Search

Snow avalanches are the primary mountain hazard for mechanized skiing operations. Helicopter and snowcat ski guides are tasked with finding safe terrain to provide guests with enjoyable skiing in a fast-paced and highly dynamic and complex decision environment. Based on years of experience, ski guides have established systematic decision-making practices that streamline the process and limit the potential negative influences of time pressure and emotional investment. While this expertise is shared within guiding teams through mentorship, the current lack of a quantitative description of the process prevents the development of decision aids that could strengthen the process. To address this knowledge gap, we collaborated with guides at Canadian Mountain Holidays (CMH) Galena Lodge to catalogue and analyze their decision-making process for the daily run list, where they code runs as green (open for guiding), red (closed), or black (not considered) before heading into the field. To capture the real-world decision-making process, we first built the structure of the decision-making process with input from guides, and then used a wide range of available relevant data indicative of run characteristics, current conditions, and prior run list decisions to create the features of the models. We employed three different modelling approaches to capture the run list decision-making process: Bayesian Network, Random Forest, and Extreme Gradient Boosting. The overall accuracies of the models are 84.6 %, 91.9 %, and 93.3 % respectively, compared to a testing dataset of roughly 20,000 observed run codes. The insights of our analysis provide a baseline for the development of effective decision support tools for backcountry avalanche risk management that can offer independent perspectives on operational terrain choices based on historic patterns or as a training tool for newer guides. [ABSTRACT FROM AUTHOR]

Avalanche warning services increasingly employ snow stratigraphy simulations to improve their current understanding of critical avalanche layers, a key ingredient of dry slab avalanche hazard. However, a lack of large-scale validation studies has limited the operational value of these simulations for regional avalanche forecasting. To address this knowledge gap, we present methods for meaningful comparisons between regional assessments of avalanche forecasters and distributed snowpack simulations. We applied these methods to operational data sets of 10 winter seasons and 3 forecast regions with different snow climate characteristics in western Canada to quantify the Canadian weather and snowpack model chain's ability to represent persistent critical avalanche layers. Using a recently developed statistical instability model as well as traditional process-based indices, we found that the overall probability of detecting a known critical layer can reach 75 % when accepting a probability of 40 % that any simulated layer is actually of operational concern in reality (i.e., precision) as well as a false alarm rate of 30 %. Peirce skill scores and F1 scores are capped at approximately 50 %. Faceted layers were captured well but also caused most false alarms (probability of detection up to 90 %, precision between 20 %–40 %, false alarm rate up to 30 %), whereas surface hoar layers, though less common, were mostly of operational concern when modeled (probability of detection up to 80 %, precision between 80 %–100 %, false alarm rate up to 5 %). Our results also show strong patterns related to forecast regions and elevation bands and reveal more subtle trends with conditional inference trees. Explorations into daily comparisons of layer characteristics generally indicate high variability between simulations and forecaster assessments with correlations rarely exceeding 50 %. We discuss in depth how the presented results can be interpreted in light of the validation data set, which inevitably contains human biases and inconsistencies. Overall, the simulations provide a valuable starting point for targeted field observations as well as a rich complementary information source that can help alert forecasters about the existence of critical layers and their instability. However, the existing model chain does not seem sufficiently reliable to generate assessments purely based on simulations. We conclude by presenting our vision of a real-time validation suite that can help forecasters develop a better understanding of the simulations' strengths and weaknesses by continuously comparing assessments and simulations. [ABSTRACT FROM AUTHOR]

The Avalanche Terrain Exposure Scale (ATES) is a system for classifying mountainous terrain based on the degree of exposure to avalanche hazard. The intent of ATES is to improve backcountry recreationist's ability to make informed risk management decisions by simplifying their terrain analysis. Access to ATES has been largely limited to manually generated maps in high-use areas due to the cost and time to generate ATES maps. Automated ATES (AutoATES) is a chain of geospatial models which provides a path towards developing ATES maps on large spatial scales for relatively minimal cost compared to manual maps. This research validates and localizes AutoATES using two ATES benchmark maps which are based on independent ATES maps from three field experts. We compare the performance of AutoATES in two study areas with unique snow climate and terrain characteristics: Connaught Creek in Glacier National Park, British Columbia, Canada, and Bow Summit in Banff National Park, Alberta, Canada. Our results show that AutoATES aligns with the ATES benchmark maps in 74.5 % of the Connaught Creek study area and 84.4 % of the Bow Summit study area. This is comparable to independently developed manual ATES maps which on average align with the ATES benchmark maps in 76.1 % of Connaught Creek and 84.8 % of Bow Summit. We also compare a variety of DEM types (lidar, stereo photogrammetry, Canadian National Topographic Database) and resolutions (5–26 m) in Connaught Creek to investigate how input data type affects AutoATES performance. Overall, we find that DEM resolution and type are not strong indicators of accuracy for AutoATES, with a map accuracy of 74.5 % ± 1 % for all DEMs. This research demonstrates the efficacy of AutoATES compared to expert manual ATES mapping methods and provides a platform for large-scale development of ATES maps to assist backcountry recreationists in making more informed avalanche risk management decisions. [ABSTRACT FROM AUTHOR]

The Avalanche Terrain Exposure Scale (ATES) is a system for classifying mountainous terrain based on the degree of exposure to avalanche hazard. The intent of ATES is to improve backcountry recreationist's ability to make informed risk management decisions by simplifying their terrain analysis. Access to ATES has been largely limited to manually generated maps in high use areas due to the cost and time to generate ATES maps. Automated ATES (AutoATES) is a chain of geospatial models which provides a path towards developing ATES maps on large spatial scales for relatively minimal cost compared to manual maps. This research validates and localizes AutoATES using two ATES benchmark maps which are based on independent ATES maps from three field experts. We compare the performance of AutoATES in two study areas with unique snow climate and terrain characteristics; Connaught Creek in Glacier National Park, British Columbia, Canada and Bow Summit in Banff National Park, Alberta, Canada. Our results show that AutoATES aligns with the ATES benchmark maps in 74.5 % of the Connaught Creek study area and 84.4 % of the Bow Summit study area. This is comparable to independently developed manual ATES maps which on average align with the ATES benchmark maps in 76.1 % of Connaught Creek and 84.8 % of Bow Summit. We also compare a variety of DEM types (LiDAR, stereo photogrammetry, Canadian National Topographic Database) and resolutions (5 m -- 26 m) in Connaught Creek to investigate how input data type affects AutoATES performance. Overall, we find that DEM resolution and type are not strong indicators of accuracy for AutoATES, with map accuracy of 74.5 % ± 1 % for all DEMs. This research demonstrates the efficacy of AutoATES compared to expert manual ATES mapping methods and provides a platform for large-scale development of ATES maps to assist backcountry recreationists in making more informed avalanche risk management decisions. [ABSTRACT FROM AUTHOR]

Danger ratings are used across many fields to convey the severity of a hazard. In snow avalanche risk management, danger ratings play a prominent role in public bulletins by concisely describing existing and expected conditions. While there is considerable research examining the accuracy and consistency of the production of avalanche danger ratings, far less research has focused on how backcountry recreationists interpret and apply the scale. We used 3195 responses to an online survey to provide insight into how recreationists perceive the North American Public Avalanche Danger Scale and how they use ratings to make trip planning decisions. Using a latent class mixed-effect model, our analysis shows that 65 % of our study participants perceive the avalanche danger scale to be linear, which is different from the scientific understanding of the scale, which indicates an exponential-like increase in severity between levels. Regardless of perception, most respondents report avoiding the backcountry at the two highest ratings. Using conditional inference trees, we show that participants who recreate fewer days per year and those who have lower levels of avalanche safety training tend to rely more heavily on the danger rating to make trip planning decisions. These results provide avalanche warning services with a better understanding of how recreationists interact with danger ratings and highlight how critical the ratings are for individuals who recreate less often and who have lower levels of training. We discuss opportunities for avalanche warning services to optimize the danger scale to meet the needs of these users who depend on the ratings the most. [ABSTRACT FROM AUTHOR]

Avalanche warning services increasingly employ large-scale snow stratigraphy simulations to improve their insight into the current state of the snowpack. These simulations contain information about thin, persistent critical avalanche layers that are buried within the snowpack and are fundamental drivers of avalanche hazard. However, the data volume, data complexity, and unknown validity have so far limited the value of the simulations for operational decisions. We attribute this at least partially to a lack of research that validates the simulations for their capability to represent the existence and instability of known critical layers at the regional scale. To address this knowledge gap, we present methods that enable meaningful comparisons between regional assessments of avalanche forecasters and snowpack simulations that are distributed across entire forecast regions. We applied these methods to operational data sets of ten winter seasons and three public forecast regions in western Canada and thereby quantified the performance of the Canadian weather and snowpack model chain to represent persistent critical avalanche layers. We found that the overall probability of detecting a known critical layer in the simulations can be as high as 75 % when accepting a low probability of 40 % that any simulated layer is actually of operational concern in reality. Furthermore, we explored patterns that characterize which layers were represented well and which were not. Faceted layers, for example, were captured well but also caused most false alarms, whereas surface hoar layers tended to be less prevalent but in return were mostly of operational concern when modeled. Overall, our results suggest that the simulations provide a valuable starting point for targeted field observations as well as a rich complementary information source that can help alert forecasters about the existence of specific critical layers or provide an independent perspective on their instability. However, we do not believe that the existing model chain is sufficiently reliable to generate assessments purely based on simulations. We conclude by presenting our vision of a real-time operational validation suite that can help forecasters develop a better understanding of the simulations' strengths and weaknesses by continuously comparing assessments and simulations in a user-friendly manner. [ABSTRACT FROM AUTHOR]