ABSTRACT: Those traveling in the winter backcountry expose themselves to avalanche danger, where the interplay of triggering probability and potential consequences of being caught determines the individual risk. To assess the risk, the focus is on the sections particularly exposed to avalanche danger, so-called cruxes. At a crux, the probability of triggering as well as the possible consequences of an avalanche are estimated by assessing the stability of the snowpack (failure initiation and crack propagation) and the characteristics of the terrain. In general, there are different approaches, ranging from pre-tour planning to on-site decision-making, which are applicable for beginners as well as experts. However, navigating the multitude of information, observations and assessment tools is a challenge. This requires a comprehensive view and a structured yet flexible approach. To address this issue, we introduce the RiskCheck framework, which systematically focuses on key issues related to avalanche triggering probability and the consequences of an avalanche, culminating in a graphical risk analysis. This modular framework allows flexibility and continuous refinement to adapt to further information and observations. On the one hand, we propose an automated to semi-automated application of the RiskCheck that uses digital avalanche terrain maps and up-to-date information from the avalanche bulletin as a starting point, with subsequent refinement and adaptation by the user following a Bayesian approach. On the other hand, we show how simple rules of thumb and process-oriented approaches can facilitate risk assessment in the field using the same framework. The integration of both automated and manual procedures provides an approach to backcountry risk assessment that considers the inherent uncertainty associated with the problem. The proposed framework ensures consistency across different information sources and application scenarios which also allows a smooth transition from automated to manual application. In summary, the RiskCheck provides a versatile tool to assess avalanche risk by promoting risk-based decisions with a universal approach (probability  consequences) based on the current state of knowledge on avalanche release. ABSTRACT: For over a decade skitourenguru.com has been assigning daily avalanche risk scores to thousands of backcountry ski tours in the alpine region. These ratings are based on the “Quantitative Reduction Method” (QRM) and on “Screening the Likelihood of Avalanches on Backcountry Ski tours” (SLABS). Both methods combine information from the avalanche forecast and the terrain to a risk score. The standard routes made available on Skitourenguru are edited and reviewed by experts in Geographical Information Systems. End users have frequently wished to draw their own routes and have them rated afterwards. However, routes drawn by end users often lack the necessary quality. Therefore we developed a routing algorithm that not only adjusts existing routes but also generates new routes from a given start to a given end point. This algorithm is based on cost optimization (Dijkstra) through a cost surface. The cost surface is dependent on slope angle, curvature, forestation and other criteria such as roads, paths, rivers, bridges, lakes or information about ski routes and tracks. The optimization algorithm is implemented with the GRASS tool r.walk. For Switzerland, the feature has been available on skitourenguru.com/rating-view since December 2023. Users can either prospectively draw a route on a map or retrospectively upload a GPX file. Subsequently the route is adjusted by the algorithm and the avalanche risk score is calculated based on an avalanche forecast chosen by the user. The feature was immediately embraced by the backcountry skiing community in Switzerland. In the winter of 2023/2024, users rated more than 50.000 routes. Starting from December 2024, this feature will become available for the entire alpine region. Given its playful nature, we are confident that this feature can make a significant contribution to "avalanche education and learning." In this contribution we present the feature, give an insight to the routing algorithm, explain its purposes, discuss its limits and give an outlook to the future. 
  ABSTRACT: Assessing snowpack stability is crucial for determining the avalanche danger level. In this context, the propagation saw test is a valuable tool, though its results can be difficult to interpret. This paper aims to provide practical guidance for interpreting PST results by establishing direct connections between test outcomes and interpretable parameters. Despite nearly two decades of use, a definitive guide for PST interpretation is still lacking, with outcomes categorized as no propagation, crack arrest, or full propagation. We address the influence of factors such as slope angle and slab geometry on critical cut length and introduce an analytical model for real-time assessment. Utilizing data from 271 PSTs, we offer an interpretation tool to enhance comparability across different terrains and conditions.
  ABSTRACT: In recent years, the integration of physical snowpack models coupled with machine-learning techniques has become more prevalent in public avalanche forecasting. When combined with spatial interpolation methods, these approaches enable fully data- and model-driven predictions of snowpack stability or avalanche danger at any given location. This prompts the question: Are such highly detailed spatial model predictions sufficiently accurate for operational use? To explore this, we assess the performance of interpolated, model-based predictions of snowpack stability and avalanche danger, comparing them to human-generated public avalanche forecasts during the 2023/2024 winter season in Switzerland. To do so, we compare human forecasts and model predictions for locations in avalanche terrain (considering coordinates, aspect, elevation) where skiers triggered avalanches (244 events) or which were skied but where no avalanche was triggered (non-events, 3173 data points from GPX tracks). While this data reflects human behavior to some extent, we consider the event ratio as a proxy for the probability of avalanche release due to human load. We observed that with increasing model-predicted danger level or decreasing model-predicted snowpack stability, the event ratio increased. Comparing model predictions with human-made forecasts showed that the predictive performance of two operationally used models was similar to the performance of human avalanche forecasts: both predicted a strong increase in the probability of human-triggered avalanches. In summary, our results indicate that models capture regional patterns of snowpack (in)stability or avalanche danger well, and that these model chains should therefore be systematically integrated in the forecasting process.
  ABSTRACT: Forecasting avalanche hazard is the prediction of the magnitude, likelihood, timing, and location of potential avalanches. Assessing the likelihood of avalanches in absolutes is not currently possible due to a lack of scientific understanding of avalanche release processes, insufficient data describing snowpack structure, and the influence of weather; consequently, likelihood assessments represent subjective probabilities assigned by the forecaster. These subjective probability assessments are commonly communicated via verbal probability expressions on an ordinal scale: 1) Unlikely, 2) Possible, 3) Likely, 4) Very Likely, and 5) Almost Certain. Interpretations of verbal expressions of probability have been shown to vary drastically amongst individuals leading to communication problems, decreases in forecasting accuracy, and ultimately can compromise decision making. The verbal expressions of probability for avalanche forecasting have yet to be defined explicitly - both the meaning as a percentage chance of release and in rates of release expected for the spatial scale being forecasted - which has the potential to improve communication. In this paper, we propose a somewhat novel definition for likelihood of avalanches: Likelihood of Avalanches: is the chance of the start zones being assessed releasing within the forecast time period, regardless of avalanche size. Additionally, we propose a new ordinal scale of verbal expressions of probability complete with suggestions for percentage chance of release and frequencies of release. The definition: 1) can be effectively applied across all common spatial scales, which supports a core forecasting challenge, and 2) explicitly states a reference class (start zones being assessed) for the probability assessment which can improve understanding. The verbal expressions of probability are chosen given our existing understanding of what terms support operational avalanche risk mitigation decisions. The definitions of these terms in percentages are in line with common human interpretations and within the reasonable bounds of uncertainty when forecasting avalanches. Helpful frequency definitions are included because they naturally force the forecaster to explicitly explain the event and context (reference class). Finally, the amended scale is balanced and has more resolution at the lower end (avalanches are rare). The ultimate purpose of this paper is to promote discussion about forecasting the likelihood of avalanches and to elicit feedback from the community. Hence, we finish with an invitation to contribute via a survey. ABSTRACT: Snow and avalanche safety is a male dominated field. The aim of this paper is to increase the knowledge on the gendered conditions and prerequisites this poses for snow and avalanche safety professionals, and to shed light on why relatively few women enter and stay in the industry. Our analysis is based on qualitative survey data from a survey sent out to male and female-identifying avalanche professionals in North America, Continental Europe, and Scandinavia (N = 299). We inductively coded and categorized the responses into themes based on patterns and commonalities, using a conventional content analysis. Our preliminary results show that the majority of our participants think that increasing the share of women and non-binary individuals would be beneficial to the industry. However, our data also show that women experience important challenges related to their gender, both during their training and in their work. Based on our findings, we propose actions that we hope can help the avalanche safety industry becoming a world for anyone with a passion for snow.  ABSTRACT: Uncertainty is inherent in avalanche forecasting. On some days, the available evidence provides a clear picture of the existing hazard situation, while on other days, various unknowns can make it difficult to understand the conditions. Yet, uncertainty is currently not communicated in a consistent way in avalanche forecasts across the world, and to our knowledge, no research has explored how forecast users understand and respond to uncertainty information. To address this knowledge gap, we conducted a detailed online survey in collaboration with the Colorado Avalanche Information Center in the spring of 2024. The core of the survey consisted of an exercise where participants were presented with simplified but realistic avalanche forecasts that included different amounts of information about uncertainty from various sources. For each forecast scenario, participants were asked to assess the level of uncertainty in the forecast, and how the presented information would affect their approach to a typical backcountry trip. We also asked questions about the understandability, usefulness, and trustworthiness of forecasts that include uncertainty information. Our analysis of the responses from 1313 participants suggests that adding explicit statements about the magnitude and characteristics of the uncertainty in the forecast has a significant impact on readers’ estimated amount of uncertainty, which then influences their decisions about whether to enter the backcountry. Our results also show that including this information increases trust in avalanche forecast centers as reliable sources of hazard information. These insights support the inclusion of uncertainty information in avalanche forecasts. Our study contributes to the growing body of applied research that aims to help avalanche forecasting centers improve the effectiveness of their communication products by explicitly testing different formats and approaches. ABSTRACT: In this study, we investigate Scottish end users’ understanding and interpretation of the five point European Avalanche Danger Scale. Our main findings are, first, that many end users report to draw on detailed avalanche information including the avalanche problem information in their planning. Second, only seven in ten end users are aware that there are five danger levels. Third, end users’ risk perception of the danger scale, which we elicited using numerical probability judgements about each danger level, increases mostly linearly—results that cohere well with recent findings by Morgan et al. (2023) who used a different response format and a North American user group. Lastly, we report the results of an exploratory analysis whether individual characteristics of end users (such as their outdoor sport experience, age, gender, avalanche education), predict the individuals’ interpretation of the danger level. We find these characteristics have little explanatory power, which suggests that avalanche education needs to take a broad and inclusive approach to improve the intended understanding of the avalanche danger scale. We finish our discussion by contextualising the main findings.