In alpine regions, avalanches endanger infrastructure such as roads, ski slopes and buildings. Some of the avalanche paths cannot be protected permanently due to financial and topographic limitations. Therefore, local experts assess daily whether additional safety measures are required, such as temporary road or ski slope closures. To support this decision-making process, we produced avalanche probability maps that show potential daily avalanche runout areas and intensities. The probability maps are generated by running multiple avalanche simulations using realistic distributions of input parameters, such as release volume and erosion depth, to capture a representative range of possible scenarios and runouts. Additionally, we account for release probability by incorporating the predicted avalanche danger scale into the analysis. We aim to identify the minimum number of input parameters needed to meaningfully represent daily conditions. To perform the numerical simulations, we can apply models with varying levels of physical details. To evaluate the quality of the produced probability maps, we recalculated well-documented avalanche events from Switzerland, using meteorological station data from the mornings prior to the avalanche occurrence. We compare the resulting predictions to the measured outlines of the avalanche cores. This study demonstrates how real-time data on weather and snow conditions can be utilized effectively to provide practitioners with a quick overview on how far current avalanches can reach considering the current conditions to support their decision-making process.

How to cite: Glaus, J., Kleinn, J., Stoffel, L., Ruttner-Jansen, P., Vicari, H., Gaume, J., and Bühler, Y.: From Data to Decisions: Enhancing Avalanche Mitigation with Probability Mapping, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-20987, https://doi.org/10.5194/egusphere-egu25-20987, 2025.

Accurate and sufficient use of climate data is essential for understanding the frequency of extreme weather events and avalanche cycles. However, current approaches to estimating avalanche release probability and fracture heights often fail to fully utilize available climate data, potentially leading to inaccurate results.

Existing methods for determining fracture heights typically rely on return values of three-day snow height increases, adjusted for wind load and slope. These methods, however, overlook the properties of the old snowpack and fail to quantify the probability of fractures occurring within it. This omission may lead to misleading fracture height estimates, especially in diverse climate conditions. Furthermore, current methods often rely purely on the height of the fracture disregarding the snow density. This introduces more often then not inconsistencies in the mass balance of observed release masses and those used in avalanche models.

To address these limitations, we developed AvaRelPro, a novel method for estimating avalanche release probabilities using gridded climate data. AvaRelPro combines daily climate variables, such as snow water equivalent and air temperature, to estimate shear strength and assess snowpack stability under various conditions. Monte Carlo simulations over 2.5 million synthetic days calculate release probabilities and corresponding fracture heights, accounting for both new and old snow layers. The model also incorporates adjustments for vegetation effects, wind drift, and snowpack stability.

AvaRelPro has been tested on several well-known avalanche paths in Norway, yielding promising results. This methodology enables the combined quantification of release probability and initial release mass for specific climates, slopes, and vegetation covers. Additionally, AvaRelPro is well-suited for studies exploring how changing climatic conditions impact avalanche release probabilities.

How to cite: Gisnås, K. G. and Gauer, P.: AvaRelPro: A Novel Method for Estimating Avalanche Release Probability and Mass, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15777, https://doi.org/10.5194/egusphere-egu25-15777, 2025.

Comprehensive and robust hazard and risk assessments require information beyond the current hazard map information. Current hazard mapping is usually limited to selected return periods and these return periods are commonly based on the return period of release and not the return period of impact.

We present a modelling approach, which allows to derive probability-based hazard information at any location in avalanche-prone areas. Such an event-based modelling approach provides continuous intensity-frequency-curves at any location and allows to calculate continuous loss-frequency-curves for individual objects as well as for groups of objects. It is therefore suited for comprehensive hazard and risk assessments.

The modelling approach is based on probabilistic seismic hazard analyses, which are common practice for seismic hazard and risk assessments since several decades. Continuous intensity-frequency information is required for hazard assessment and continuous loss-frequency curves are required for risk assessment. Therefore, we simulate numerous possible avalanches, covering the entire range of release volumes, from small to very large, and various possible model parameters. The probabilities of the different avalanche events are based on the exceedance probabilities of different snow accumulation heights, represented by the three-day snow accumulation, which is also the key information used for the generation of traditional hazard maps. The combined evaluation of avalanche intensities and probabilities in the area of impact allows to derive continuous intensity-frequency-curves at any given location. This approach contrasts with the common practice of assigning the return period of a single release volume to all possible outcomes of this release volume. In combination with exposed values and their vulnerability functions, this modelling approach allows to derive continuous loss-frequency-curves for comprehensive risk assessment and risk communication.

We present the simulation results from a case study in the Bernese Oberland in the Swiss Alps. The case study results highlight the advantages of this new modelling approach in hazard and risk management and the additional analyses being possible with this new modelling approach.

How to cite: Kleinn, J., Bühler, Y., Glaus, J., Aller, D., Peter, A., and Hählen, N.: A modelling approach for deriving continuous intensity-frequency and loss-frequency curves in the area of impact for robust hazard and risk assessment, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15365, https://doi.org/10.5194/egusphere-egu25-15365, 2025.

Winter nature-based tourism is a vital segment of the tourism industry, offering unique activities like ski touring, snowshoeing, winter hiking, or ice climbing. These activities not only provide recreational opportunities but also foster a deeper appreciation for winter landscapes and natural environments. Beyond its economic value, winter tourism promotes environmental awareness, nature conservation values, along with physical and mental well-being. However, challenges such as managing avalanche risks and ensuring tourist safety require understanding of both: 1) avalanche hazards and 2) visitors’ behaviour in order to introduce effective risk management strategies.

Therefore, our study aims to investigate determinants of avalanche risk and support preparedness planning, based on systematic long-term visitor and avalanche hazard monitoring in the Tatra National Park, Poland. Our work is based upon empirical data which comprise following datasets: 1) avalanche accidents records, 2) daily visitor counts, 3) daily meteo data, 4) daily avalanche danger scale in the winter seasons from 2018/19 to 2023/24. Avalanche accidents records were obtained from the mountain rescue database (TOPR) and included anonymized information on date, place, avalanche danger level, type of accident, number of affected people and the consequences of the accident. Visitor counts (grouped by recreation activity) were systematically collected at the entrance points to the Tatra National Park. Meteorological data contained measurements from 2 meteo stations located in the Tatra National Park: Kasprowy Wierch (1989 m a.s.l.) and Hala Gasienicowa (1508 m a.s.l.). Daily avalanche danger level was obtained from TOPR avalanche bulletin.

Our results show that visitation volumes ranged between 8’000 – 40’000 per winter season and have gradually increased in recent years. Dominant winter nature-based recreational activities were hiking (incl. snowshoeing), followed by ski touring. Climbing and speleology were the least frequent winter outdoor activities. Significant increase in nature-based winter activities was observed during COVID-19 pandemic (2021). During this time especially, ski touring visits increased by 400% in comparison to the 5-year average before the COVID-19 pandemic. Visits took place mainly on the weekends. Weather and avalanche risk level had less influence on visitation numbers.

To conclude, systematic long-term monitoring of avalanche risk determinants is necessary to develop successful risk management strategies. Combing knowledge on physical environmental conditions like snowpack, weather and human factors is critically important to address winter tourism risks and to improve safety outcomes in mountain destinations.

How to cite: Taczanowska, K., Bielanski, M., and Reiweger, I.: Combining long-term visitor and avalanche hazard monitoring to support risk management in winter mountain tourism: a case study of the Tatra Mountains, Poland, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8013, https://doi.org/10.5194/egusphere-egu25-8013, 2025.

Avalanche forecasters mostly rely on human observations of avalanche activity, but these reports are typically incomplete during periods of poor visibility, delayed, and not automated. Automated detection systems equipped with seismic sensors can improve monitoring efficiency, providing accurate avalanche data to support avalanche forecasting, regardless of visibility and weather conditions. Additionally, such systems could be implemented as early warning tools to enhance safety measures in mountain regions. While seismic detection systems have been widely used for avalanche monitoring, there is currently no automated method to reliably identify signals originating from avalanches. To address this, we developed a deep neural network to automatically detect avalanches in real-time continuous seismic recordings. This model was trained using seismic data collected over 13 winter seasons at the avalanche test site of Vallée de la Sionne, in Switzerland. Avalanches of varying sizes and paths are monitored using four seismometers strategically placed within and outside the avalanche path. The system simultaneously acquires continuous seismic data and event alarms. The alarms are based on amplitude thresholds recorded by the two seismometers near frequent release zones. While these alarms provide preliminary insights into avalanche activity, they require manual verification to filter out false alarms caused by events such as earthquakes or other unknown sources.

To overcome this limitation, we trained an end-to-end deep learning-based seismic waveform classifier on normalized, 40-second signal snippets extracted from the event database. The model architecture includes a convolutional encoder, a convolutional feature extractor with attention mechanisms, and a fully connected classification head. The network reliably distinguishes between avalanches and non-avalanche signals, achieving an accuracy of 0.97 on held-out events from the 2022/23 and 2023/24 winter seasons. The model was also deployed during the latest winter season to classify signals in near real-time, providing daily avalanche detection rates and demonstrating its feasibility as an automatic detection system. Finally, we plan to investigate the transferability of the classifier to seismic data collected by a distributed acoustic sensing (DAS) system installed on the avalanche test site, exploring its potential for broader applications.

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.: Development of a deep learning-based seismic waveform classifier to automatically detect snow avalanches., EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12940, https://doi.org/10.5194/egusphere-egu25-12940, 2025.

Snow avalanches are one of the most impactful natural hazards in mountainous areas. Avalanche characteristics are likely to change in a changing climate, especially in the Arctic where changes are more rapid, posing a severe challenge for local adaptation. Here we train machine-learning (ML) models to predict avalanche danger in northern Norway and then apply these models to dynamical downscalings of future climate projections.

We utilise regional expert avalanche-danger level assessments differentiating two different avalanche problems: wind slab, and wet (loose and slab combined). The ML models are trained on the 3-km Norwegian reanalysis (NORA3) to estimate the linkage between avalanche danger in the Troms region of Norway and local meteorological conditions. For the future climate simulations, we employ the Nordic Convection Permitting Climate Projections (NorCP), providing a 3-km dynamical downscaling of the Representative Concentration Pathway (RCP) scenarios performed with two global climate models.

To obtain a rough estimate of the trend of avalanche danger, the European Avalanche Danger Services (EAWS) 5-level avalanche danger scale is changed into a binary setup with levels 1 and 2 aggregated to 0 and levels 3, 4, and 5 to 1. The overall accuracy of the ML model for the wind slab problem is about 80 % and considerably higher than for the wet problem with about 67 %. This indicates that while the wind slab problem is to a high degree determined by the recent weather, this is less so for the wet problem. Including information from sophisticated snowpack modelling in the training data may thus increase the prediction accuracy.

By applying the ML models in a hindcast setting to the whole NORA3 record (1970-2024), we find a correlation between avalanche danger and a well-known climate mode, namely the Arctic Oscillation (AO). Given recent advances in model skill in representing the AO, this has potential implications for the seasonal predictability of avalanche danger in northern Norway.

Moreover, by applying the ML model to the NorCP simulations (2040-2060 and 2080-2100), the results differ per avalanche problem: While the wind slab avalanche danger declines in all scenarios, the wet avalanche danger remains relatively constant and even increases in some cases. The former appears to be related to decreasing snowfall and wind speed, while the latter is likely connected to increasing temperatures and rain.

How to cite: Eiselt, K.-U. and Graversen, R. G.: Predicting past and future avalanche danger in northern Norway with machine-learning models, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-3712, https://doi.org/10.5194/egusphere-egu25-3712, 2025.

How to cite: Aichele, J., Simeon, A., van herwijnen, A., Volpi, M., Sovilla, B., Huguenin, P., Gaume, J., Fichtner, A., Pérez, C., and Edme, P.: In Situ Snow Avalanche Monitoring and Characterization, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7061, https://doi.org/10.5194/egusphere-egu25-7061, 2025.

The mechanics of snow plays a critical role in assessing and mitigating hazards associated with snow and avalanches in mountain and cold environments. Accurately modelling the mechanical behaviour of snow remains a significant challenge, requiring the development of reliable constitutive models and advanced numerical methods. These efforts are particularly important because of the unique characteristics of snow, such as strain-rate sensitivity, sintering and degradation phenomena, localisation of the deformation, etc.

This study employs the elasto-visco-plastic constitutive model recently proposed by the authors [1, 2] to simulate compaction band formation in snow samples subjected to confined compression conditions. We refer to the experimental findings described by Barraclough et al. (2017) [3], which demonstrated the onset of compaction bands in snow specimens subjected to compression in plane strain conditions. The results revealed a clear dependency of the response on the applied strain rate: at low strain rates (typically below the threshold of 10^-4 s^-1) the samples exhibited homogeneous deformation. Conversely, when the strain rate exceeded this threshold, deformation localised, forming distinct compaction bands. The new constitutive model, implemented within the finite element software Abaqus/Standard through a custom User MATerial (UMAT) subroutine, has been used to simulate the behaviour observed in the tests.

The application of the proposed constitutive model to this problem demonstrates its ability to accurately replicate the specific phenomenon observed in laboratory experiments. This study underscores the potential of numerical simulations to enhance the understanding of snow deformation mechanisms, with promising implications for simulating the behaviour of snow at both in-situ and laboratory scales, as well as for improving avalanche hazard and risk assessment.

[1] Vallero, G., Barbero, M., Barpi, F., Borri-Brunetto, M., De Biagi, V. (2025). An elasto-visco-plastic constitutive model for snow: Theory and finite element implementation. Computer Methods in Applied Mechanics and Engineering, 433, 117465.

[2] Vallero, G. (2024). A visco-plastic constitutive model for snow. Theoretical basis and numerical implementation. PhD thesis. Politecnico di Torino.

[3] Barraclough, T. W., Blackford, J. R., Liebenstein, S., Sandfeld, S., Stratford, T. J., Weinländer, G., Zaiser, M. (2017). Propagating compaction bands in confined compression of snow. Nature Physics, 13(3), 272-275.

How to cite: Vallero, G., Barbero, M., Barpi, F., Borri-Brunetto, M., and De Biagi, V.: Modelling compaction band formation in snow: a FEM application of an elasto-visco-plastic model, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8180, https://doi.org/10.5194/egusphere-egu25-8180, 2025.

How to cite: Huitorel, C., Vicari, H., Di Pietro, T., Sovilla, B., and Gaume, J.: Modeling the effect of entrainment and air pore pressure on the mobility of snow avalanches: new insights from DEM and CFD-DEM simulations, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-16897, https://doi.org/10.5194/egusphere-egu25-16897, 2025.