in avalanche release dates and restricting the training phase to a recent period with homogeneous observations are key prerequisites to obtain consistent results. Once trained, the model is first applied to reconstruct avalanche activity over the period 1958–2023 using reanalysed snow and meteorological data. It is then used to compute the evolutions of the avalanche activity between 1950 and 2100 using a downscaled ensemble of snow–climate simulations. Changes in avalanche activity are assessed using three complementary indicators corresponding to annual, monthly and weekly time scales. The reconstructed historical time-series indicates a marked decline in avalanche activity, with an average reduction of about 9 % per decade in the annual number of avalanches since 1958, mainly due to a decrease in the spring avalanche activity, while extreme avalanche cycles exhibit a more moderate decline. Future projections suggest a continued downward trend. Under the RCP4.5 and RCP8.5 scenarios, annual avalanche counts are projected to decrease by roughly 5 % and 9 % per decade, respectively, again largely driven by changes in spring conditions. Extreme avalanche activity is also expected to weaken, although at slower rates, with projected decreases in the 30-year return level of about 2 % per decade for RCP4.5 and 5 % per decade for RCP8.5. These climatic trends are associated with climate-induced changes in snowpack and meteorological variables through the use of machine-learning interpretation approaches. Overall, this study provides a quantitative assessment of climate-driven changes in avalanche activity for a representative alpine valley, combining machine-learning approaches with physically based snow-climate simulations.

How to cite: Doussot, F., Viallon-galinier, L., Eckert, N., and Hagenmuller, P.: Climate-driven changes in avalanche activity in the Haute-Maurienne valley (French Alps) over the period 1950–2100 based on machine-learning modelling, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17206, https://doi.org/10.5194/egusphere-egu26-17206, 2026.

The south-facing Arzler Alm avalanche path above Innsbruck, Tyrol, has a release area at approximately 2300 m a.s.l. on the Nordkette range and a runout zone reaching the immediate vicinity of an Innsbruck city district at about 650 m a.s.l. Following several very large avalanche events during the last century, extensive mitigation measures were constructed, including multiple breaking mounds and a catching dam, designed to withstand avalanche volumes of up to 1 million cubic meters. An exceptional snowstorm during the winter of 2018/19 triggered a large avalanche that caused substantial forest damage and highlighted the continued necessity of these protection structures.

During recent renovation works, a comprehensive measurement system was installed in the uppermost breaking mound. This structure is 5 m high, wedge-shaped, and approximately 10 m wide. Force measurements on a structure of this scale provide a rare opportunity to improve current engineering standards by describing avalanche-obstacle interactions. The measurement system consists of multiple 0.25 m² pressure plates installed on the mountain-facing sides of the breaking mound and record data at 10 kHz. These sensors capture both frontal impact forces and shear forces generated during flow deflection along the lateral flanks. To place local force measurements into the context of entire avalanche events and to quantify avalanche occurrence frequency, a state-of-the-art avalanche radar observes the full avalanche track continuously. The setup is complemented by video cameras from multiple viewing angles and allows future expansion with additional instrumentation, such as seismic sensors, fiber-optic sensing systems, and optical velocity measurements.

The test site has been operational since early 2026. We primarily present the technical design, site layout, and instrumentation and in case discuss data collected from avalanches during the current winter season. However, as the experiments rely entirely on natural avalanche activity and the statistical recurrence interval for avalanches reaching the mitigation structures is approximately one event per season, the availability of force measurements cannot be guaranteed for every year.

The Arzler Alm Avalanche Test Site is part of the Avalanche Laboratory Nordkette (Innsbruck, Austria), where recording of forces in snow fences, observations of glide snow activity and measurements with in-flow sensors in artificially released avalanches are performed. It complements other existing full-scale European avalanche research facilities such as Vallée de la Sionne (Valais, Switzerland) and Ryggfonn (western Norway). The year-round accessibility enables detailed manual field campaigns, including investigations of avalanche deposits, snow compaction in front of breaking mounds, and volumetric and entrainment processes. The south facing, large altitude range combined with significant new snow amounts during northwesterly snowstorms promotes avalanches that may transition from cold, dry flow regimes at release to warm, wet conditions in the deposition zone, making this new site particularly valuable for the large variety of avalanche release types, flow regimes and their evolution.

How to cite: Gleirscher, E., Köhler, A., Granig, M., Tollinger, C., Aust, J., Demmler, C., Walter, G., and Fischer, J.-T.: Avalanche Laboratory Nordkette/Innsbruck – On the technical design of the Arzler Alm avalanche test site, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-22941, https://doi.org/10.5194/egusphere-egu26-22941, 2026.

Mountain snow surfaces evolve through a combination of continuous processes, such as accumulation, melt, and redistribution, as well as abrupt, impulse-driven events, including snow avalanches. While C-band Synthetic Aperture Radar (SAR) time series have proven effective for avalanche mapping, most existing approaches rely on bi-temporal change detection or data-driven classification, treating avalanches as isolated events and neglecting the underlying temporal persistence and terrain-controlled spatial distribution of avalanche debris. In this study, we present a modified dynamical algorithm (m-DYN) for avalanche debris detection, which builds upon earlier applications of dynamical systems for mapping wet snow. The algorithm takes a SAR backscatter intensity ratio (BSR) image as an initial condition and iteratively evolves it toward a segmented image following a bistable dynamic.  This dynamics incorporates both a thresholding effect that tends to classify large BSR as avalanche debris and small BSR as background snow, and a physically consistent (topography-constrained) exchange of information between pixels at each iteration. Pixel coupling decreases with distance and is modulated by elevation differences, slope, and aspect, guiding the propagation of information along realistic avalanche flow paths and deposition zones. 

The methodology was evaluated over a study region of approximately 180 km² in the vicinity of Davos, Switzerland, within the Swiss Alps, an area of high relevance for avalanche research. Two periods of pronounced avalanche activity were analysed: 20–24 January 2018 and 13–16 January 2019. During both intervals, high-resolution SPOT-6 imagery was acquired by the WSL Institute for Snow and Avalanche Research (SLF), and precise avalanche boundaries were mapped and are publicly available. These independently mapped avalanche outlines, together with associated ground-truth coverage information, serve as validation datasets. Sentinel-1 Level-1 Ground Range Detected (GRD) imagery at 20 m spatial resolution and a nominal 6-day revisit interval was used for the SAR time-series analysis.

Initial results demonstrate that the m-DYN algorithm generates spatially coherent and terrain-consistent avalanche debris maps, effectively suppressing noise and seasonal variability while preserving fragmented and low-contrast deposits. Compared to traditional threshold-based approaches, which produce patchy detections, the dynamical segmentation approach substantially improves the reconstruction of continuous avalanche flow paths while maintaining robust precision values. The method converges reliably within a limited number of iterations and shows strong agreement with independent validation datasets across both study periods, underscoring the potential of topography-aware dynamical algorithms for avalanche mapping using a SAR time series.

How to cite: Kaushik, S., James, G., Karbou, F., and Mauss, A.: Dynamical segmentation of avalanche debris from a time-series of Sentinel-1 SAR images, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12382, https://doi.org/10.5194/egusphere-egu26-12382, 2026.

Current avalanche risk assessment relies heavily on manual snowpack observations and point-based meteorological data. Traditional forecasting requires personnel to physically enter high-risk terrain to perform stability tests and snow pit analyses, a process that is inherently dangerous, time-consuming, and limited by low spatial resolution. These manual "in-situ" measurements often fail to capture the complex stratigraphy and spatial variability of snow stability across entire slopes, leading to potential gaps in regional safety models.

To address these limitations, this research presents an integrated remote sensing framework utilizing Unmanned Aerial Systems (UAS) equipped with a multi-sensor payload. By synthesizing data from Ground Penetrating Radar (GPR), LiDAR, and multispectral cameras, we developed a non-invasive methodology for comprehensive snowpack characterization. The LiDAR sensors provide high-precision surface topography and snow depth measurements, while the GPR allows for the identification of internal stratigraphic boundaries and the estimation of snow density through electromagnetic wave propagation analysis. Concurrently, multispectral imaging assesses surface albedo and moisture content, offering insights into thermal degradation and surface hoar development.

The results of this integration are high-resolution 3D snow profile maps that allow for the quantitative assessment of snow hardness and density across broad, inaccessible slopes. By digitizing the snowpack structure at a granular level, this system provides forecasters with the data density required for accurate stability modeling without the necessity of human exposure to avalanche-prone zones. Ultimately, this UAS-based approach represents a paradigm shift in mountain safety, transitioning from discrete, high-risk manual sampling to continuous, remote, and data-driven hazard mitigation.

How to cite: Pistovčák, M.: High school students revolutionizing avalanche risk prediction using drones with mounted GPR, LiDAR and multispectral cameras, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2622, https://doi.org/10.5194/egusphere-egu26-2622, 2026.

How to cite: Spannring, P., Hesselbach, C., Huber, A., and Fischer, J.-T.: Towards a model chain for an automated thalweg identification, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10580, https://doi.org/10.5194/egusphere-egu26-10580, 2026.

Flow channels on the snow surface are a common phenomenon frequently observed and reported to avalanche warning services. An accurate interpretation of field observations and an understanding of the underlying physical processes are crucial for the assessment of snowpack stability. Flow channel formation is typically associated with rain-on-snow (ROS) events and is often interpreted as an indicator of the approximate elevation of the snow line, a key factor in forecasting wet-snow avalanche activity and the formation of crusts and crust-related weak layers. However, recent observations of flow channel development, without significant liquid precipitation, challenge the assumption that ROS events are the sole cause of their formation.

In this study, we quantitatively compare liquid water input into the snowpack from melt processes to the amount of rain during a documented flow channel formation event. Using a combination of field observations, energy balance calculations and model simulations, we demonstrate that, at least in our case study, meltwater was the predominant driver of flow channel formation. Our results indicate that more than 97% of the total liquid water input originated from melt, while rain contributed only roughly 2%. These findings highlight the need for a revised interpretation of flow channel formation, suggesting that meltwater-driven flow channels may be more significant than previously assumed.

How to cite: Hatvan, V., Gobiet, A., and Reiweger, I.: Drivers of Flow Channel Formation on Snow Surfaces: Rain versus Melt-Water – A Case Study in the Austrian Alps, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-21871, https://doi.org/10.5194/egusphere-egu26-21871, 2026.

Rapid mass movements in cold mountain environments are an important geomorphic factor that shapes the landscape, as well as a potential source of geohazard. Although their triggering mechanisms have been investigated in multiple studies, the spatial and temporal distribution of these events is poorly known in most Arctic regions. Among rapid mass movements that can be triggered abruptly by precipitation and snowmelt, we report a large-scale event that involved Central West Greenland in July 2023. Rain-on-snow linked to an atmospheric river event triggered more than 150 slushflows in Disko Island (Qeqertarsuaq). Pre-post event remote sensing imagery was used to map the affected areas, while environmental monitoring data and climate reanalysis products provided insights into the atmospheric river event and its impact on a snow-dominated landscape. During the 18-hour event, cumulative precipitation peaked at 115 mm, with more than 80 mm in several areas among the most affected by debris and snowpack mobilization. We show how the extreme precipitation rates reached during this event are similar to those experienced during a few other rapid mass movement events (i.e., wet-slush avalanches, slush flows, and debris flows) documented in various locations across West Greenland. The implications of such extreme rain-on-snow events in the Arctic are still poorly known. We suggest increasing remote-sensing based efforts to monitor and map rapid mass movements in critical areas of the Arctic, as most of them go unnoticed when infrastructure or large-scale damage are not involved.

How to cite: Securo, A., Colucci, R. R., Sigsgaard, C., Del Gobbo, C., Nielsen, S. R., Svennevig, K., and Citterio, M.: Rapid mass movements triggered by rain-on-snow events associated with atmospheric rivers in West Greenland, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6418, https://doi.org/10.5194/egusphere-egu26-6418, 2026.

In the Alps, numerous studies indicate that winter precipitation, measured in millimetres of water equivalent, does not exhibit any specific temporal trend. However, in recent years, the altitude of reliable snow cover has increased, whereas the snow depth on the ground, the amount of fresh snow, and the snow water equivalent (SWE) have decreased.

An investigation of the databases of Italian avalanche services affiliated with the Interregional Association for coordination and documentation of snow and avalanche problems (AINEVA), revealed an increase in rain on snow (ROS) events at all altitudes, even the highest above 2200 m. The analysis of data collected throughout the Italian Alps over a 20-year period (2006-2025) confirmed that spontaneous avalanche activity is greater on days with ROS.

Analysis conducted in the Dolomites, a range of the eastern Italian Alps, at six stations at different altitudes, further confirmed the increase in rainy days, especially in February and at medium and low altitudes. In the study area, the number of days with precipitation in the period 2016-2025 decreased by 25% compared to the previous decade, indicating more intense episodes, as total precipitation does not show significant variations. The number of snowy days decreased by 33%, further indicating a change in the snow-to-rain regime. Over the last 30 years, the snowfall limit increased by as much as 240 m in January and over 420 m in February.

How to cite: Bee, C. and Valt, M.: Rain on snow (ROS) events in the Italian Alps, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18300, https://doi.org/10.5194/egusphere-egu26-18300, 2026.