Snow avalanches affect transportation corridors and settlements worldwide. In many mountainous regions, robust records of avalanche frequency and magnitude are sparse or non-existent. However, dendrochronological methods can be used to fill this gap and infer historic avalanche patterns. In this study, we developed a tree-ring based avalanche chronology for large magnitude avalanche events using dendrochronological techniques for a sub-region of the northern United States Rocky Mountains. We used a strategic sampling design to examine avalanche activity through time and across nested spatial scales (i.e. from individual paths, four distinct sub-regions, and the region). We analysed 673 total samples from 647 suitable trees collected from 12 avalanche paths, from which 2,134 growth disturbances were identified over years 1636 to 2017 Common Era (C.E.). Using existing indexing approaches, we developed a regional avalanche activity index to discriminate avalanche events from noise in the tree-ring record. Large magnitude avalanches common across the region occurred in 30 individual years and exhibited a median return interval of approximately three years (mean = 5.21 years). The median large magnitude avalanche return interval (3–8 years) and the total number of avalanche years (12–18) vary throughout the four sub-regions, suggesting the important influence of local terrain and weather factors. We tested subsampling routines for regional representation, finding that sampling eight random paths out of a total of 12 avalanche paths in the region captures up to 83 % of the regional chronology, whereas four paths capture only 43 % to 73 %. The greatest value probability of detection for any given path in our dataset is 40 % suggesting that sampling a single path would capture no more than 40 % of the regional avalanche activity. Results emphasize the importance of sample size, scale, and spatial extent when attempting to derive a regional large magnitude avalanche event chronology from tree-ring records.
Travel in avalanche terrain requires considered and careful selection of appropriate terrain to reduce exposure to avalanche danger. In many parts of the world, recreational backcountry skiers in avalanche terrain are aided by a regional avalanche forecast. The overall aim of an avalanche forecast is for users to adjust their terrain choices in response to the avalanche danger rating and avalanche problem, thereby reducing their risk of an avalanche involvement. In this paper we present a novel passive observation technique to assess how lift assisted backcountry skiers adjust their terrain use in response to the avalanche danger rating.
This paper develops and demonstrates a method to record the terrain metrics of all skiers on an avalanche-prone backcountry slope. Using a remote time-lapse camera focused on a high skier-use backcountry slope, we anonymously recorded the descent route of skiers in ten-second increments. Using 31,966 images over 13 days and 7499 skier point locations, skier locations were digitized from the images, then transformed onto a geo-referenced digital elevation model (DEM) such that terrain metrics could be extracted for each anonymous skier location.
When these location points are compared to simultaneous GPS measurements, the horizontal accuracy was estimated to be within a 49-m horizontal accuracy, with a 95% confidence interval. Analysis of the terrain metrics for each skier point compared slope, profile curvature (downslope), and plan curvature (cross slope) over days with different forecasted avalanche danger ratings. This statistical analysis was qualitatively supported by a review of the spatial patterns of the terrain choices on these days. Furthermore, we used this technique to estimate group size, and found a surprising number of solo skiers, even on Considerable avalanche danger days. By remotely photographing all skiers on a slope, the data collected provides a large and diverse data set of the terrain preferences of backcountry skiers under varying avalanche conditions, with limited bias. These results have implications for avalanche education by enhancing our understanding of specific terrain management skills by backcountry skiers.
Our research seeks to expand existing knowledge of travel behavior and decision-making in avalanche terrain. We have done this by using GPS tracking to observe the travel behavior of out-of-bounds (OB) skiers and collecting survey data to investigate their terrain choices. We sampled participants in the field by distributing hand-held GPS units and surveys along the southern boundary of Bridger Bowl Ski Area, Southwest Montana, USA. In total, we used data from 136 participants that volunteered over the course of 19 field days, from February 2017 to February 2018. We analyzed the resulting GPS data using a GIS, and we derived terrain metrics from elevation and land cover data. We fit a multiple linear regression model using GPS track downhill starting distance from the ski area boundary as the response variable and survey responses, interaction with complex avalanche terrain (as defined using the Avalanche Terrain Exposure Scale), weather conditions, and avalanche hazard level as explanatory variables. This approach evaluates travel behavior as a function of human factors, terrain, weather, and snowpack, providing a holistic perspective on decision-making drivers. Our results show that gender (female), formal avalanche education, and perception of avalanche mitigation are statistically significant (p < 0.05) survey responses which indicate that participants travel further from the ski area boundary before descending Saddle Peak, which effects individuals avalanche terrain exposure. Downhill starting distance is also significantly correlated with time and distance in complex avalanche terrain (p < 0.05). Our results provide a case study of the terrain preferences and avalanche awareness of OB skiers and highlight specific “human factors” that are correlated with terrain selection. Two practical applications of this research are: 1) tailoring of targeted avalanche education outreach based on our results specific to the OB setting, and 2) designing new signage to illustrate the avalanche terrain near the ski area boundary for skiers who are inexperienced in the backcountry or unfamiliar with the specific area.
Most dry slab avalanches occur during or immediately following loading by snowfall or wind deposition. In the absence of further loading avalanche activity decreases over time. This suggests that loading favorably changes snow cover properties for avalanche release over short time scales (e.g., minutes and hours), and that changes over longer time scales (days or longer) help to stabilize the snowpack. This study quantifies both: 1) the effect of increasing load on the interaction of the slab and weak layer over short time scales, and 2) the longer term stabilizing changes following loading. We developed a field method to rapidly increase the load on existing weak layers, and conducted two different sets of experiments. For the first set of experiments we used a cardboard frame the dimensions of a standard Propagation Saw Test (PST) and added 5, 10, 15, or 20 cm of disaggregated snow on top of PST columns on 11 sampling days. We allowed the added snow to sinter for approximately 30 to 60 min before completely isolating the block and performing a PST. In the second set of experiments we used the same technique to add 10 cm of disaggregated snow on over 30 isolated columns. We then conducted PSTs in the minutes, hours and days following isolation, with tests ranging from 15 min to 4 days. For both experiments we
filmed each test at 120 fps for particle tracking velocimetry analysis. We also utilized a model simulating the experiments to better interpret our results. In the first set of experiments, critical crack lengths dramatically decreased with increasing load while crack propagation speed increased, a finding consistent with previous work. In the second set of experiments, we found that critical crack lengths increased rapidly at first and then more slowly over time. Simulations of the experiments suggest that changes in critical crack length over time are caused by an increase in slab elastic modulus in the first hours following loading, and then caused by both increasing slab elastic modulus and weak layer specific fracture energy in days following loading. Overall, our
results help to illustrate changes in critical crack lengths immediately after and in the days following loading. Our results are consistent with field observations of increasing avalanche activity during and immediately following loading events and decreasing avalanche activity afterwards.
In avalanche operations, ski cutting involves a single avalanche practitioner attempting to trigger a snow avalanche by skiing across the upper part of a slope. There are two types of ski cutting: test skiing to determine if the snow is unstable and mitigation to remove unstable snow before the avalanches get bigger or before less skilled people (e.g. clients) get to the specific slopes. To address the wide differences in the perceived risk of injury during ski cutting, we conducted a quantitative survey that helped avalanche practitioners estimate the number of ski cuts over many winters and asked them to recall their near misses and three classes of injuries. Over 150 practitioners completed the survey with a combined career total of 1.5 million ski cuts. From the responses, we calculated various results. The median number of ski cuts per respondent is 300 per winter. The rate of triggering a size D1 to 1.5, D2 to 2.5 and D3+ avalanche was 300, 4 and 0.1 per thousand ski cuts, respectively, indicating that smaller avalanches are triggered much more often than larger ones. The rate of being caught in a size D1 to 1.5, D2 to 2.5 and D3+ per thousand triggered avalanches was 7, 25 and 80, respectively, indicating that the probability of being caught increases with the size of an avalanche triggered during ski cutting. When the survey results are scaled to a million ski cuts, about 23 resulted in light duty, 7 resulted in missed work and 3 resulted in career ending injuries. Practitioners at lift-based ski areas (ski patrollers) had lower risk per ski cut than guides for helicopter and snowcat skiing.
Deep persistent slab avalanches are a natural hazard that are particularly difficult to predict. These avalanches are capable of destroying infrastructure in mountain settings, and are generally unsurvivable by humans. Deep persistent slab avalanches are characterized by a thick (> 1 m) slab of cohesive snow overlaying a weak layer in the snowpack, which can fail due to overburden stress of the slab itself or to external triggers such as falling cornices, explosives, or a human. While formation of such snowpack structure is controlled by persistent weather patterns early in the winter, a snowpack exhibiting characteristics capable of producing a deep persistent slab avalanche may exist for weeks or months before a specific weather event such as a heavy precipitation or rapid warming pushes the weak layer to its breaking point. Mountain weather patterns are highly variable down to the local scale (1-10 m), but they are largely driven by atmospheric processes on the continental scale (1000 km). This work relates atmospheric circulation to deep persistent slab events at Mammoth, CA; Bridger Bowl, MT; and Jackson, WY. We classify 5,899 daily 500 millibar geopotential height maps into 20 synoptic types using Self-Organizing Maps. At each location, we examine the frequency of occurrence of each of the 20 types during November through January during major deep persistent slab seasons and compare those frequencies to seasons without deep persistent slab avalanches. We also consider the 72-hour time period preceding deep persistent slab avalanches at each location and identify synoptic types occurring frequently, as well as those rarely occurring prior to onset of activity. At each location, we find specific synoptic types that tend to occur at a higher rate during major deep persistent slab years, while minor years are characterized by different circulation patterns. We also find a small number of synoptic types dominating the 72-hour period prior to onset of deep slab activity. With this improved understanding of the atmospheric processes preceding deep persistent slab avalanches, we provide avalanche practitioners with an additional tool to better anticipate a difficult to predict natural hazard.
Most dry slab avalanches occur during or immediately following loading by snowfall or wind deposition. In the absence of further loading avalanche activity decreases over time. This suggests that avalanche release is facilitated by changes in snowpack properties during loading, and that changes following loading generally help to stabilize the snowpack. This study quantifies these longer term stabilizing changes following loading. We developed a field method to rapidly increase the load on existing weak layers, and used the technique to add 10 cm of disaggregated snow on over 30 isolated columns. We then conducted Propagation Saw Tests (PSTs) in the minutes, hours and days following loading and isolation, with tests ranging from 15 minutes to 4 days. We filmed each test at 120 fps for particle tracking velocimetry analysis, and we utilized a finite element (FE) model to simulate the experiments and better interpret our results. We found that critical crack lengths increased rapidly at first and then more slowly over time. FE simulations of the experiments suggest that changes in critical crack length over time are caused by an increase in slab elastic modulus in the first hours following loading, and then caused by both increasing slab elastic modulus and weak layer specific fracture energy in days following loading. Our results help to illustrate changes in critical crack lengths in the days following loading, and are consistent with field observations of increasing avalanche activity immediately following loading events and decreasing avalanche activity afterwards.
Avalanches not only pose a major hazard to people and infrastructure, but also act as an important ecological disturbance. In many mountainous regions in North America, including areas with existing transportation corridors, reliable and consistent avalanche records are sparse or non-existent. Thus, inferring long-term avalanche patterns and associated contributory climate and weather factors requires the use of dendrochronological methods. Through the collection of regionally distributed tree-ring data recording avalanche events, we aim to address the following questions: 1) What is the regional and path specific frequency of large magnitude avalanches in the U.S. Northern Rockies?, and 2) Are there specific seasonal weather or climate variables that contribute to large magnitude and regional avalanche events? We collected 617 cross sections and 56 cores from 12 different avalanche paths in four mountain ranges in Glacier National Park and the Flathead National Forest across northwest Montana, USA. Six of these paths affect major transportation corridors, and the other six impact heavily used winter backcountry recreation zones. We identified, cross-dated, and quality ranked damage events recorded within and between samples from each path. We then implemented a double threshold analysis to account for decreasing sample numbers through time and to ensure accurate identification and dating of avalanche events. Finally, we developed an avalanche chronology of large magnitude events for each path, mountain range, and ultimately a composite record for the region. Preliminary samples from five paths in Glacier National Park reveal 1308 growth disturbances over 27 major avalanche years from 1795 to 2017. Within the major transportation corridors of GNP, a large magnitude avalanche (size 3 or greater) is likely to occur every 5 years in at least one of the five avalanche paths. These avalanche years coincide with winters characterized by high regional snowpack anomalies. Using this developing network to understand the spatiotemporal behavior of large magnitude avalanches and the contributory climate and weather factors will ultimately improve avalanche forecasting and backcountry safety efforts within the region.
Understanding snow depth distribution and change is useful for avalanche forecasting and mitigation, runoff forecasting, and infrastructure planning. Advances in remote sensing are improving the ability to collect snow depth measurements. The development of structure from motion (SfM), a photogrammetry technique, combined with the use of uninhabited aerial systems (UASs) allows for high resolution mapping of snow depth over complex terrain. The primary objective of this study was to determine the feasibility and efficacy of SfM to examine snow depth distribution and variability in complex terrain such as avalanche path starting zones at multiple times during the season. We used a 3DR Solo quadcopter UAS equipped with a Ricoh GR II camera at 90 m above ground level to acquire images of one avalanche starting zone in northwest Montana, USA. We also placed 4 to 13 ground control points (GCPs) around the area of interest to avoid traveling in steep, avalanche terrain. Ground control measurements resulted in 5 to10 cm horizontal accuracy and 5 to 15 cm vertical accuracy for 90 to 95 % of the collected points (a minimum of 100 points collected at each GCP). In-situ measurements of snow depth difference between sampling days ranged from 20 to 60 cm. We processed the images to create point clouds and digital surface models (DSMs). The resolution of the resultant DSMs was approximately 5 cm. Preliminary DSM and point cloud differencing efforts suggest relative change detection of snow depth at 5 to 15 cm resolution. The use of these relatively low cost and easily accessible methods of snow depth data collection will enhance accuracy of snow depth change estimates in starting zones and can be used to inform avalanche forecasting and mitigation efforts.
This paper presents a method developed to capture the terrain metrics of all visible skiers on an avalanche-prone backcountry slope. A remote time-lapse camera focused on Saddle Peak, a high skier-use backcountry slope in the Bridger Mountain Range of southwest Montana, USA captured 31,960 photos of 525 skiers descending in ten-second increments on 13 unique days. Skier locations (7,499 location-points) were digitized from the photos, then transformed onto a geo-referenced digital elevation model (DEM) such that terrain metrics could be applied to each of the 7,499 skier locations. Analysis of terrain metrics for each skier point compared slope, profile curvature (downslope), and plan curvature (cross-slope) over days with three different forecasted avalanche hazards (Con, Mod, Low). Terrain metrics on Considerable avalanche hazard days differed significantly from Moderate or Low avalanche hazard days (p-value < 0.001). Skier location-points transformed from the oblique photos to a geographic coordinate system had an observed horizontal spatial accuracy of 49-m with a 95% confidence interval. By capturing all visible skiers on a slope anonymously, the data provides a large and diverse data set of the terrain preferences of backcountry skiers under varying conditions.
Time lapse photography presents a simple and inexpensive tool for effectively monitoring skiers in avalanche terrain. Skier images can be useful in determining high-use areas of skier traffic, crowding or congestion issues, documenting avalanches, and recording avalanche control operations. It has also proven to be useful with assisting avalanche emergencies by providing visual survey of the avalanche path including determining number of people (if any) involved, identifying triggers and last seen points, and assessing residual risk to responders.
Avalanches are one of the greatest hazards for those recreating in snow covered mountainous terrain. In the past 20 years an average of 13 people in Canada and 27 people in the US are killed in avalanches each winter. Meanwhile, uncontrolled backcountry avalanche terrain use has significantly increased demonstrated by increased demand for avalanche education and increased sales in backcountry equipment. Lift-accessed backcountry (LABC), or avalanche terrain easily accessed from the ski resort, has seen increased usage since resorts opened boundaries in the mid-1990s. This has led to increased research interest in how people are using backcountry avalanche terrain. A simple method to reduce exposure to avalanche hazard is avoidance, however total avoidance is seldom practical. Professionals and recreational skiers alike mitigate avalanche hazard by managing exposure to terrain containing the avalanche hazard. Current research studies use GPS tracking to study the terrain metrics of backcountry skiers. This GPS research is limited to studying volunteers and professionals that willingly track and submit their trips. This approach ignores many users and thus presents a biased picture of use. This paper develops a method to capture the terrain metrics of all skiers on an avalanche-prone backcountry slope. A remote time-lapse camera focused on a high skier-use backcountry slope, (Saddle Peak, in the Bridger Mountain Range of southwest Montana, USA) captured skiers descending Saddle Peak in ten-second increments. Skier locations were digitized from the photos, then transformed onto a geo-referenced digital elevation model (DEM) such that terrain metrics could be applied to each skier location. Analysis of terrain metrics for each skier point compared slope, profile curvature (downslope), and plan curvature (cross slope) over days with different forecasted avalanche hazard. Terrain metrics on Considerable avalanche hazard days differed significantly from Moderate or Low avalanche hazard days (p-value < 0.001). Transformed data fell within a 49-m horizontal accuracy for all skier point locations with a 95% confidence interval. By capturing all skiers on a slope without their knowledge, the data collected provides a large and diverse data set of the terrain preferences of backcountry skiers under varying conditions.
Natural wet slab avalanches release when rain or melt water decreases snowpack strength, and natural dry slab avalanches release when an increased load overcomes snowpack strength. This study investigates avalanche activity resulting from mixed rain and snow falling on a faceted snowpack. This scenario produced an extensive slab avalanche cycle in March 2018 in the mountains near Ketchum, Idaho, when a 24 hour storm deposited 50 to 65 mm of water. We investigate the contributions of the pre-existing snowpack structure and weather to avalanching, and suggest possible mechanisms for the observed slab avalanche activity. At upper elevations, expected widespread, 0.5 to 3 m deep, dry slab avalanche activity occurred on many aspects. However, at middle elevations (2300 m to 2700 m) near the fluctuating rain-snow line, a low frequency return period avalanche cycle occurred in a much smaller geographical area, and was concentrated around north-northwest aspects. This differs significantly from avalanches above this elevation that spanned all aspects. This scenario illustrates the challenges forecasting and communicating these events. In our experience, some avalanche cycles exist in a continuum of avalanche types that are not easily sorted into simple “wet” and “dry” categories. We discuss challenges in using current advisory and bulletin communication tools. Furthermore, it is possible that a changing climate will increase the frequency of mixed rain-snow events in areas with traditionally drier and colder climates. We believe the avalanche community will benefit from the refinement and development of tools and techniques to describe and forecast this challenging problem.
Deep slab avalanches result from season-long weather beginning with the formation of snow stratigraphy conducive to deep slab avalanches. The onset of deep slab activity is commonly preceded by heavy snowfall, wind, rain-on-snow, or rapid warming events. Local slope weather (or surface weather) is mainly driven by atmospheric circulation at higher altitudes. The field of synoptic climatology relates these atmospheric processes to surface observations. Previous work has investigated the meteorology of deep slab avalanches, as well as the synoptic climatology of snowfall and avalanche occurrence in several different locations around the world. However, there is limited research that specifically relates atmospheric circulation to deep slab avalanches. This work aims to bridge a knowledge gap by investigating the synoptic climatology of deep slab avalanches across the western United States. We obtained daily 500 mb geopotential height maps from the NCEP/NCAR Reanalysis dataset for 38 consecutive winter seasons from 1979-2017, resulting in over 6,000 days of data. We have classified these daily 500 mb height maps using Self-Organizing Maps (SOM) to obtain 20 summary map patterns characterizing the observed modes of atmospheric variability over western North America and the eastern Pacific. We used Westwide Avalanche Network (WAN) data from three ski resorts across the western U.S. that represent the three dominant snow climates (maritime, continental, and intermountain) and count the number of deep slab avalanche events for each synoptic type at each location. We investigate which synoptic types tend to occur immediately prior to deep slab avalanche events. Our work improves our understanding of the relationship between atmospheric circulation and deep slab avalanches, thereby enabling practitioners to better anticipate deep slab avalanche events.
This conceptual model of avalanche hazard identifies the key components of avalanche hazard and structures them into a systematic, consistent workflow for hazard and risk assessments. The method is applicable to all types of avalanche forecasting operations, and the underlying principles can be applied at any scale in space or time. The concept of an avalanche problem is introduced, describing how different types of avalanche problems directly influence the assessment and management of the risk. Four sequential questions are shown to structure the assessment of avalanche hazard, namely: (1) What type of avalanche problem(s) exists? (2) Where are these problems located in the terrain? (3) How likely is it that an avalanche will occur? and (4) How big will the avalanche be? Our objective was to develop an underpinning for qualitative hazard and risk assessments and address this knowledge gap in the avalanche forecasting literature. We used judgmental decomposition to elicit the avalanche forecasting process from forecasters and then described it within a risk-based framework that is consistent with other natural hazards disciplines.
Our research expands existing knowledge of travel behavior and decision-making in avalanche terrain. We utilize GPS tracking to observe the travel behavior of lift access backcountry (LABC) skiers and in person survey data to investigate the human factors that influence their terrain choices. Our study area is the southern boundary of Bridger Bowl Ski Area, Southwest Montana, USA (N=139). We analyze travel behavior by avalanche hazard rating, and identify human factors that affect avalanche terrain choices. A subset of our results are presented here, which illustrate the demographics of our sample and highlight their travel behavior through heat maps. Our results provide a case study example of the terrain preferences and avalanche awareness of LABC skiers, and highlight specific human factors that are correlated with terrain selection. Two practical applications of this research are: 1) targeted avalanche education outreach based on our results, and 2) designing new signage to illustrate the avalanche terrain near the ski area boundary for skiers who are inexperienced in the backcountry or unfamiliar with the area.
Backcountry skiers recreate in a complex environment, with the goal of minimizing the risk of avalanche hazard and maximizing recreational opportunities. Traditional backcountry outings start and end in uncontrolled backcountry settings, with responsibility for avalanche safety and rescue falling in the hands of each group of skiers. Lift access backcountry skiing (LABC) is a particular genre of the sport in which ski resort lifts are utilized to access backcountry recreation sites. By shifting skiers mentality from the traditional backcountry setting to a LABC setting, the line between whether the ski resort provides avalanche mitigation and rescue services or not, becomes less clearly defined in the minds of skiers.
We observe the travel behavior and evaluate the decision-making biases of LABC skiers via GPS tracking and survey responses. Participants were recruited in the field, at the boundary between the relative safety of the ski resort and the uncontrolled backcountry terrain beyond. A geographic information system (GIS) is implemented to analyze the travel behavior of participants, with the aim to detect changes in behavior, as indexed via terrain used under different levels of avalanche hazard. Logistic regression and multiple linear regression are used to model travel behavior and decision-making biases as a function of observed terrain metrics.
Data was collected over 19 days from February 2017 to February 2018 at Saddle Peak backcountry area, a prime LABC location at the southern boundary of Bridger Bowl Ski Area, Montana, USA. Avalanche hazard during data collection was either moderate (119 tracks) or considerable (20 tracks). Regression models indicate subtle changes in the terrain preferences of participants under elevated avalanche hazard, with increased travel on ridge features and decreased use of convex features. These indicate a positive response, minimizing the risk of an avalanche involvement by managing slope shape. Survey responses indicate that female participants and those with greater backcountry experience have a significantly lower percentage of their total GPS track in complex avalanche terrain as defined using the avalanche terrain exposure scale. Participants who perceived the ski patrol as providing avalanche mitigation in the backcountry area adjacent to the resort had a significantly higher percentage of GPS track in complex avalanche terrain.
Avalanche climatology is defined as the study of the relationships between climate and snow avalanches, and it contributes in aiding avalanche hazard mitigation efforts. The field has evolved over the past six decades concerning methodology, data monitoring and field collection, and interdisciplinary linkages. Avalanche climate research directions are also expanding concerning treatment in both spatial scale and temporal timescales. This article provides an overview of the main themes of avalanche climate research in issues of scale from local to global, its expanding interdisciplinary nature, as well as its future challenges and directions. The growth of avalanche climatology includes themes such as its transformation from being mostly descriptive to innovative statistical methods and modeling techniques, new challenges in microscale efforts that include depth hoar aspects and increased field studies, expanding synoptic climatology applications on studying avalanche variations, efforts to reconstruct past avalanches and relate them to climatic change, and research on potential avalanche responses to recent twentieth-century and future global warming. Some suggestions on future avalanche climatology research directions include the expansion of data networks and studies that include lesser developed countries, stronger linkages of avalanche climate studies with GIScience and remote sensing applications, more innovative linkages of avalanches with climate and societal applications, and increased emphases on modeling and process-oriented approaches.
Avalanches failing on buried weak layers tend to be easier to trigger during or immediately following snowfall, suggesting that loading favorably changes snow cover properties for avalanche release. However, previous research showed that thicker, denser slabs (with more load) tend to have longer critical crack lengths in propagation saw tests (PSTs). To address these seemingly contradictory observations, this study quantifies the effect of increasing load on weak layer fracture. In previous research weak layer strengthening may have affected the relationship between load and critical cut lengths. We therefore developed a method to rapidly increase the load on existing weak layers to ensure reasonably constant weak layer strength. On 11 sampling days we used a cardboard frame the size of a PST and added 5, 10, 15, or 20 cm of disaggregated snow on top of PST columns. We allowed the added snow to sinter for 30 to 60 minutes before performing PST experiments and filming each test at 120 fps for particle tracking velocimetry (PTV) analysis. In all cases critical cut lengths dramatically decreased with increasing load (sometimes to almost 10% of unloaded cut lengths), a finding consistent with observations of easier avalanche triggering immediately after loading. Crack propagation speeds also increased with increasing load, consistent with previous work showing higher speeds with increasing slab density. From a practical perspective, this research presents an objective, repeatable, and inexpensive method for testing the effect of loading on surface weaknesses or shallow weak layers. From a scientific perspective, these experiments offer the first controlled, field-based measurements investigating the effect of loading on weak layer fracture, providing results that may be useful for calibrating theoretical models.
The open-source Know Before You Go (KBYG) avalanche awareness program provides a 50 minute introduction to avalanche safety for motorized and non-motorized users and can be presented during a school period. Intended for a 14 year-old audience with minimal avalanche experience, the program is easily adapted for older audiences and provides a valuable tool for influencing the behavior of potential, aspiring, and beginning backcountry riders. Major revisions in 2015 include messaging based on 5 points: Get the Gear Get the Training Get the Forecast Get the Picture Get out of Harm’s way The program begins with a video featuring prominent athletes, big avalanches, extreme riding footage, music, animation, and an introduction to backcountry social responsibility. The introductory video was posted online (kbyg.org) and has received 463,000 views since November. KBYG was presented 288 times to over 16,000 people in Utah and Colorado last winter, three-quarters of them in schools. We distributed the program content across North America, with major media sources helping to promote the program. The KBYG program content is available to all avalanche educators.
Avalanche activity in snow climates dominated by direct-action avalanches is primarily controlled by the local and synoptic scale meteorological conditions just prior to and during winter storm events. Previous work on Svalbard characterized the region’s unique, direct-action snow climate as “High Arctic maritime” and demonstrated an association between periods of snow drift and regional avalanche activity. This study uses a record of road closures due to drifting snow on a mountain road to further investigate Svalbard’s snow climate and avalanche regime by: 1) characterizing synoptic meteorological conditions leading to regional snow drift events, and 2) exploring the relationship between these periods of snow drift and regional avalanche activity using a case study approach. We couple a nine-year (2007- 2015) record of road closures with local meteorological observations and NCEP/NCAR synoptic composite maps to characterize the local and synoptic weather conditions leading to and occurring during periods of snow drift near Longyearbyen, Svalbard’s primary settlement. Then we compare this record of snow drift events with regional avalanche observations to illustrate the relationship between snow drift and avalanche activity on Svalbard. The results of this study will improve the understanding of Svalbard’s unique maritime snow climate and will help advance avalanche forecasting efforts throughout the region.
Snow drift endangers human life and infrastructure in alpine and arctic environments by contributing to snow avalanche formation in steep terrain and impacting transportation through reduced visibilities and drift deposition on roadways. Understanding the local and synoptic scale meteorological conditions just prior to and during hazardous snow drift conditions is a crucial element in forecasting for — and mitigating the hazards associated with — snow drift processes. This is especially true in Svalbard, a High Arctic Norwegian archipelago, where snow drift processes have been linked to avalanche activity and hazardous travel conditions in the region’s unique, direct-action maritime snow climate. This study uses a record of road closures due to drifting snow on a mountain road to further investigate Svalbard’s snow climate and avalanche regime by characterizing meteorological conditions leading to regional snow drift events and exploring the relationship between these periods of snow drift and regional avalanche activity. A nine-year record of road closures is coupled with local meteorological observations and NCEP/NCAR synoptic composite maps to characterize the local and synoptic weather conditions leading to and occurring during periods of snow drift near Longyearbyen, Svalbard’s primary settlement. This record of snow drift events is then compared with regional avalanche observations using a case study approach to illustrate the relationship between snow drift and avalanche activity in Svalbard. Results show snow drift events result from five distinct synoptic circulation types and are characterized by increased wind speeds, higher precipitation totals, and elevated air temperatures relative to average winter conditions. Four case studies qualitatively illustrate the interactions between local and synoptic weather patterns, snow drift processes, and regional avalanche activity. In addition to the suggested mitigation strategies provided, these results will help advance avalanche forecasting efforts throughout the region.
Knowing the Extended Column Test’s (ECT’s) effectiveness at different slab thicknesses is critically important for practitioners. To better understand the limitations of the ECT, we used the SnowPilot dataset to investigate the utility of ECTs for providing an index of crack initiation and propagation on varying weak layer depths. The database currently contains 5013 ECTs conducted by 386, primarily professional, users worldwide between 2007 and 2016. The broad range of observers and snowpacks in the dataset allow us to examine variations in ECT results with changing weak layer depth across seasons and locations. We found 25% of ECTP (propagating ECT) results have weak layer depths of less than 30cm, 45% have depths between 30 and 70 cm, and 30% propagate on weak layers deeper than 70 cm. We also found an increasing ratio of ECTP to ECTN (non-propagating ECT) results as depth increased. The results make intuitive sense as fracture initiation takes more force and becomes more infrequent as slab depths increase, especially once depths exceed about 1.2 m. In addition, we used these same data to look at the repeatability of ECT results in individual snow pits. In the 582 pits where two ECTs were performed on the same weak layer, 86% have similar results (either two ECTPs or two ECTNs), showing a high degree of repeatability. Our results suggest the ECT can be effectively used over a fairly wide range of weak layer depths.
The North American Avalanche Danger Scale is a tool used by backcountry avalanche forecasters to communicate the potential for avalanches to cause harm or injury to backcountry travelers. Danger ratings are the most basic component of the public forecast, providing the foundation for more nuanced descriptions of avalanche conditions. In 2010, the United States, Canada, and New Zealand adopted a consistent, five-tiered danger scale. Although widely used, we do not know how consistently the danger scale is applied both within and between avalanche forecasting operations. To address this question, we developed ten scenarios capturing a variety of avalanche conditions at the mountain range scale. We derived the scenarios from real avalanche forecasts issued by various avalanche centers throughout North America. Avalanche forecasters in the United States, Canada, and New Zealand reviewed each scenario and assigned a single danger rating for the forecast period. Results indicate that although most respondents choose ratings within one step of each other, individual forecasters can arrive at different conclusions when presented with identical information. Additionally, it appears that there are regional and/or cultural differences in how forecasters assign danger ratings.
The interaction between snowpack layers determines the snowpack stability. It follows that disrupting layers may improve stability. Ski patrols and guiding operations employ mechanical compaction methods such as boot packing, ski compaction, and explosives aiming to increase snowpack stability. High-usage backcountry areas (but still uncontrolled avalanche terrain) may become compacted, developing a snowpack with different characteristics than in lower usage adjacent terrain. A challenge in avalanche forecasting is determining how this compaction alters snowpack stability, if at all. Our results show that snow pack stability increases if compaction penetrates and impacts a weak layer, disrupts the failure plan, or affects slab cohesion. This is likely due to an increased probability of fracture arrest in a compacted snowpack from either slab or weak layer heterogeneity. While several compaction methods exist, specific research addressing different compaction techniques is lacking. This study compares the effects of different compaction methods on a snowpack. The snowpack for our case study consisted of an approx. 30 cm 1F wind slab over approx. 30 cm depth hoar. After applying mechanical compaction methods to nine slopes, we conducted ECTs and PSTs over six weeks to assess stability. We found ECTXs more common in compaction-dense areas (boot packed) than in compaction-light (skied and compaction free) areas. This research quantifies some of the effects of different compaction strategies, and provides preliminary guidance for avalanche practitioners on the most useful techniques for mitigating avalanche hazard.
Once a weak layer has fractured, slope steepness largely dictates whether or not an avalanche will release. Exceeding the critical slope angle, i.e. angle at which the slab overcomes friction, is an essential condition for dry-snow slab avalanche release. Practitioners continually take this into consideration when assessing avalanche terrain and safe travel routes in alpine environments. However, thus far, such assessments rely on rules of thumb, such as avoiding terrain steeper than 30°, rather than actual measurements. Further, research into critical slope angles has been limited and confined to dry-snow and mostly persistent weak layers and harder slabs. To address these limitations, we developed a simple smart phone app to measure the kinetic friction between the detached slab and the bed surface. We used Optical Flow method to track motion of sliding slabs. We assumed Coulomb friction to calculate the friction between the slab and bed surface and derive the critical slope angle. Using our app on existing videos allowed us to compare our measurements with prior friction research, and we found the app was able to estimate friction to within +/- 1° of previous work. Our preliminary field data consist of 16 measurements on decomposed fragments (DF) and moist faceted crystal (FC) weak layers from four pits spaced 10 meters apart on a single slope. The critical slope angle was 35 +/- 1° for the DF layer and 39 +/- 0.5° for the moist FC layer. Our data also show a constant critical slope angle within the initial 0.1 – 0.2 m. of the down-slope motion, and a decreasing critical slope angle shortly afterward. Our smartphone tool provides a method to quickly estimate critical slope angles in the field. Our goal is to link critical slope angles to specific snow cover properties, and to assess spatial and temporal variability in critical slope angle.
In 2013 the Forest Service National Avalanche Center created guidelines designed to reduce worker risk at Forest Service (FS) backcountry avalanche forecasting operations. These guidelines establish context for field operations, define worker safety philosophy and responsibility, and suggest specific risk management practices. At the conclusion of the 2015-2016 season, we conducted a survey to examine current safety practices and how operations integrated the 2013 guidelines into their safety programs. The survey results illustrate field practices and safety management at FS avalanche centers, and provide insight into how these avalanche centers might further improve workplace safety.
Particle tracking velocimetry (PTV) is a measurement technique widely used to determine displacement and velocity fields from video recordings. It is largely nonintrusive and capable of simultaneously measuring the state of deformation over an entire cross section of a sample. PTV has been used in field and laboratory experiments since the mid-1990s to study snow deformation and fracture. Initial studies focused primarily on documenting weak layer collapse and crack propagation velocities. However, recent technological and computational advances allow researchers to determine essential mechanical properties relevant to the processes involved in snow slab avalanche release. Indeed, PTV has been used to estimate the effective elastic modulus of the slab, weak layer specific fracture energy, crack propagation distance and speed, and the friction between the slab and the bed surface after fracture. In this contribution, we will give an overview of over 500 field experiments performed in Canada, USA, and Switzerland over the last 15 years, with an emphasis on relating derived snow mechanical properties to commonly observed snow cover characteristics. For instance, our results suggest that crack propagation speed, which increases with slab density, strongly correlates with crack propagation distance. Furthermore, crack face friction, which determines the critical slope angle at which an avalanche releases, is affected by the hardness differences across the weak layer. While PTV has improved our understanding of the fundamental processes involved in snow fracture, we will also highlight topics that have received little attention to date.
Measurements of the mechanical properties of snow are essential for improving our understanding and the prediction of snow failure and hence avalanche release. We performed fracture mechanical experiments in which a crack was initiated by a saw in a weak snow layer underlying cohesive snow slab layers. Using particle tracking velocimetry (PTV), the displacement field of the slab was determined and used to derive the mechanical energy of the system as a function of crack length. By fitting the estimates of mechanical energy to an analytical expression, we determined the slab effective elastic modulus and weak layer specific fracture energy for 80 different snowpack combinations, including persistent and nonpersistent weak snow layers. The effective elastic modulus of the slab ranged from 0.08 to 34 MPa and increased with mean slab density following a power-law relationship. The weak layer specific fracture energy ranged from 0.08 to 2.7 J m−2 and increased with overburden. While the values obtained for the effective elastic modulus of the slab agree with previously published low-frequency laboratory measurements over the entire density range, the values of the weak layer specific fracture energy are in some cases unrealistically high as they exceeded those of ice. We attribute this discrepancy to the fact that our linear elastic approach does not account for energy dissipation due to non-linear parts of the deformation in the slab and/or weak layer, which would undoubtedly decrease the amount of strain energy available for crack propagation.
Predicting avalanche danger depends on knowledge of the existing snowpack structure and the current and forecasted weather conditions. In remote and data sparse areas this information can be difficult, if not impossible, to obtain, increasing the uncertainty and challenge of avalanche forecasting. In this study, we coupled the Weather Research and Forecasting (WRF) model with the snow cover model SNOWPACK to simulate the evolution of the snow structure for several mountainous locations throughout western Montana, USA during the 2014-2015 and 2015-2016 winter seasons. We then compared the model output to manual snow profiles and snow and avalanche observations to assess the quantitative and qualitative accuracy of several snowpack parameters (grain form, grain size, density, stratigraphy, etc.) during significant avalanche episodes. At our study sites, the WRF model tended to over-forecast precipitation and wind, which impacted the accuracy of the simulated snow depths and SWE throughout most of the study period. Despite this, the SNOWPACKWRF model chain managed to approximate the snowpack stratigraphy observed throughout the two seasons including early season faceted snow, the formation of various melt-freeze crusts, the spring transition to an isothermal snowpack, and the general snowpack structure during several significant avalanche events. Interestingly, the SNOWPACK-WRF simulation was statistically comparable in accuracy to a SNOWPACK simulation driven with locally observed weather data. Overall, the model chain showed potential as a useful tool for avalanche forecasting, but advances in numerical weather and avalanche models will be necessary for widespread acceptance and use in the snow and avalanche industry.
Predicting avalanche danger depends on knowledge of the existing snowpack structure and the current and forecasted weather conditions. In remote and data sparse areas this information can be difficult, if not impossible, to obtain, increasing the uncertainty and challenge of avalanche forecasting. In this study, we coupled the Weather Research and Forecasting (WRF) model with the snow cover model SNOWPACK to simulate the evolution of the snow structure for several mountainous locations throughout western Montana, USA during the 2014-2015 and 2015-2016 winter seasons. We then compared the model output to manual snow profiles and snow and avalanche observations to assess the quantitative and qualitative accuracy of several snowpack parameters (grain form, grain size, density, stratigraphy, etc.) during significant avalanche episodes. At our study sites, the WRF model tended to over-forecast precipitation and wind, which impacted the accuracy of the simulated snow depths and SWE throughout most of the study period. Despite this, the SNOWPACK-WRF model chain managed to approximate the snowpack stratigraphy observed throughout the two seasons including early season faceted snow, the formation of various melt-freeze crusts, the spring transition to an isothermal snowpack, and the general snowpack structure during several significant avalanche events. Interestingly, the SNOWPACK-WRF simulation was statistically comparable in accuracy to a SNOWPACK simulation driven with locally observed weather data. Overall, the model chain showed potential as a useful tool for avalanche forecasting, but advances in numerical weather and avalanche models will be necessary for widespread acceptance and use in the snow and avalanche industry.
Extended Column Tests (ECTs) are used to assess crack initiation and propagation. Previous research shows that tests 90 cm in length may propagate, suggesting instability, while tests 2min length may not propagate, suggesting stability, for identical snowpacks. A practical question is: are 90 cm ECTs optimal for assessing stability? To test the added value of 2 m ECTs for stability evaluation, we collected data on 220 ECTs, with 136 side-by-side standard length ECTPs (full propagation indicating instability) followed by 2 m ECTs. We only performed 2 m ECTs after a standard ECT propagated because we assumed 2 m ECTs would not propagate if standard length tests did not. These tests were preceded by an a priori stability assessment. Our results show imbalances for both tests. The ECT had a similar probability of detection (0.88-0.92, POD), i.e. the ability to detect unstable conditions, as in previous studies, but a much lower probability of null events (0.54-0.75, PON), i.e. the ability to detect stable conditions, with variation due to the binary classification of “Fair” stability as stable or unstable. Adding a 2m test after an ECTP result lowered the POD (0.49-0.58), but substantially raised the PON (0.88-0.98) of the combined tests. The proportion of tests in agreement, i.e. ECTP and 2 m ECTP, increases with decreasing stability. We conclude that an ECTP followed by a 2m ECTP is a clear red flag, indicating instability. the interpretation of an ECTP followed by a 2 m ECTN/X (no propagation) is not clear. Though this result suggests stability, a high potential for a false stable result means we cannot recommend the 2 m ECT for binary stability assessments.
The distribution of snow depth in avalanche starting zones exerts a strong influence on avalanche potential and character. Extreme depth changes over short distances are common, especially in wind-affected, above-treeline environments. Snowdepth also affects the ease of avalanche triggering. Experience shows that avalanche reduction efforts are often more successful when targeting shallow trigger point areas near deeper slabs with explosives or ski cutting. Our paper explores the use of high-resolution (cm scale) snow depth and snow depth change maps from terrestrial laser scanning (TLS) data to quantify loading patterns for use in both pre-control planning and in post-control assessment.
We present results from a pilot study in three study areas at the Arapahoe Basin ski area in Colorado, USA. A snow-free reference data set was collected in a summer TLS survey. Mapping multiple times during the snow season allowed us to produce time series maps of snow depth and snowdepth change at high resolution to explore depth and slab thickness variations due to wind redistribution. We conducted surveys before and after loading events and control work, allowing the exploration of loading patterns, slab thickness, shot and ski cut locations, bed surfaces, entrainment, and avalanche characteristics.We also evaluate the state of TLS for use in operational avalanche control settings.
Dry-snow slab avalanches are generally caused by a sequence of fracture processes including (1) failure initiation in a weak snow layer underlying a cohesive slab, (2) crack propagation within the weak layer and (3) tensile fracture through the slab which leads to its detachment. During the past decades, theoretical and experimental work has gradually led to a better understanding of the fracture process in snow involving the collapse of the structure in the weak layer during fracture. This now allows us to better model failure initiation and the onset of crack propagation, i.e., to estimate the critical length required for crack propagation. On the other hand, our understanding of dynamic crack propagation and fracture arrest propensity is still very limited.
To shed more light on this issue, we performed numerical propagation saw test (PST) experiments applying the discrete element (DE) method and compared the numerical results with field measurements based on particle tracking. The goal is to investigate the influence of weak layer failure and the mechanical properties of the slab on crack propagation and fracture arrest propensity. Crack propagation speeds and distances before fracture arrest were derived from the DE simulations for different snowpack configurations and mechanical properties. Then, in order to compare the numerical and experimental results, the slab mechanical properties (Young’s modulus and strength) which are not measured in the field were derived from density. The simulations nicely reproduced the process of crack propagation observed in field PSTs. Finally, the mechanical processes at play were analyzed in depth which led to suggestions for minimum column length in field PSTs.
Deep slab avalanches are particularly challenging to forecast. These avalanches are difficult to trigger, yet when they release they tend to propagate far and can result in large and destructive avalanches. We utilized a 44-year record of avalanche control and meteorological data from Bridger Bowl ski area in southwest Montana to test the usefulness of meteorological variables for predicting seasons and days with deep slab avalanches. We defined deep slab avalanches as those that failed on persistent weak layers deeper than 0.9 m, and that occurred after February 1st. Previous studies often used meteorological variables from days prior to avalanches, but we also considered meteorological variables over the early months of the season. We used classification trees and random forests for our analyses. Our results showed seasons with either dry or wet deep slabs on persistent weak layers typically had less precipitation from November through January than seasons without deep slabs on persistent weak layers. Days with deep slab avalanches on persistent weak layers often had warmer minimum 24-hour air temperatures, and more precipitation over the prior seven days, than days without deep slabs on persistent weak layers. Days with deep wet slab avalanches on persistent weak layers were typically preceded by three days of above freezing air temperatures. Seasonal and daily meteorological variables were found useful to aid forecasting dry and wet deep slab avalanches on persistent weak layers, and should be used in combination with continuous observation of the snowpack and avalanche activity.
Snow avalanches pose a significant hazard to winter recreationalists travelling in the backcountry and are difficult to predict on individual slopes. Weak layers responsible for these avalanches may form at the surface multiple times during the winter season and are buried by subsequent snowfall. Understanding causes of slope-scale weak layer variations during formation and destruction periods is crucial for gaining an understanding of their distribution after burial. Persistent weak snow layers, such as surface hoar, can pose hazards for months after burial. This study examines surface hoar crystals on the surface, directly after formation, in two small meadow openings in southwest Montana. Data collection occurred during two winter seasons for three surface hoar formation events. Three environmental metrics associated with surface hoar growth processes in meadow openings are explored and their relationships with crystal size examined using spatial regression and regression tree analysis. The spatial structure for each event is described using multiple crystal sizing measures through semi-variograms. Surface hoar crystals tended to grow largest in areas that were both shaded and possessed large unobstructed views of the sky on north and south aspects. The range of spatial autocorrelation for surface hoar crystal sizes varied from 7 m to beyond 25 m and differed depending on event or crystal sizing measure. Results vary between events and suggest the drivers controlling surface hoar growth are unique to each area and not consistent between events. This research highlights the need for multiple slope-scale snow stability assessments for understanding the distribution of a buried weak layer of surface hoar in a meadow opening. Targeted areas for assessment should incorporate a basic understanding of a meadow’s shading and canopy openness and how this varies over a winter season.
Extended Column Tests (ECT) and Propagation Saw Tests (PST) are used to assess crack propagation; that is, the likelihood of a crack self-propagating. Yet, we present findings that show that full crack propagation to the end of the beam depends on beam length. The practical question is: are beams about 1 m long optimal for assessing stability? Finite element modeling shows that the so called “far edge attraction” becomes insignificant for beams longer than 2 m. In other words, full crack propagation be-comes independent of beam length when beams ≥ 2 m are used. To test the accuracy of 2 m tests for stability evaluation, we collected data on 135 side-by-side standard length ECTPs (full propagation) fol-lowed by 2 m ECTs. We only focused on ECTPs because we assumed 2 m ECTs would not propagate if standard length tests did not. These tests were preceded by an a priori stability assessment to reduce circularity or stability ratings based on test results. Our results show that the proportion of tests in agree-ment, i.e. ECTP and 2 m ECTP, increase with decreasing stability. We conclude that an ECTP followed by a 2 m ECTP is a clear red flag. The interpretation of an ECTP followed by a 2 m ECTN/X (no propaga-tion) is not clear. The main finding for practitioners is that a 2 m ECT can be used to give additional infor-mation about slope stability following an ECTP.
The Extended Column Test (ECT) and the Propagation Saw Test (PST) are two commonly used tests to assess the likelihood of crack propagation in a snowpack. Guidelines suggest beams with lengths of around 1 m, yet little is known about how test length affects propagation. Thus, we performed 163 ECTs and PSTs 1.0–10.0 m long. On days with full crack propagation in 1.0–1.5 m tests, we then made videos of tests 2.0–10.0 m long. We inserted markers for particle tracking to measure collapse amplitude, propagation speed, and wavelength. We also used a finite element (FE) model to simulate the strain energy release rate at fixed crack lengths. We find that (1) the proportion of tests with full propagation decreased with test length; (2) collapse was greater at the ends of the beams than in the centers; (3) collapse amplitude was independent of beam length and did not reach a constant value; (4) collapse wavelengths in the longer tests were around 3 m, two times greater than what is predicted by the anticrack model. We also confirmed the prediction that centered PSTs had double the critical length of edge PSTs. Based on our results, we conclude that cracks propagated more frequently in the shorter tests because of increased stress concentration from the far edge. The FE model suggests this edge effect occurs for PSTs of up to 2 m long or a crack to beam length ratio ≥ 0.20. Our results suggest that ECT and PST length guidelines may need to be revisited.
Although all avalanche types are difficult to forecast and mitigate, wet slab avalanches are particularly problematic. The dynamics of wet slabs are poorly understood and large wet slab cycles are relatively rare in comparison to dry slab cycles. We provide a case study of a historic wet slab cycle at Crystal Mountain Ski Area in Washington State in March of 2014. This cycle resulted in numerous large natural and explosive-triggered avalanches, including one particularly notable slide that destroyed a 34-year old chairlift and had an alpha angle of only 20º. The avalanches occurred after a brief period of warming followed by rain. Though the amount of precipitation (1.4 inches or 36 mm) on the preceding day was not atypical for this location, we believe the existing snow structure was a key contributing factor to the avalanches. The snowpack contained numerous crusts and layers of facets near the base due to a lack of early-season snowfall, and a deep and relatively homogenous slab formed by a 10-day storm that dropped 12′ (3.6 m) of new snow. Water penetration through the slab progressed much more rapidly than expected, perhaps due to the homogeneous nature of the slab. We discuss the snowpack evolution, contributory weather factors, operational challenges, mitigation efforts (including whether or not to use mitigation), and difficult decision-making that allowed the risks of this avalanche cycle to be managed without loss of life. A thin margin for error existed in this event, and we hope other practitioners can learn from our case study.
Though recent work on fracture has provided us with new stability tests and improved our knowledge of avalanche release, our understanding of fracture arrest is still limited. We studied fracture arrest by making modifications to weak layers and slabs in a series of propagation saw tests (PSTs). We conducted more than 100 tests at a single study plot over eight weeks during which cracks along a surface hoar weak layer consistently propagated. Slab characteristics changed dramatically over time, with increasing slab depth, density, and hardness. High speed videos of more than 70 tests allow the analysis of fracture arrest dynamics. Our results show that removing the weak layer had no effect on propagation. Fracture arrest only occurred when we replaced a 30 cm long section of the weak layer with a non-collapsible structure. Modifying the slab by introducing slope normal cracks (either from the surface down or from the weak layer up) showed that sometimes small changes to soft (F hardness) parts of a slab were sufficient to arrest fractures through slab fracture, while other times only a strong thin portion of a thick slab was capable of communicating fracture farther down the beam. Our results suggest that the tensile strength of the upper layers of the slab is a key component for crack propagation. This work demonstrates the importance of the slab in slope stability, and it also suggests avenues for developing tests capable of assessing slab characteristics conducive to fracture propagation.
Conducting stability tests in avalanche terrain is inherently dangerous since it exposes the observer to the potential of being caught in an avalanche. Recent work shows that such exposure may be unnecessary since the results of extended column tests (ECTs) and propagation saw tests (PSTs) are largely independent of slope angle, allowing for data collection in safer locations. Conversely, some past work shows that compression tests (CTs) are slope angle dependent. In this paper, we test the effect of slope angle on CTs using similar methods as the recent ECT work. We collected field data on three sep-arate days with persistent weak layers in Montana and California. Our slopes exhibited gradual changes in steepness, allowing us to sample a variety of slope angles with minimal snow structure changes. We also employed a second method to reinforce our results. Utilizing the SnowPilot dataset, we analyzed the difference between propagating ECTs and CTs on the same layer, and compared that difference with slope angle. Our fieldwork shows that the CT test results either did not change or increased slightly with increasing slope angle. Further, the SnowPilot data demonstrate that the difference between ECTs and CTs is not statistically dependent on slope angle, reinforcing conclusions from our field work. Our results have significant theoretical implications, but the practical implications are even more important since this work suggests that, in addition to ECTs and PSTs, CTs can be conducted in safer low-angle terrain.
Understanding what controls coarse scale snowpack properties, such as surface hoar distribution, is imperative for predicting snow avalanches. Due in part to the inherent difficulties of winter travel in mountainous terrain, most spatial variability investigations of snow properties have been limited to relatively fine scales. To quantify snow surface spatial variability at the basin, region, and mountain range scales, a team of heli-skiing guides recorded observations describing surface hoar presence or absence coordinates, crystal size, and elevation throughout four major surface hoar formation periods over two heli-skiing seasons in rugged alpine terrain near Haines, Alaska across an extent of nearly 60 km. Geostatistical analysis yielded spherical semivariogram autocorrelation ranges from approximately 3-25 km, which is similar in size to many of the basins and regions within the study area. Kriging models built from the semivariograms were produced to aid geographic visualization of coarse scale snowpack processes. Geographically Weighted Regression revealed a positive relationship between elevation and surface hoar crystal size with adjusted R 2 values averaging near 0.40. The results of this research suggest it may be possible to identify areas with greater surface hoar growth and persistence potentials as a consequence of synoptic onshore or offshore flow, and glacially influenced katabatic winds. Additionally, larger surface hoar crystals may be found in the higher elevation avalanche starting zones in the alpine glaciated terrain near Haines, Alaska. These results can help in future efforts to forecast snow stability patterns over large areas.
The distribution of snow depth in avalanche starting zones exerts a strong influence on avalanche potential and character. Extreme depth changes over short distances are common, especially in wind-affected, above-treeline environments. Snow depth also affects the ease of avalanche triggering. Experience shows that avalanche reduction efforts are often more successful when targeting shallow trigger point areas near deeper slabs with explosives or ski cutting. Our paper explores the use of high resolution (cm scale) snow depth and snow depth change maps from terrestrial laser scanning (TLS) data to quantify loading patterns for use in both pre-control planning and in post-control assessment.
Experience suggests that shallow, steep zones on slopes are potential dry slab avalanche trigger points. However, a scientific understanding of this common knowledge is not well quantified due to the spatial variability of snowpack stability, which is governed by various internal and external processes. Currently, the best way to investigate these processes is through point stability testing on small slopes. We thus performed Compression and Extended Column Tests (CTs and ECTs) on three small, wind-affected alpine slopes in central Svalbard. While one study slope (Gangskaret) had smooth ground topography, the other two (Fardalen and Larsbreen) exhibited irregular, rugged ground topography. Our resuts show that weak layer reactivity was largely influenced by the ground topography, as snow depth is a function of terrain on wind-affected slopes. Slab thickness determines weak layer sensitivity where the ground topography is rugged. Thus, the most unstable spots on these slopes coincided with the shallower zones characteristic of steeper ground surfaces inclinations where the snowpack is thin and the weak layers are close to the surface. This was not as pronounced on slopes with smooth ground topography. However, as snowpack develops and thickens to a “snow depth threshold X”, the group irregularities are leveled out and their influence diminishes. Thus, knowing the terrain is crucial. Moreover, it is crucial to follow the seasonal snowpack development and extreme weather events that influence it. We found inverse relationships between stability and slab thickness for weak layers that developed early in the season. These early instabilities displayed discontinuity due to melt out over topographic highs during rain-on-snow events, but were left in a preserved state in topographic lows that became overlain by shielding refrozen melt form layers.
Dry-snow slab avalanches are generally caused by a sequence of fracture processes including (1) failure initiation in a weak snow layer underlying a cohesive slab, (2) crack propagation within the weak layer and (3) tensile fracture through the slab which leads to its detachment. During the past decades, theoretical and experimental work has gradually led to a better understanding of the fracture process in snow involving the collapse of the structure in the weak layer during fracture. This now allows us to better model failure initiation and the onset of crack propagation, i.e. to estimate the critical length required for crack propagation. On the other hand, our understanding of dynamic crack propagation is still very limited. For instance, it is not uncommon to perform field measurements with widespread crack propagation on one day, while a few days later, with very little changes to the snow-pack, crack propagation does not occur anymore. Thus far, there is no clear theoretical framework to interpret such observations, and it is not clear how and which snowpack properties affect dynamic crack propagation. To shed more light on this issue, we performed numerical propagation saw test (PST) experiments applying the discrete element (DE) method and compared the numerical results with field measurements based on particle tracking. The goal is to investigate the influence of weak layer failure and the mechanical properties of the slab on crack propagation. Crack propagation velocities and distances before fracture arrest derived from the DE simulations were in good agreement with experimental data suggesting that the simulations can reproduce crack propagation in PSTs.
Observations by practitioners suggest that fracture speeds are often much faster than re-cently reported measurements. These measurements include speeds along isolated beams ranging from 15 to 45 m/s, as well as the speed of a meadow collapse of 20 m/s. Since reported velocities appear to be slower than many observed avalanches, we analyzed a sample of 27 videos to estimate avalanche fracture speeds. Time was measured by counting video frames from the explosive trigger to visible slab fractures, and distances by on-site measurements, picture scaling, and Google Earth. Though our speeds vary widely (from 18 to 428 m/s), most of our values fall in the range of 50 to 125 m/s, which clearly exceeds previously reported values. We also investigated the relationship between fracture speed and other characteristics, such as explosive size. Interestingly, in videos with visible crowns and stauch-walls, the stauchwall opens first or at the same time 33% of the time. Our results may improve our under-standing of avalanche release, as well as providing practical guidance for explosive placements.
Most avalanche accidents are the results of an avalanche triggered by the victim, or a member of the victim’s party. Many of these accidents are the result of uncertainty regarding the stability of the snowpack. Spatial variability of snow stability is a significant cause of this uncertainty. There has been significant previous work on the spatial variability of snow stability at multiple spatial scales, but most of these studies have focused on measures of fracture initiation. This study investigates the spatial variability of Extended Column Test (ECT) results (an index of fracture propagation). We measured the spatial variability of ECT results in 23 grids across southwest Montana over the course of two winters. These slopes were all topographically uniform, wind sheltered clearings, with snowpacks relatively undisturbed by skiers or snowmobiles. Twenty eight ECTs were spaced across each grid in a standardized layout with a 30 m x 30 m extent. Our results are consistent with previous work, with some grids showing high levels of variability as well as other grids with relatively homogenous results. We found no consistent spatial pattern to our test results. We tested slopes with a variety of weak layers (surface hoar, depth hoar, new snow, and near surface facets), slab characteristics (slab hardness, slab depth), and snow depths and found no correlations with ECT results. We found a relationship between the forecasted regional avalanche danger and the percent of ECTs showing propagation in a grid. As the regional danger increases the percent of ECTs propagating in a grid does as well. ECT results are most variable under moderate danger. When the regional avalanche danger is either considerable or low, results are likely to be more consistent. The key practical implication of our results is that ECTs, like all other stability tests, should be interpreted with an appropriate level of caution and in consideration of all other relevant variables. The spatial variability of this test has the potential to be high on some slopes, while on other slopes test results will be entirely in agreement.
Though a significant amount of research examines the spatial variability of snow stability at the slope scale, most of that work focuses on tests primarily related to fracture initiation. The small number of studies examining the spatial variability of fracture propagation (utilizing tests such as the ex-tended column test (ECT)) are inconsistent. Some work reports homogenous ECT results, while later studies showed more variable and difficult to explain results. Because of these inconsistencies, we measured the spatial variability ECT results on multiple slopes in southwest Montana over the course of two winters. We sampled 23 grids, with each grid containing 28 ECTs for a total of 644 ECTs using a pre-defined semi-random 30m by 30m extent. We tested slopes with a variety of weak layers (surface hoar, depth hoar, new snow, and near surface facets), slab characteristics (slab hardness, slab depth), and snow depths. Further, we sampled during varying levels of forecasted regional avalanche danger. Our data demonstrate that considerable spatial variability in ECT potential exists on many slopes, even with-out substantial variation in snowpack structure. When the regional avalanche danger is either considerable or low, results are likely to be more consistent, but when the regional danger is moderate, results tend to be more variable. Further, the ratio of ECTPs to ECTNs can be correlated with the forecasted danger level. The practical implications of our results are that ECTs, like all other stability tests, should be inter-preted with an appropriate level of caution.
Deep slab avalanches are a particularly challenging avalanche forecasting problem. These avalanches are typically difficult to trigger, yet when they release they tend to propagate far and can result in large and destructive avalanches. For this work we define deep slab avalanches as those that fail on persistent weak layers deeper than 0.9m (3 feet), and that occur after February 1st. We utilized a 44-year record of avalanche control and meteorological data from Bridger Bowl Ski Area in southwest Montana to test the usefulness of meteorological variables for predicting seasons with deep slab avalanches. While previous studies often exclusively use data from the days preceding deep slab cycles, we include meteorological metrics over the early months of the season when persistent weak layers form. We used classification trees for our analyses. Our results showed that seasons with avalanches on deep persistent weak layers typically had drier early months, and often had maximum snow depth greater than 88cm in November, which provided ideal conditions for persistent weak layer development. This paper provides insights for ski patrollers, guides, and avalanche forecasters who seek to understand the seasonal conditions that are conducive to deep slab avalanches on persistent weak layers later in the season.
Snow avalanches are a potentially fatal and highly destructive natural hazard. Snow slab avalanches occur in steep alpine terrain due to an unstable layered snowpack. When a consolidated layer of snow forms a slab above a weak layer of snow the slab may collapse and slide downhill due to gravitational and applied forces (e.g., the weight of a skier, explosive, or new snowfall). Persistent weak layers form in the snowpack due to strong vapor pressure gradients, and they can last for weeks to months as a slab builds above them. Avalanches on persistent weak layers become less frequent, yet are typically larger and more destructive the longer and deeper the layer is buried. Deep slab avalanches on persistent weak layers pose a difficult forecasting problem due to their low likelihood of occurrence and potentially high consequences. This thesis aims to identify meteorological metrics that are associated with deep slabs on persistent weak layers. We used univariate analysis, classification trees, and random forests to explore relationships between seasons with deep slabs and summaries of meteorological metrics over the beginning of the season during weak layer formation. We also looked at the relationship between days with these avalanches and summaries of meteorological metrics over the days prior to them. In addition, we reviewed a case study of a season that had multiple deep slabs on a persistent weak layer and a historic wet slab avalanche cycle on the same layer, at Bridger Bowl ski area. Seasons with deep slabs typically had relatively low precipitation throughout the early part of the season (i.e., November – January), and a snowpack in the beginning of the season that was sufficiently deep, but shallow enough for a weak layer to develop. Our results also showed warmer twenty-four hour temperatures and more precipitation over seven day prior to days with dry deep slabs, and extended periods of above freezing temperatures were seen prior to days with deep wet slabs. These results are in line with previous research and are suggestive of meteorological summaries that may be useful to forecast deep slab avalanches on persistent weak layers.
Dry slab avalanches start when a catastrophic failure along a weak snowpack layer separates the slab from the underlying snowpack over a large section of the slope. Most research and papers focus on crack propagation, but little work has been done on fracture arrest, which is critically important for understanding slab boundary locations. This study investigates some of the factors contributing to fracture arrest at boundaries of actual avalanches. We targeted these areas by conducting a series of 2m ECTs (ECT200s) along crown walls, and working toward the flanks of 15 Storm-Slab and seven Persistent-Slab avalanches with maximum crown depth between of 45 and 80 cm. At 16 of our avalanches we recorded snowpack and terrain changes in 32 locations where fractures arrested as evidenced by our tests transitioning from ECT200Ps to ECT200Ns. In four ECT200Ns (13%), we observed a sharp increase in slab density down column from fracture arrest, in 21 ECT200Ns (65%) there was a decrease in slab thickness, density or both, and in seven ECT200Ns (22%) the weak layer disappeared within the scope of the column. The slab boundaries in the other six of our 22 avalanches were dictated by the slab and weak layer friction rather than weak layer fracture arrest. In these avalanches, slab fractures appeared beyond the release areas, but slope steepness around these fractures wasn’t sufficient to overcome friction. Our work helps improve terrain management strategies and suggests terrain-related safety margins for different dry slab avalanche problems.
Workplace avalanche accidents claimed the lives of 11 avalanche workers in the United States over the last five years. Though none of the victims worked at a backcountry avalanche center, each accident reminds us of the danger of working in avalanche terrain. In response to these accidents, and in particular to an accident involving a Utah Department of Transportation avalanche forecaster, the Forest Service National Avalanche Center started a dialogue between the U.S. avalanche centers aimed at improving worker safety and increasing consistency in safety procedures. Our discussion began with the practice of solo travel by avalanche workers, but quickly transitioned into a more comprehensive project. The result was the creation of guidelines designed to reduce risk during field work. The guidelines establish context for field operations, define worker safety philosophy and responsibility, and improve risk management by requiring documentation of procedures related to check-in/check-out, required safety equipment, working alone, and emergency response. One of the key components of the guidelines is a pre-field work checklist and critical thinking exercise. U.S. avalanche centers are unique and diverse operations, adding complexity to the project. In the end, reducing or preventing accidents requires a holistic approach to safety. This approach must address the fundamental questions of who we are as a professional group, to what extent are we willing to expose our workers to potential hazards, and what methods are at our disposal to mitigate “acceptable risk”.
The Extended Column Test (ECT) has become increasingly popular for assessing snowpack stability. What happens to the snow slab and the underlying weak layer during the test remains largely unknown. Such work has been done for the Propagation Saw Test (PST), but not for the ECT. We therefore analyzed high-speed videos of ECTs and adjacent PSTs using particle tracking to better understand the mechanics of the ECT. Our results show that in an ECT, tapping on one end of the column had no observable effect on the opposite end, and that fracture initiates in an area of the weak layer directly under the shovel at the free edge of the column. We observed no signs of progressive damage accumulation in the weak layer during tapping, but rather a single rapid collapse when fracture initiated. In contrast, in PSTs, we observed slab bending prior to weak layer fracture. During fracture, weak layers in ECTs compact on the order of several mm, similar to measurements obtained from PSTs. Measured propagation speeds, on the order of 20 to 30 m s-1, we also similar to those from PSTs. The similarities between ECT results and those with PSTs give us confidence that the fracture mechanics are similar regardless of the triggering mechanism. From a practical perspective, our results suggest that the ECT is indeed measuring the propensity of a crack to propagate at the small scale of the ECT block, thus providing information on a critically important property of snow stability in our tests.
Understanding variations in slope scale snowpack properties influences stability assessments at the slope and regional scale. Previous studies have shown that surface hoar, a prominent weak layer type, can vary in initial crystal size and height across small, seemingly homogenous meadows and sparsely forested areas on northerly aspects. The differences in size have generally been attributed to the radiation balance, which is difficult to estimate in the field. This study aims to further investigate the effect of canopy cover and shading on the growth of surface hoar in small forested meadow openings just after initial growth and prior to burial. Two small (approx. 30m x 50m) study plots located in southwest Montana were selected. The study plots, one northern and one southern aspect, are mostly planar 10° meadows surrounded by heavy tree cover. We collected 200 samples with 2119 individual crystal observations, and estimated shading and hemispheric sky visibility to explain the difference in sizes of surface hoar in each meadow. Findings indicate that the strength of these determinants varies depending on aspect and how the surface hoar size is determined.
Surface hoar (SH) crystals, once buried, often result in a persistent weak layer within the snowpack that contributes to instability on slopes in the mountains of southwestern Montana as well as many other mountainous regions of the world. The influence of local meteorological conditions and gross-scale topographical features (e.g. slope and aspect) that influence the formation of surface hoar are relatively well understood. However, the relationships between synoptic atmospheric conditions and the formation of surface hoar are not very well known, nor have they been extensively studied. To investigate these relationships, atmospheric patterns from NCEP/NCAR synoptic composite maps were obtained for periods with varying amounts of surface hoar presence (four intervals from <20% to >80%) for our study area in SW Montana. This study used a comprehensive suite of 127 days of manual observations of surface hoar presence and absence from Pioneer Mountain within the Yellowstone Club in southwestern Montana from December 2012 – April 2014. Each of these days provides detailed observations from 16 sites distributed across several different elevations and all aspects on Pioneer Mountain. The composite maps show that higher sea level pressure, northerly winds and higher than average 500 hPa geopotential heights tend to favor more extensive surface hoar formation. This knowledge along with further research and analysis could potentially provide better insight into long-term forecasts of the spatial distribution of surface hoar presence.
More winter recreationists are venturing into “extreme” terrain each year, and avalanche fatalities in that terrain are increasing. The slope-scale spatial variability of snow stability and how it relates to this complex terrain is critically important but poorly understood. In this study, we use terrain parameters to model potential trigger locations (PTLs) of slab avalanches, which are defined based on a minimum slab thickness overlying a persistent weak layer or the presence of a weak layer on the snow surface which could be subsequently buried. In a sample of seventeen couloirs from Lone Mountain, Montana, field teams tracked and mapped persistent weak layers and slabs with probe and pit sampling. We used terrain parameters derived from a one-meter digital elevation model to explore the relationships between PTLs and terrain, and our results show strong statistical relationships exist. However, results varied widely from couloir to couloir, suggesting that the relationships between terrain and PTLs in each couloir are unique and highly complex. For these steep alpine couloirs, parameters relating to wind deposition, wind scouring, and sluffing are most strongly associated with PTLs. The influences of these and other terrain parameters vary, depending on broader-scale terrain characteristics, prior weather patterns, and seasonal trends. Win an understanding of the broader scale influences and physical processes involved, we can use terrain to optimize stability test locations, explosive placements, or route selection. The unique nature of each couloir means that simple rules relating terrain to PTLs will not apply, although couloirs in the same cirque generally share similarities. This study will help to improve practical decision-making as well as future modeling efforts.
Explosives are a critically important component of avalanche control programs. They are used to both initiate avalanches and to test snowpack instability by ski areas, highway departments and other avalanche programs around the world. Current understanding of the effects of explosives on snow is mainly limited to shock wave behavior demonstrated through stress wave velocities, pressures and attenuation. This study seeks to enhance current knowledge of how explosives physically alter snow by providing data from field-based observations and analyses that quantify the effect of explosives on snow density, snow hardness and snow stability test results. Density, hardness and stability test results were evaluated both before and after the application of 0.9 kg cast pentolite boosters as surface and air blasts. Changes in these properties were evaluated at specified distances up to 5.5 meters (m) from the blast center for surface blasts and up to 4 m from the blast center for air blasts. A density gauge, hand hardness, a ram penetrometer, Compression Tests (CTs), and Extended Column Tests (ECTs) were used. In addition to the field based observations, the measurement error of the density gauge was established in laboratory tests. Results from surface blasts did not provide conclusive data. Air blasts yielded statistically significant density increases out to a distance of 1.5 m from the blast center and down to a depth of 50 centimeters (cm). Statistically significant density increases were also observed at the surface (down to 20 cm) out to a distance of 4 m. Hardness data showed little to no measurable change. Results from CTs showed a statistically significant decrease in the number of taps needed for column failure 4 m from the blast center in the post-explosive tests. A smaller data set of ECT results showed no overall change in ECT score. The findings of this study provide a better understanding of the physical changes in snow following explosives, which may lead to more effective and efficient avalanche risk mitigation.
At ski areas, a majority of avalanches fail in storm snow. We investigate these avalanches using stability tests and avalanche observations from California and Alaska. Collapse amplitudes during fracture, measured using particle tracking, were 1 mm for a failure layer of precipitation particles and 7 mm for a layer of unrimed sectored plates. Stability test results showed little dependence on slope angle, suggesting that both precipitation particles and older faceted crystals (persistent weak layers) fail as described by the anticrack model, with collapse providing energy. Using observations from avalanche control work at Mammoth Mountain, CA USA, a large coastal ski area where 9/10 avalanches fail in storm snow, we examined Extended Column Test (ECT) results and their relation to avalanche activity. ECT propagation was a powerful predictor; days with ECTs that propagated had significantly more and larger avalanches. Since other studies have shown that the ECT is an effective predictor of avalanches involving persistent weak layers, we suggest that the ECT is an effective test to predict both types of avalanches, those that fail in storm snow and those that fail on persistent weak layers.
Storm snow often avalanches before crystals metamorphose into faceted or rounded shapes, which typically occurs within a few days. We call such crystals nonpersistent, to distinguish them from snow crystals that persist within the snowpack for weeks or even months. Nonpersistent crystals can form weak layers or interfaces that are common sources of failure for avalanches. The anticrack fracture model emphasizes collapse and predicts that triggering is almost independent of slope angle, but this prediction has only been tested on persistent weak layers. In this study, dozens of stability tests show that both nonpersistent and persistent crystals collapse during failure, and that slope angle does not affect triggering (although slope angle determines whether collapse leads to an avalanche). Our findings suggest that avalanches in storm snow and persistent weak layer share the same failure mechanism described by the anticrack model, with collapse providing the fracture energy. Manual hardness measurements and near-infrared measurements of grain size sometimes showed thin weak layers of softer and larger crystals in storm snow, but often showed failures at interfaces marked by softer layers above and harder layers below. We suggest collapse often occurs in crystals at the bottom of the slab. Planar crystals such as sectored plates were often found in failure layers, suggesting they are especially prone to collapse.
Birkeland, K.W.. 2012. Crust thoughts. The Avalanche Review 30(3), 1,20.
Extended Column Tests (ECTs) have become increasingly popular for assessing snowpack stability. However, we still do not fully understand what happens to the block and the underlying weak layer during the test. Such work has been done for the Propagation Saw Test (PST), but not for the ECT. In order to better understand and interpret ECT results, we analyzed high-speed video (240 frames/second) of propagating ECTs using particle tracking. Our results show several things: 1) fractures initiate in an area of the weak layer directly under the shovel at the free edge of the block, 2) at the resolution of our measurements we do not see signs of progressive damage accumulation in the weak layer during tapping, but rather a single rapid collapse when fracture is initiated, 3) tapping on one side of the ECT does not affect the far side of the ECT, and 4) measured fracture speeds are similar to previously reported values, including those measured using the PST. From a practical perspective, our results suggest that the ECT is indeed measuring the propensity of a crack to propagate at the small scale of the ECT block, giving us greater confidence that we are capturing a critically important property of snow stability in our tests.
Professionals and recreationists utilize stability tests to assess snow stability. Our goal is to determine whether or not people are changing the types of tests they conduct. We utilized the SnowPilot database of over 3,600 snow pits from nine winters, with about 83% of these pits being dug by avalanche professionals. We found a dramatic shift in the tests conducted since 2004. SnowPilot users have moved away from rutschblocks and stuffblocks and moved more toward extended column tests (ECTs) and propagation saw tests (PSTs), while still conducting a large number of compression tests (CTs). ECTs are now the most popular test, being conducted in nearly 80% of all pits. Not surprisingly, this shift toward ECTs and PSTs has coincided with an increasing emphasis on the importance of propagation potential in our stability assessments. As we learn more about snow and the way it fractures, newer and more effective tests might well be advanced. Our results demonstrate that our community will quickly adopt new tests when they are useful and scientifically validated.
Understanding what controls coarse scale snowpack properties, such as surface hoar distribution, is imperative for predicting snow avalanches. Due in part to the inherent difficulties of winter travel in mountainous terrain, most spatial variability investigations of snow properties have been limited to relatively fine scales. To quantify snow surface spatial variability at the basin, region, and mountain range scales, a team of heli-skiing guides collected data throughout four major surface hoar formation periods over two heli-skiing seasons in rugged alpine terrain near Haines, Alaska across an extent of nearly 60km. Geographically weighted regression revealed a positive relationship between elevation and surface hoar crystal size with adjusted R2 values averaging near 0.40. Geostatistical analysis yielded spherical semivariogram autocorrelation ranges from approximately 3-25km, which is similar in size to many of the basins and regions within the study area. Kriging models built from the semivariograms were produced to aid geographic visualization of coarse scale snowpack processes. The results of this research suggest it may be possible to identify areas with greater surface hoar growth and persistence potentials as a consequence of synoptic onshore or offshore flow, and glacially influenced katabatic winds. These results can help in future efforts to forecast snow stability patterns over large areas.
Experience suggests that shallow, steep zones on slopes are likely spots for artificially triggering slab avalanches. However, a scientific understanding of this observation is not well quantified. We performed 108 point stability tests on a 30 x 30 m slope in central Svalbard. The slope has a rugged underlying topography and frequent wind influence both by top and cross loading. We found three persistent weak layers at different depths in the snowpack. Due to the rough nature of the study slope, snow surface does not resemble the ground topography. Weak layers forming early in the season follow the ground topography closely. As snow depth increases, the influence of ground topography diminishes. We further found a decrease in slab thickness with increasing slope and bed surface inclination. We therefore investigated the influence of slab thickness on slope stability. Our data shows that stability decreased significantly with decreasing slab thickness, which correlates to how deeply the weak layer is buried. Thus the weakest spots on the slope coincide with the shallowest and steepest spots, where the deeper buried weak layers are buried “less deep”. Such spots often occur around topographic heights such as large rocks, which are thus potential trigger zones.
We have developed a robust meteorological tower for deployment in locations with extreme conditions and for applications that require relatively maintenance-free structures. The basic design consists of a triangular base with two horizontal rails on each side, and uprights at the triangle vertices for various instrument configurations. The fabrication materials include 6061-T6 aluminum pipe (schedule 40 or 80), and cast aluminum connectors. The design is self-supporting, but may be guyed. Advantages of the design compared to conventional towers include easy assembly, readily available materials, easy adjustment for rugged location or terrain, transportable components by backpack, helicopter, boat, snowmobile, ATV or alternate method, and no need for concrete. Like all structures, the design is vulnerable to snow creep in deep-snow environs, but judicious site selection will mitigate this issue. We deployed these stations with minor variations in design in Alaska and Colorado, USA and on Baffin Island, Nunavut, Canada. The towers have survived relatively severe conditions in terms of cold temperature, wind, rime, and snow deposition. Locations have included subalpine, alpine, and arctic maritime environments. Regular data transmission has been achieved via satellite modem and satellite transmitter. We discuss design and installation.
During this past winter, southwest Montana had many large avalanches and days of high avalanche danger. In the Bridger Mountains the most prominent and active weak layer was a 30-55cm thick layer of depth hoar. This layer developed in November 2011 and persisted throughout the season. Precipitation was below average between December and early February, but each storm consistently produced artificially triggered avalanches on this layer. Higher snowfall rates and above average SWE in late February and March produced natural avalanche activity in the backcountry on the depth hoar. In late March 2012 a near isothermal snowpack, combined with a period of above freezing temperatures and heavy snowfall, produced an historic skier triggered and explosive controlled full depth wet slab cycle on the depth hoar. This event occurred on in-bounds terrain that was closed at the time, and had previously been heavily skied and controlled with explosives. We will review meteorological and snowpack factors that were associated with avalanche activity on the depth hoar within Bridger Bowl ski area and in the surrounding backcountry. Weather data from stations at Bridger Bowl and snowpack observations taken throughout the ski area will be used to discuss the factors associated with the timing of avalanche activity on the depth hoar layer. Patterns, trends, and outcomes will be highlighted which may be of wider value to the industry in managing these types of instabilities in future years.
Weak snow of interest to avalanche forecasting often forms and changes as thin layers. Thermometers, the current field technology for measuring the temperature gradients across such layers – and for thus estimating the expected vapour flux and future type of crystal metamorphism – are difficult to use at distances shorter than 1 cm. In contrast, a thermal imager can provide thousands of simultaneous temperature measurements across small distances with better accuracy. However, a thermal imager only senses the exposed surface, complicating its methods for access and accuracy of buried temperatures. This paper presents methods for exposing buried layers on pit walls and using a thermal imager to measure temperatures on these walls, correct for lens effects with snow, adjust temperature gradients, adjust time exposed, and calculate temperature gradients over millimetre distances. We find lens error on temperature gradients to be on the order of 0.03 °C between image centre and corners. We find temperature gradient change over time to usually decrease – as expected with atmospheric equalization as a strong effect. Case studies including thermal images and visual macro photographs of crystals, collected during the 2010–2011 winter, demonstrate large temperature differences over millimetre-scale distances that are consistent with observed kinetic metamorphism. Further study is needed to use absolute temperatures independently of supporting gradient data.
Avalanche researchers and practitioners have long measured snowpack temperatures in snow pits with thermometers about 10 cm apart. This led to the assumptions that temperature gradients are smooth and that temperature changes are regular. For this study, we used a thermal imager in standard snow pits in the Canadian Rocky Mountains during two seasons between 2010 and 2012. We collected the first season of data in a very shallow, below treeline snowpack study plot, and the second season of data in a deeper, treeline study plot. Data included thousands of thermal images, as well as visual macro images of the snow crystals in each pit layer to monitor changes.
We observed strong temperature gradients on the scale of individual snow crystals. We found that these small scale gradients correlated with future snow crystal changes. We also found that these gradients changed quickly with the weather, even at depth. This paper focuses on our most recent findings from the 2011-12 season, and describes our overall progress in extracting data from thermal images to use for research and forecasting. We use correlations to present very general relationships between thermal data, crystal size, and layer stability tests. We also present temperature and gradient changes at depth during a period of clearing.
Shea, C., B. Jamieson, and K.W. Birkeland. 2012. The life of a shallow snowpack. Poster showing the temporal changes in the thermal evolution of a thin snowpack.
Field observations suggest that avalanches releasing on non-persistent weak layers tend to slide on steeper slopes than other dry slab avalanches. Recent research indicates that the initiation of fracture in a weak snowpack layer is almost unaffected by slope angle. However, this field research included only persistent weak layers. Thus, the question remains: Why are avalanches involving non-persistent weak layers slope angle sensitive? Our paper investigates the following for non-persistent weak layers: 1) the effect of slope angle on fracture initiation and 2) the friction coefficient. We performed Extended Column Tests (ECTs) on slopes where we could track the number of taps needed for weak layer fracturing on a variety of slope angles (from 8° to 46°) with minimal snow structure change. Our results show that the number of taps required to initiate a fracture that crosses the entire column (ECTP) was mostly independent of slope angle. In addition, we measured the friction coefficient of two different non-persistent weak layers types (PPsd and DFdc) within minutes after avalanches were triggered and in the following days. Our results show that the friction coefficient of non-persistent weak layers was higher than published values for persistent weak layers. From a theoretical perspective, our results are in line with the mixed-mode anticrack model for fracture in a weak snowpack layer. From a practical perspective, our results can contribute to safer pit site selection, terrain and snowpack analysis, and they give us insight into the value of observing downslope block movement after fracturing in stability tests.
Depth hoar is a persistent weak layer that is a common instability problem in the snowpack of many areas around the world. Boot packing or compaction via explosives is a technique that is widely used in an attempt to disrupt this weak layer and increase its variability across a slope, thereby increasing overall slope stability. While some data have been gathered on the results of boot packing and explosive use on slopes, no recent work has concentrated directly on the effect of boot packing on layer density, hand hardness and stability test scores. Therefore, we devised an experiment to test the changes in these metrics directly on a side-by-side boot-packed and non-boot packed slope.
A 50m x 25m, relatively uniform slope was split into two equal areas with one being extensively boot packed, while the other remained undisturbed. Observations were made in both boot packed and non-boot packed plots until both areas were unreactive with respect to results using Extended Column Tests (ECT) on this basal layer. Our results show that density and hand hardness increased in the boot packed area in comparison to the non-boot packed area. Furthermore, in the boot packed area we also observed a marked increase in ECT scores and a change in fracture character of the basal layers. In the non-boot packed area our ECTs propagated (ECTPs) for a full nine weeks longer than in the boot packed area. While this is just one case study, we are encouraged by our results, which provide quantitative data that indicate that boot packing can be an effective tool for helping to stabilize persistent basal weak layers.
Explosives are a critically important component of avalanche control programs. They are used both to initiate avalanches and to test snowpack instability by ski areas, highway departments and other avalanche programs around the world. Current understanding of the effects of explosives on snow is mainly limited to shock wave behavior exhibited through stress wave velocities, pressures and attenuation. This study aims to enhance current knowledge of how explosives physically alter snow by providing practical, field-based observations and analyses that quantify the effect of explosives on snow density and snow stability test results. Density and stability test results were evaluated both before and after the application of 0.9 kg cast pentolite boosters as air blasts. Changes in these properties were evaluated at specified distances up to 4 meters (m) from the blast center using a density gauge and Compression Tests (CTs). Statistically significant density increases occurred out to a distance of 1.5 m from the blast center and down to a depth of 60 centimeters (cm). Statistically significant density increases were also observed at the surface (down to 20 cm) out to a distance of 4 m. Results from CTs showed a decrease in the number of taps needed for column failure in the post explosive tests. The results of this study provide a better understanding of the physical changes in snow following explosives, which may lead to more effective and efficient avalanche risk mitigation.
Nonpersistent snow crystals are deposited during storms, last for a few days, and are common fracture planes in avalanches. Because avalanches on persistent crystals cause more deaths, avalanches on nonpersistent crystals receive less study despite being responsible for the majority of avalanche fatalities in some U.S. states such as California. Avalanches are studied here using thousands of avalanche control records, dozens of crown face profiles, and field work with near-infrared (nIR) imaging and particle tracking. New (24-hr) snow and Extended Column Test propagation (ECTP) are the best predictors of avalanche activity on nonpersistent crystals, while changes in temperature and density during storms have no effect. Failure planes with nonpersistent crystals are the same hardness as layers immediately above, but are significantly softer than layers immediately below. ix In one nIR profile, nonpersistent crystals form a thin weak layer, but in others, there is no grain size difference between the fracture plane and adjacent strata. Lack of a weak layer makes fracture in nonpersistent crystals similar to other materials; fractures cut through the bulk without any specific weakness other than a flaw. Failures in nonpersistent crystals follow behavior predicted by the anticrack model, a model emphasizing collapse that has only been tested on failures of persistent crystals. This suggests nonpersistent and persistent crystals share the same failure mechanism. Critical cut lengths during Propagation Saw Tests (PSTs) vary little with slope angle, consistent with anticrack predictions. A PST on nonpersistent crystals collapsed about 1 mm, half the collapse of a PST on persistent crystals. Easily collapsible planar crystals such as sectored plates and stellars are the most common nonpersistent crystals in failure planes. During high instability, fractures are easier to trigger on nonpersistent than on persistent crystals. Sustainability of propagation depends on new snow depth because the slab must withstand bending; otherwise tensile slab fracture occurs. Measurement of displacement during PSTs yield some of the lowest elastic moduli recorded for snow: 0.04-0.41 MPa for slabs with densities from 73-143 kg m-3 . The median fracture energy for nonpersistent crystals is 0.09 J m-2 , comparable to published values for persistent crystals.
More winter recreationists are venturing into steep avalanche chutes and “extreme” terrain each year, and avalanche fatalities are increasing. The slope-scale spatial variability of weak layers and slabs and how it relates to this complex terrain is of critical importance but poorly understood. In this study, I use terrain parameters to model potential trigger locations (PTLs) of slab avalanches, which are defined based on slab thicknesses and presence of weak layers.
In a sample couloirs and chutes in Montana and Wyoming, field teams tracked and mapped persistent weak layers and slabs with probe sampling. Terrain parameters derived from a one meter DEM were used to explore the relationships between PTLs and terrain. Exploratory analysis, multi-model classification trees, and logistic regression models support strong relationships between terrain and PTLs.
Modeling of PTLs was highly successful for individual couloirs, with terrain-based model success rates frequently exceeding 70% for depth hoar PTLs and 85% for near-surface weak layers. However, models varied widely from couloir to couloir, with generally poor cross-validation results between couloirs, suggesting that the relationships between terrain and PTLs in each couloir are unique and highly complex. For these 21 couloirs in steep alpine terrain, parameters relating to wind deposition and scouring have the strongest association with PTLs.. Parameters with the greatest ability to discriminate PTLs are distance from the edge of a couloir, relative elevation, degree of wind exposure, and degree of terrain exposure. The influences of these and other terrain parameters vary, depending on broader-scale terrain characteristics, prior weather patterns, and seasonal trends.
Practical implications from this study are numerous. With an understanding of the broader scale influences and physical processes involved, we can use terrain to optimize stability test locations, explosive placements, or route selection. The unique nature of each couloir means that simple rules relating terrain to PTLs will not apply, although couloirs in the same cirque generally share similarities. This work increases our understanding of how each parameter relates to the physical processes causing PTLs and how these relationships can vary. This information will help to improve practical decision-making ability as well as future modeling efforts.
Skiers caught in a slab avalanche often trigger the avalanche themselves. Preventing those accidents necessitates a better understanding of the factors contributing to the failure of the snowpack under the action of a skier. In the present work, a mathematical model based on the principles of mixed-mode anticracking is proposed for skier triggering. The respective influences of the slope-normal and slope-parallel components of the load exerted by a skier on the prospective fracture plane are taken into account. A criterion for fracture propagation under typical skier loads is derived. It manifests a small number of factors that, combined, multiply the risk of triggering an avalanche. The criterion indicates, contrary to a common perception, that fracture is not more difficult to trigger in gentle slopes than in steep slopes. The major result of the model is confirmed by data obtained from field experiments.
Avalanche hazard evaluation relies in part on representative snowpack stability observations. Thus, understanding the spatial patterns of snowpack instabilities and their environmental determinants is crucial. This case study integrates intensive field observations with spatial modeling to identify associations between incoming radiation, surface hoar development and its subsequent shear strength across an inclined forest opening. We examined a buried surface hoar layer in southwest Montana, USA, over five sampling days, collecting 824 SnowMicroPen resistance profiles and performing 352 shear frame tests. Spatial models of incoming long- and shortwave radiation were generated for the surface hoar formation period using modeled hemispheric sky visibility, physically based parameters and the Bird Clear Sky Radiation Model in a Geographic Information System. Before burial, the surface hoar persisted despite moderate winds and relatively high air temperatures. The buried surface hoar layer thickness varied between 3 and 21mm within a distance of 30 m. Modeled incoming radiation explained spatial variations in layer thickness and shear strength. In areas exposed to large amounts of radiation, the surface hoar layer was strong and thin, while areas with limited incoming radiation (due to high sky visibility and shading) possessed a thicker surface hoar layer that sheared more easily. This demonstrates the usefulness of microclimate modeling for slope-scale avalanche hazard evaluation. We also identify that over the 3 week sample period, strengthening occurred without thinning of the surface hoar layer.
This thesis presents four applied methods for seasonal snow observation with respect to avalanches. Previous avalanche-related spatial variation and scale studies have shown a clear need for observation and methods to focus on the scale of interest to human triggering. These methods have the common goal to reveal spatial variation of interest to avalanche formation and human triggering in an efficient, accessible manner.
The four methods are: (1) A minimally destructive slope-scale sampling method, (2) A method to relate Google Earth terrain images to surface hoar formation in sparse trees, (3) A method of accessibly presenting complex GIS warming model data over real terrain, and (4) A method of measuring heat in the snowpack using a thermal imager. Despite their common goal of spatial visualization, each new method draws on a different subset of background literature and employs very different methods in development and use. Thus, each method is presented as a self-contained paper with independent results. Of note, these methods have all subsequently received active use, and conclusions from such use are discussed at the end of the thesis.
Despite its fundamental importance, crown depth is often treated as a scalar rather than a distributed variable in avalanche run-out and fracture models. To date, no studies have examined the distribution of depth across crown transects. We present results on geometry, depth distribution, and spatial correlation for transects along the crowns of small to large avalanches. Crown heights are fit well by normal or Weibull distributions and are spatially correlated. Transects are thinnest and decreasing toward the flanks, perhaps associated with fracture arrest. Underlying and adjacent terrain seems to have little influence on our transects. Instead, wind transported snow and upwind features play a dominant role. We suggest deposition of new snow by wind is a Gaussian process that drives transect shape. Comparing strength distributions and coefficients of variation from previous crown face studies, we suggest differences in overburden drive distributions of strength measured across crown faces.
Assessing snow stability requires a holistic approach, relying on avalanche, snowpack and weather observations. Part of this assessment utilizes stability tests, but these tests can be unreliable due in part to the spatial variability of test results. Conducting more than one test can help to mitigate this uncertainty, though it is unclear how far apart to space tests to optimize our assessments. To address this issue we analyze the probability of sampling two relatively strong test results over 25 spatial datasets collected using a variety of stability tests. Our results show that the optimal distance for spacing stability tests varies by dataset, even when taking the sampling scheme and stability-test type into account. This suggests that no clear rule currently exists for spacing stability tests. Our work further emphasizes the spatial complexity of snow stability measurements, and the need for holistic stability assessments where stability tests are only one part of a multifaceted puzzle.
Avalanche practitioners often assume that weak snowpack layers fracture more easily in steeper terrain. We are typically reluctant to conduct stability tests in safer, gentler terrain because we believe the results may not be reliably extrapolated to steeper areas. However, recent fracture mechanics research as well as propagation saw testing suggest that increasing slope angle has a small effect on fracture initiation. This paper investigates the effect of changes in slope angle on extended column test (ECT) results. We conducted ECTs on slopes in Montana and Alaska where gradual changes in slope angle allow us to sample a variety of angles with minimal changes in snow structure. The slope angles in the Montana datasets range from 7° to 35°, while those in the Alaska datasets range from 30° to 44°. On all slopes the weak layer consisted of buried surface hoar. The results show that the number of ECT taps required to initiate fracture that propagates across the entire column (ECTP) either did not change or increased slightly as the slope angle increased. From a theoretical perspective, this result is in line with the predictions of the mixed-mode anticrack model of fracture propagation in snow. From a practical perspective, our results suggest that, as long as the snow structure remains reasonably consistent in space, observers can conduct dependable tests in gentler, safer terrain before committing themselves to more exposed areas.
Understanding the spatial distribution of weak layers is a significant challenge for avalanche forecasters. Thus, improving our understanding of the processes that dictate the formation and persistence of surface weak layers across large areas is critically important for improving backcountry avalanche forecasting accuracy. For this work, heli-skiing guides mapped unburied surface hoar and near-surface facets across the Chilkat and Takhinsha Mountains of southeast Alaska during two major formation events of the 2010 season. After burial, we monitored weak layer persistence and avalanche activity. Our study area encompasses 900 km2 of rugged, glaciated, alpine terrain, at elevations ranging from 300 m to 2000 m. Guides collected information on crystal attributes and terrain characteristics at each location, and used handheld GPS units to reference locations. Incorporating the data into a Geographic Information System (GIS) proved to be invaluable for managing and visualizing observations. For example, the GIS allowed the creation of surface condition maps that we used operationally in guide meetings and in the helicopter for making run decisions. Data analyses quantified observed patterns. In particular, we often found that surface hoar crystal size lessened with decreasing elevation, possibly due to stronger katabatic winds in valley bottoms. Additionally, we observed areas of inhibited surface hoar formation and persistence which may have resulted from the influence of synoptic inflow and outflow drainage winds. By better understanding the distribution of surface hoar and near-surface facets, this work provides insights into improved backcountry forecasting of avalanche conditions over large areas.
On February 16th, 2010 a person accidentally kicked off a cornice that triggered a large slide on Saddle Peak immediately outside Bridger Bowl’s boundary. This slide narrowly missed killing several people, and the immense snow cloud was seen throughout the ski area. While the avalanche surprised many, avalanche professionals at the Gallatin National Forest Avalanche Center (GNFAC) and Bridger Bowl Ski Area have anticipated such an event since this sidecountry area was opened up two years ago. We have tried to educate the public about the accident potential through signage, articles, a special DVD, and education efforts targeting youth and their parents. Despite our best efforts Saddle Peak is heavily skied by all ages, with multiple people on the slope at the same time and with no regards for, and apparent knowledge of, safe backcountry travel techniques. Moreover, the slide demonstrated the fallacy of skier compaction for preventing large hard slab avalanches. In December 2009 a layer of facets was buried by a supportable hard slab impervious to the thousands of skiers who skied the slope during the season. A mid-February storm dropped over three inches of SWE, setting up the conditions leading to the slide.
Snowfall, temperature and wind are three factors that quickly change avalanche conditions. Ridge-top winds have been used to assess avalanche conditions with mixed success due to high variability. Few analyses have tested the effect upper atmospheric winds have on avalanche conditions. This study attempts to quantify the effect upper atmospheric wind direction and speed have on the spatial pattern of natural avalanching in the mountains near Gothic, Colorado. The Gothic dataset consists of over 3,300 natural avalanches spanning a time period of 33 years. These data are associated with daily new snow water equivalence (SWE) collected at Gothic, as well as 500mb wind direction and speed from the National Centers for Environmental Protection/National Center for Atmospheric Research (NCEP/NCAR) gridded reanalysis project. I hypothesize that prevailing 500mb winds are more likely to be associated with natural avalanches in avalanche paths with a starting zone aspect lee to the prevailing free air winds. In addition, I hypothesize that the odds of a natural avalanche occurring increases with increasing new SWE, increasing 500mb wind speeds, and 500mb wind direction. These hypotheses are tested using simple probability analysis as well as a two-component hurdle model.
As expected, avalanche paths lee to the 500mb wind direction have an increased probability of avalanching in relation to paths facing in other directions. However, exceptions do occur, some of which can be explained by cross-loading. The hurdle model results indicate that after accounting for new SWE, 500mb wind speed is significant in determining whether or not a day will be considered an avalanche day. Once a day is determined to be an avalanche day, 500mb wind direction is an important determinant for the daily avalanche hazard after accounting for new SWE. These results have practical significance. They give avalanche forecasters confidence that upper air wind direction is a useful predictor of the pattern of avalanche activity at the valley or mountain range scale, and the exceptions observed show that such predictions cannot be applied at the path scale. These scale issues demonstrate the general nature of backcountry advisories and why they cannot be applied at the scale of individual slopes.
In the fall of 2009 I became interested in when offices are staffed by forecasters at the various centers and what time of day products are issued. In other words are the hours inhumane at all the centers? How much do product issuance times vary between the centers? I decided to survey the various centers regarding this and a few other topics. Later I learned of a similar survey compiled by Knox Williams in 1998. So I added some questions to my survey in order to update his work. My survey turned out to have more questions than I originally imagined; but later I thought of more questions I could have asked.
So the results of this survey include various topics such as administration, pay, forecast area size, budget, hours of operation, weather stations, weather and avalanche product information, dissemination methods, product issuance days and times, avalanche education, and length of seasonal employment. I hope the results are informative and useful to any interested persons.
I would like to emphasize that I did not undertake this survey in order to show any particular results. Rather the work was undertaken out of interest in similarities and differences between the various centers.
Understanding the spatial variability of the snowpack is critical for avalanche prediction and mitigation. Previous spatial variability research focused primarily on relatively low angle slopes, many of which had fairly uniform characteristics. With snow sports progressing to steeper and more complicated terrain, a need exists to better understand the relationship between this “extreme” terrain and the snowpack variability. This research utilizes nine couloirs from Big Sky, Montana and Teton Pass, Wyoming. We used a probe to measure weak layer heights, slab thicknesses, and snow depths, we cross-verified those measurements with pits, and we georeferenced our sampling points using a sub-meter accuracy GPS. LiDAR data are used to derive terrain parameters, such as slope, aspect, elevation, and curvature in a GIS, and these data are then compared with our snow observations. Our analyses quantify the distribution of snow in the couloirs, and suggest that weak layer thickness normalized by snow depth is significantly correlated with the distance from the windward boundary when all other terrain parameters are accounted for in our sampled population of couloirs. Our results provide insights into the distribution of weak layers and snow depth in steep couloirs, which is a first step in optimizing snow pit locations and explosive placements in this terrain.
Most skiers trigger the slab avalanche in which they are caught. Preventing those accidents necessitates a better understanding of the factors contributing to the failure of the snowpack under the action of a skier. Our contribution gives a number of insights into a newly developed mathematical model of skier triggering based on the principles of mixed-mode anticracking. We give various examples of how the direction of the applied force and the penetration depth of the skis influence the chances of triggering fracture in the weak layer, and we investigate how the skier’s stance influences that risk. We also ask how the critical loads for triggering fracture depend on slope angle in general. We find that, for weak layers prone to anticracking, fracture is not easier to trigger on steep slopes, but is equally difficult or marginally more difficult to trigger the steeper the slope. We carried out extensive field experimentation to test this proposition using the extended column test method, the detailed results of which are given in a companion paper. As usual, our formulation includes simple shear cracking as a limiting case, so that the anticrack model is well-suited to emphasize the differences between the two fracture mechanisms. The results emphasize that the anticrack mechanism for fracture in snow requires scientists and practitioners to rethink previously accepted – and practically relevant – paradigms.
The number of skier triggered dry slab avalanches typically increases during or shortly after snow loading events. However, field observations and research suggest that a skier is less likely to trigger a weak layer fracture as the depth to that weak layer increases. This begs the question: Why does skier triggered avalanche activity increase when the likelihood of initiating fracture seemingly decreases? This paper presents preliminary evidence that new snow loading may decrease the chances for fracture arrest once initiated. During the winter of 08/09 in Colorado’s continental snowpack we used Extended Column Tests (ECT) and Propagation Saw Tests (PST) to track changes in the snowpack’s ability to propagate facture before and after loading events. In addition, we present two case studies from Southeast Alaska’s maritime snowpack. We used Extended Column Tests to measure the amount of additional loading required for a fracture to cross the entire column (ECTP). We compared these measurements to the natural loading at the end of the loading event and a day after the precipitation stopped. We also compared our data to avalanche activity on the same slopes. Our results suggest that in some cases the snowpack’s propensity for fracture arrest decreased with the additional loading, and that artificial loading of an extended column may be a useful tool to estimate loading thresholds for full fracture propagation.
Glide avalanches are a significant hazard that threatens people and property in many snowy climates. They are hard to control, poorly understood, and extremely challenging to forecast. This paper presents meteorological and environmental data associated with three glide avalanche cycles. It also discusses hazard reduction techniques from an operational perspective and provides possible explanations why previous attempts to artificially trigger glide avalanches rarely succeed. During Southeast Alaska’s winter of 09/10, we witnessed three glide avalanche cycles with over 35 total avalanches. During those cycles we collected data on snowpack, precipitation, temperature, relative humidity, sky coverage and streamflow, as well as slope aspect, elevation, steepness, shape and ground cover. We also recorded visual snow surface observations leading to the transition of some of the glide cracks to avalanches. Although glide avalanche activity is clearly somehow related to atmospheric events, we found no direct correlation between meteorological data and avalanche occurrences. However, we did find a rough correlation between snowpack, terrain and avalanche time distribution in two out of the three cycles. Our lack of reliable forecasting and control tools for glide avalanches implies that limiting the potential destructive size of glide avalanches throughout the entire winter may be the most effective approach to managing the hazard for some operations.
Statham, G., P. Haegeli, K.W. Birkeland, E. Greene, C. Israelson, B. Tremper, C. Stethem, B. McMahon, B. White, and J. Kelly. 2010. The North American avalanche danger scale. Proceedings of the 2010 International Snow Science Workshop, Squaw Valley, California.
The Avalanche Danger Scale is an ordinal, five-level warning system that is a cornerstone of public avalanche information. The system was developed in Europe in 1993, and introduced to North America in 1994. Although both Canada and the United States adopted the system, different descriptors of the danger levels were developed in each country. Fifteen years of practical use revealed numerous deficiencies in this danger scale, most notably a lack of clarity during low probability/high consequence avalanche conditions. In 2005, a group of Canadian and American avalanche forecasters and researchers began to revise the system, with the goal of improving clarity and developing a single standard for North America. Initial explorations to define the problem resulted in more questions and uncovered an almost complete absence of formal underpinnings for the danger scale. The magnitude of the project subsequently changed, and in 2007 the project objectives were clarified as: 1) definitions of avalanche hazard, danger and risk; 2) methodology for assessing avalanche danger; and 3) revisions to the danger scale as a public communication tool. This paper concentrates on the third and final objective, and describes the methods and results of producing the North American Public Avalanche Danger Scale. Emphasis is placed on best practice in warning system design and the principles of risk communication, which helped reshape the avalanche danger scale into a more effective communication tool. The revised danger scale will be implemented across Canada and the United States for the 2010/11 season.
Statham, G., P. Haegeli, K.W. Birkeland, E. Greene, C. Israelson, B. Tremper, C. Stethem, B. McMahon, B. White, and J. Kelly. 2010. A conceptual model of avalanche hazard. Proceedings of the 2010 International Snow Science Workshop, Squaw Valley, California.
Conventional avalanche forecasting is practiced as a mix of deterministic treatment for snow and weather parameters and inductive logic to reach actual forecast decisions (LaChapelle, 1980). Avalanche forecasters subjectively integrate a complex array of data and evidence to reach their decisions, often operating under a high degree of uncertainty. Spatial scales in avalanche forecasting are widely variable, ranging from slope specific predictions to large, regional areas characterized by significant variation across space, and over time. Thus, forecasters must synthesize the available evidence, and extrapolate this across the landscape by relying on their knowledge of the terrain.
We present results on the geometry and depth distributions for transects along the crowns of small to large avalanches. Previous research suggests that power laws in aggregate distributions of crown depths result from observers recording maximum depth. We therefore studied individual crown transects to obtain information about the parent distributions, using in situ and photogrammetric measurements. The geometry shows that crowns are thinnest and decreasing toward the flanks. Depth distributions of individual crowns do not follow power laws. Instead, we find transects fit normal distributions. To reconcile this finding with observed aggregate distributions, we propose a mixture of maxima, taken from normally distributed crown depths, with power law mean parameters. We suggest normally distributed transects arise from spatially-correlated new snow depths.
Understanding the spatial variability in fracture initiation and fracture propagation is critical for avalanching as both are required for an avalanche to release. Most of the previous research looked at the spatial variability of fracture initiation. We focus on understanding the spatial variability of the fracture propagation potential using the Extended Column Test (ECT). This work uses a new overlapping grid methodology which allowed us to make repeat data collection on the same slope to collect data on two separate days at the slope scale from two environmentally different sites (windy and sheltered), thereby capturing temporal changes in the spatial variability of our results. In contrast to previous fracture propagation test research, our data demonstrates considerable spatial variability in fracture propagation potential. Interestingly, at both the windy and sheltered sites the first sampling day demonstrated a relatively random distribution of fracture propagation potential results, while the second sampling day for both sites showed evidence of increased resistance to propagation as well as increased spatial clustering at the scale of our observations. Since distinct clustering or pockets of propagation and non propagation exist on some slopes, the practical implication of our work is that it is often necessary to dig more than one snow pit on suspect slopes to assess stability, and those slopes might be more accurately assessed by widely (greater than 10 m) spaced measurements. Though our data are limited, these results represent the first statistically demonstrated temporal change in snowpack spatial variability at the slope scale. However, in order to definitively address the question of temporal changes in spatial patterns, much more work is needed on many slopes with varying weak layers and snowpack conditions.
Avalanche hazard evaluation relies on snowpack stability observations. Because snowpack properties can vary extensively over time and space, estimating slope-scale stability is difficult. This study addressed these challenges by implementing a methodology that 1. quantified spatial and temporal patterns of snowpack stability, 2. identified spatial associations between the strength and stability of a weak layer and slab load, and radiation properties, 3. identified internal associations between weak layer thickness, shear strength, microstructural properties, and slab load.
An instability associated with a buried surface hoar weak layer was examined on an inclined forest opening at Lionhead, southwest Montana, during February and March, 2005. During five sampling days, 824 snow depth and SnowMicroPen resistance profiles and 352 shear frame tests were performed. An objective texture-based stratigraphic sampling approach was developed to obtain microstructural estimates of a stratigraphic weakness and instability from SnowMicroPen profiles, utilizing the coefficient of variation of rupture force. Spatial models of hemispheric sky visibility, and incoming long- and shortwave radiation were generated for the surface hoar formation period using a Geographic Information System and independent optical observations.
Despite relative topographic uniformity, in a distance of 30 m, the buried surface hoar weak layer thicknesses varied between 3 – 21 mm. Before burial, the surface hoar persisted despite moderate winds and above freezing air temperatures. Spatial patterns of modeled incoming longwave and shortwave radiation explained the large variation in weak layer thickness and strength properties. Areas exposed to large amounts of radiation contained a strong, thin buried surface hoar layer, while in areas with limited incoming longwave (due to high sky visibility) and shortwave radiation (due to shading), the layer was thicker and possessed low shear and microstructural strengths.
Over time, the shear frame stability index and SnowMicroPen-derived microstrength of the surface hoar layer increased and values became spatially more variable (divergence): it became harder to predict stability as the snowpack became more stable. A loading event then decreased stability and micro-strength and caused spatial uniformity (convergence), thereby increasing predictive strength of observations. The findings illustrate the usefulness of the SnowMicroPen for evaluating spatial patterns and load related changes in snowpack stability.
Researchers and practitioners have long utilized a variety of penetrometers to investigate the snowpack. Identifying definitive relationships between penetrometer-derived microstructural information and stability has been challenging. The purpose of this study is two-fold: 1. We propose a simple field test to establish relationships between load and penetrometer-derived microstructural estimates, 2. We utilize the SnowMicroPen (SMP) to quantify changes in weak layer residual strength and microstructural dimension associated with an artificial loading event. Our dataset is from Moonlight Basin, Montana and includes three modified loaded-column tests, each paired with 5 SMP profiles. Depth hoar comprised the targeted weak layer. Results indicate that loading caused the residual strength and rupture frequency to decrease significantly. Much like a compression test at a micro-scale, the force required for the SMP to rupture individual structures as well as the micro-scale strength decreased significantly when the slab stress was increased by artificially adding blocks of snow. A decrease in observed rupture frequency within the weak layer (or an increase in the distance between ruptured structures) also occurred after the loading event, probably because some structures within the weak layer had already failed or were so close to failing that the penetrometer could not detect their rupture. Due in part to the large difference in loads, microstructural differences between the natural and loaded columns were significant enough that only one profile would have been necessary to determine a significant difference in residual strength. Artificial removal of slab stress resulted in greater rupture forces and larger microstructures, likely due to elastic rebound.
Wet snow avalanches are dangerous and can be particularly difficult to predict. The rate of change from safe snow conditions to dangerous snow conditions occurs rapidly in a wet snowpack, often in response to water production and movement. This research focused on the relationship between snow stratigraphy and water movement in an inclined snowpack. Concentrating on transitions that impede water and flow finger formation within the snowpack, dye tracer was mixed with water and applied to a stratified snowpack to observe and measure the movement of water in various snow grain types, sizes, densities, and temperatures. There were two types of layer transitions that impeded water. Water was impeded at capillary boundaries caused by fine grains over coarse grains. It was also impeded at hydraulic conductivity boundaries, such as ice layers. In layer transitions that impeded water, the grain size of the layer above was significantly smaller than the layer below. The layer above a transition that impeded water was also significantly less dense than the layer below the transition. A qualitative analysis of grain type showed that there was no relationship between grain types in the layer above or below a transition and whether they will or will not impede water.
A SnowMicroPen (SMP) was used to measure changes in structural element length to identify capillary boundaries. Results from SMP measurements indicate that microstructural analysis of the snowpack aids in characterizing capillary boundaries that impede water flow. The step change, rate of change, and percent increase were significantly larger in capillary boundaries than transitions that did not impede water for the entire dataset from all 8 sessions. When all transitions were ranked according to absolute change for each profile, capillary boundaries consistently ranked in the top two of all transitions evident within each SMP profile.
The amount of water needed to produce flow fingers was highly variable. There was no significant relationship between the amount of water necessary to form flow fingers and snow density, snow grain size, snow temperature, or grain type. Layer transitions that impeded vertical water movement and flow finger formation may both play a large role in wet slab avalanche formation.
The Extended Column Test (ECT) is a new stability test that aims to assess the fracture propagation potential across a 0.90 m wide isolated column. This paper: 1) describes the test procedure and presents new recording standards for the test, 2) uses two independent datasets (each consisting of over 300 tests) to assess the effectiveness of the test, 3) looks at the spatial variability of ECT results from several test grids, and 4) compares adjacent results between the ECT and the Propagation Saw Test (PST) on stable and unstable slopes. Our results indicate that the ECT is an effective stability test, with a false-stability rate less than other standard snow stability tests. Results are sometimes quite spatially uniform, though occasionally slopes may exhibit variable ECT results. In comparison to the PST, our data suggest that the ECT has a lower false-stability rate, but a higher false instability rate. Overall, the ECT is better at discriminating between stable and unstable slopes in our dataset. No test is perfect and all tests must be used in conjunction with additional data, but our results show that the ECT is valuable additional tool for assessing snow stability.
The literature disagrees about the statistical distribution of snow avalanche crown depths. Large datasets from Mammoth Mountain, California and the Westwide Avalanche Network show that the three-parameter generalized extreme value distribution provides the most robust fit, followed by a two-parameter variation, the Fre´chet distribution. The most parsimonious explanation is neither self-organized criticality nor other complex cascades, but the maximum domain of attraction, implying that distributions of individual avalanche crown depths are scaling. We also show that crown depths do not have a universal tail index. Rather, they range from 2.8 to 4.6 over different avalanche paths, consistent with other geophysical phenomena such as wildfires, which show similar variability.
Bair, E.H., J. Dozier, and K.W. Birkeland. 2008. Avalanche crown depth distributions. Proceedings of the 2008 International Snow Science Workshop, Whistler, British Columbia, Canada.
The Extended Column Test (ECT) is a new stability test that aims to assess the fracture propagation potential across a 90 cm wide isolated column. Initial results with a dataset consisting of over 300 tests collected by one observer demonstrated the ECT’s effectiveness for differentiating between stable and unstable slopes. Further, we have received positive feedback on the test from a world-wide network of observers. This paper: 1) presents new recording standards for the test, 2) uses the SnowPilot dataset to further assess the effectiveness of the test by analyzing over 300 tests performed by several observers in different snow climates, 3) looks at the spatial variability of ECT results from several test grids, and 4) compares side-by-side results between the ECT and the Propagation Saw Test on stable and unstable slopes. Our results indicate that the ECT is an effective stability test, with a false stability ratio generally less than other standard snow stability tests. Results are sometimes quite spatially uniform, though occasionally slopes may exhibit variable ECT results. In comparison to the PST, our data suggest the ECT has a lower false stability rate, but a higher false instability rate. No test is perfect and all tests must be used in conjunction with additional data, but our results show the ECT is valuable additional tool for assessing snow stability.
Swimming in avalanches has recently been questioned, with detractors stating that “swimming leads to dying”. Since no direct scientific evidence exists to either refute or support the idea of swimming, we combine the practical experience of avalanche survivors with our emerging knowledge of avalanche dynamics to arrive at possible survival strategies for different parts of flowing avalanches. Practical experience and avalanche dynamics theory are largely consistent and suggest the following strategies: 1) Once an avalanche is released, every effort must be made to get off the moving slab, 2) After being caught, the victim must do everything possible to try to get toward the back, or tail, of the avalanche since this is where avalanches run out of mass and where a victim is more likely to be left behind by the slide, 3) Experience shows that in some avalanches a backstroking and log rolling motion may help the victim stay near the surface and move toward the flanks of the avalanche, and 4) If at all possible, the head of the avalanche should be avoided since the turbulent flow and large forces in this area increase the odds of injury and deep burial. Though it cannot be definitively proven, experience and avalanche dynamics theory suggest that swimming – or as some call it, “struggling” – is part of a viable strategy for surviving an avalanche once you are caught.
Understanding the spatial variability of the snowpack is a crucial step to improve accuracy in field data collection and avalanche forecasting. While there has already been a large volume of literature assessing the spatial variability of the snowpack, inconsistent sampling designs make comparing results difficult. This work uses an overlapping 10 by 10 m grid to collect Extended Column (ECT), Compression (CT) and Stuffblock (SB) test data at the slope scale across a range of environmental settings and climatic regimes in Montana and New Zealand. The overlapping grid methodology standardizes data collection between our sites, as well as allowing for repeat data collection on the same slope, thereby providing a new method for attempting to assess changes in spatial variability over time. Preliminary results suggest that the spatial variability of fracture propagation and fracture initiation may increase over time, and that the spatial variability of the fracture propagation propensity may be related to the processes causing the instability. As we collect more data, these results will provide further insight into the problem of snow pit location and representivity, both in terms of space and time.
Researchers and practitioners have long utilized a variety of penetrometers to investigate the snowpack. Identifying definitive relationships between penetrometer-derived microstructural information and stability has been challenging. The purpose of this study is two-fold: 1. We propose a simple field test that can be implemented by the scientific community to establish relationships between load and penetrometer-derived microstructural strength, 2. Utilizing the SnowMicroPen (SMP) data, we quantify changes in weak layer residual strength and structural dimensions associated with a loading event. Our dataset is from Moonlight Basin, Montana and includes three modified loaded-column tests, each paired with 5 SMP profiles. Depth hoar comprised the targeted weak layer. Results indicate that loading events cause the residual bond strength and bond frequency in large-grained weak layers to decrease significantly. Much like a compression test at a micro-scale, the force required for the SMP to rupture individual bonds as well as the micro-strength decrease significantly when the slab stress is increased by artificially adding blocks of snow. A decrease in observed bond frequency within the weak layer (or an increase in the distance between bonds) also occurs after a loading event, probably because some bonds within the weak layer have already failed or are so close to failing that the penetrometer cannot detect their rupture. Artificial removal of slab stress resulted in greater rupture forces and distances between bonds, likely due to elastic rebound. This indicates that long after a natural loading event has occurred, elastic deformation still exists within the weak layer.
Due to the time-consuming nature of traditional snowpit measurements, and the large spatial variability that often exists in alpine snowpacks, tools which can rapidly characterize snowpack properties are in great need. Microwave radar has an additional advantage in that it is non-destructive and measurements can be made remotely, providing the opportunity to make measurements over large areas rapidly from an airborne platform. Signal interpretation can be difficult, however recently several ground-based studies have shown that the technique can be used to accurately measure snow depth, snow water equivalent, and snow stratigraphy. Accurate measurements of these properties from the air is much more challenging, especially in steep terrain. We present results from two field campaigns near Valdez, Alaska, in which measurements in mountainous terrain were made from a helicopter with two different FMCW radar systems. Snow depth and stratigraphy was visible at altitudes of less than 100 feet, while steep terrain made interpretation difficult at typical flying altitudes, due to the footprint size of the radar. Refinements of helicopter-based radar measurements may eventually provide a useful tool to assist in stability evaluations for helicopter ski operators.
Understanding the spatial variability of snow at many different scales is critically important for avalanche forecasting. This research uses new techniques to reanalyze an existing spatial dataset collected in southwest Montana’s Bridger Range during the 1996-97 winter. Recent developments in statistical software have greatly increased the ease with which mixed effects and spatially correlated models can be run. Employing such advanced statistical procedures can lead to more beneficial use of available data sets and a more efficient use of limited financial funding. We reanalyzed the data using recently developed packages in SAS and the freely available software package R. The generalized linear model provides a suitable framework for categorical and / or dependent response variables. We analyzed the data using a fixed effect repeated measures model, a random effect clustered data model, and a spatially dependent fixed effects model using normally distributed continuous data. In addition, using the GLIMMIX procedure in SAS, we analyzed the data using a spatially dependent – random effects model with multinomial errors. This method allows for analysis of untransformed ordinal data, such as the data created when performing Rutschblock and stuffblock tests. Comparisons to the original analysis and suggestions for improving analytical efficiency are discussed. Our analyses help to provide a methodological context for future analyses of similar regional spatial data.
Power-laws provide a means for investigating snow avalanche frequency-magnitude relationships and their contributing factors. This research uses power laws to explore variations in avalanche size proportions through space and time, as well as investigating factors which may contribute to these variations. Data utilized for this work includes the Westwide Avalanche Network data from the western United States for regional analyses, with path-specific analyses focused on data from Utah’s Little Cottonwood Canyon. Results show power-law exponents vary through space both at the regional level and between individual avalanche paths. Avalanche size proportions, with respect to space, are the product of terrain based variables at both the mountain range and the path levels, with alpha angles significantly correlated to the proportion of small to large avalanches. This research also indicates that variation in exponents through time is indicative of changes in seasonal weather and snowpack characteristics, with mean snow height also significantly correlated to the proportion of small to large avalanches. Knowledge of power-law exponents for particular avalanche paths, and their relationship to seasonal snowpack depth, may be helpful for managing avalanches along highway corridors, in ski areas, or in backcountry forecasting operations.
Most avalanche fatalities occur due to dry slab avalanches. However, wet snow avalanches are also dangerous and can be particularly difficult to predict. The rate of change from safe snow conditions to dangerous snow conditions occurs rapidly in a wet snowpack, often in response to water production and movement. This research focuses on the relationship between snow stratigraphy and water movement in an inclined snowpack. Concentrating on the capillary barrier effect and flow finger formation within the snowpack, dye tracer was mixed with water and applied to a stratified snowpack to observe and measure the movement of water in various snow grain types, sizes, densities, and temperatures. Experiments show that even a slight textural change within dry snow grains produce a capillary barrier. The amount of water needed to produce flow fingers depends on the snow structure. Both capillary barriers and flow finger formation may play a large role in wet slab avalanche formation. Increasing global mean temperatures may increase the frequency of wet snow avalanches of all types, so a better understanding of the processes involved is important.
The seasonal snow cover is spatially variable. Spatial variability of layer properties is due to various external and internal process drivers interacting with terrain and ground cover during and after the deposition process. Many processes that act as process drivers such as radiation and wind cause spatial variations of the snowpack at several scales. The most challenging process is probably wind that might hinder prediction of variability at the slope scale. The complexities and uncertainties involved in snow slope stability evaluation and avalanche prediction are largely due to the variable nature of the snow cover. Many studies have tried to quantify spatial variability. Different methods have been used and the studies covered a variety of scales. Accordingly, some results appear contradictory, suggesting that the degree of spatial variation varies widely. This is not surprising, and is partly due to the methods used and of course, due to natural conditions. For example, the variation will strongly depend on the measurement scale – the so called support – of the method which varies from 10-4 m2 for the SnowMicropen to 3 m2 for the rutschblock test. The layering was found to be less variable than, for example, the stability of small column tests. Whereas it is often perceived that the results of the studies were not conclusive, they completely changed our view of spatial variability. The importance of scale issues, in particular for avalanche formation became evident. Geostatistical analysis has been introduced and used to determine the length of spatial autocorrelation and to derive appropriate input data for numerical models. Model results suggest that spatial variation of strength properties has a substantial “knock-down” effect on slope stability and that the effect increases with increasing spatial correlation. The focus on scale has also revealed that spatial variations can promote instability or inhibit it. With the awareness of scale the causes of spatial variability can now be addressed. We will review the present state of knowledge, discuss consequences for avalanche forecasting and snow stability evaluation, and recommend future research directions.
Many avalanches are triggered from shallower parts of a slope. Past research demonstrates that initiating fractures in such areas is easier than in deeper places where the applied stresses must penetrate through more overlying snow before affecting the weak layer. However, to our knowledge there is no work on whether fractures more effectively propagate from deeper to shallower areas, or from shallower to deeper areas. During the 2006/07 and 2007/08 winters, we looked at fracture propagation using standard Extended Column Tests, modified Extended Column Tests with column widths of 200 and 300 cm, and Propagation Saw Tests. We tested fracture propagation on slopes with highly variable weak layer depth and with reshaped slab thickness. Our results suggest that fractures are more likely to propagate further when traveling from under thin to thick slabs than in the opposite direction. We support our test results with four case studies. In these cases, slopes only partially released with big explosives applied where the weak layer was deeper, but then avalanched entirely a few days later when tested with small loads where the weak layer was shallower. Thus, shallower areas of the slab may be both the easiest place to initiate a fracture and also the best place from which to propagate a fracture. These results have broad implications for backcountry travel, avalanche avoidance, and avalanche control work.
Dry slab avalanche stability typically increases over time in the absence of active loading from new snow or wind. However, field observations suggest that occasionally slopes showing no signs of instability in the morning avalanche later in the day when the snow surface is warmed by the sun. In this paper we present evidence that dry snowpack fracture propagation propensity may increase during sunny days as the snow surface warms up and becomes wet. During four warm, sunny days in the winters of 06/07 and 07/08, we tracked changes in results for both Extended Column and Propagation Saw tests. Our data suggest that snow surface temperature affects fracture propagation propensity on inclined slopes, with fractures more likely to propagate when the snow surface is wet. We support our test results with two case studies where explosives and ski cuts produced no avalanches when the snow surface was cold and dry, but when those same slopes were re-tested after the snow surface warmed to zero degrees they avalanched. In both cases the weak layer was dry and had a temperature below zero. We hypothesize that fracture propagation propensity may increase due to increased surface creep or due to changes in the mechanical properties of the slab.
Spatial patterns are an inherent property of most naturally occurring materials at a large range of scales. To describe spatial patterns in the field, several observations are made according to a certain sampling design. The spatial structure can be described by the semivariogram range, and nugget and sill variances. We test how reliably seven sampling designs estimate these parameters for simulated spatial fields with predefined spatial structures using a Monte Carlo approach. Five designs have been used previously in the field for snow cover sampling, whereas two designs with semi-random sampling locations have not been used in the field. The designs include 84–159 sampling locations covering small mountain slopes typical of snow avalanche terrain. The results from the simulations show that all designs: (a) give reasonably unbiased estimates of the semivariogram parameters when averaged over many simulations, and (b) show considerable spread in the semivariogram parameter estimates, causing large uncertainty in the semivariogram estimates. Our results suggest that any comparisons of the estimated semivariogram parameters made with the sampling designs will be associated with large uncertainties. To remedy this, we suggest that optimal sampling designs for sampling slope scale snow cover parameters must include more sampling locations and a stratified randomized sampling design in the future.
Determining whether the snowpack is becoming more spatially variable or uniform is important for accurate avalanche forecasts. Greater variability increases uncertainty in extrapolation and prediction. Our results offer a look at the evolution of the spatial variability of shear strength of two buried surface hoar layers in southwestern Montana, USA, over time. We studied the layers from shortly after burial until they were no longer the weakest layer in the snowpack. We selected study sites with planar slopes, uniform ground cover, and wind-sheltered locations. This simplified the comparison of the plots by minimizing initial spatial differences so we could focus on temporal change. Within each site, we sampled four 14 m×14 m plots with more than 70 shear frame tests in a layout optimized for spatial analysis. At both sites, the layers gained strength at a rate that slowed as the layers aged. Although there was little change in the relative variability, absolute variability increased through time. Temporal change was more pronounced when the layers were younger and were gaining strength more rapidly. Additional tests at one plot suggested a correlation length, or the distance at which test results are related, for shear strength of just a few meters. At the other plot, the surface hoar layer collapsed during the initial sample. An initial dramatic decrease in shear strength occurred after this collapse
followed by strengthening during that day and into the following day. Though we measured increasing absolute variability through
time, uncovering changes in our other measures of spatial variability proved elusive. Developing methods and techniques for
adequately characterizing variability, and temporal changes in that variability, will continue to be challenging.
This study aims to improve analytical techniques for studying stratigraphic dimensions and hardness characteristics of thin weak layers in the mountain snowpack, with particular interest to buried surface hoar layers. By determining which structural characteristics of such weak layers are associated with shear strength, we may be better able to monitor and predict stability, which is relevant for avalanche forecasting and management. We utilize moving window statistical operations to analyze SnowMicroPen (SMP) penetrometer hardness profiles of a buried surface hoar layer. Results indicate that significant weak layer thinning and hardening of the interface between the weak layer and its substratum coincided with significant increases in shear strength, as measured using a size corrected shear strength index derived from concurrent stability tests. With aging, variations in slab thickness appeared to positively affect the hardness and inversely affect the coefficient of variation (CoV) of hardness of weak layer boundaries. These findings support previous research that proposed the strengthening of buried surface hoar layers results from the gradual penetration of the surface hoar crystals into the substratum which allows stronger bonds to form at this critical interface. These analytical techniques allow stratigraphic dimensions and hardness characteristics to be quantified and analyzed, improving our ability to monitor stratigraphic characteristics associated with shear strength and stability of the mountain snowpack.
Wet loose snow avalanches are a significant hazard within many ski areas. Wet snow stability changes dramatically over short time periods which typically coincide with operating hours, and few quantitative tools exist for avalanche workers attempting to predict the onset of wet snow avalanching. Field work was conducted at two study sites in southwestern Montana during the springs of 2003, 2004, 2005, and 2006. The study is composed of three separate experiments. The first documents stratigraphic boundary conditions present during periods of wet loose instability. Results show that melt-water accumulation within the upper 15cm of the snowpack increases the likelihood of wet loose avalanche occurrence. The second focuses on the mean daily and minimum daily air temperatures, and how well each variable indicates wet loose avalanche activity. Results are consistent with prior research and clearly show that temperature alone is not a good indicator. The third relates wet loose snow avalanching to surficial shear strength. A 250cm2 shear frame was used to make as many as 210 surficial shear strength measurements of melt-freeze snow per day. Changes occurred rapidly within the meltfreeze cycle as shown by highly significant changes in shear strength within half hour intervals. Most importantly, the data shows an apparent association between surficial shear strength and avalanche activity. When shear strength measurements dropped below 250 Pa wet loose avalanches were observed, and triggered, in the immediate vicinity of study slopes. Conversely, surficial stability on the study slope improved when shear strength values exceeded 250 Pa. This research provides insights into wet loose snow avalanching and the development of possible tools for better predicting wet loose snow avalanche occurrence.
The worst nightmare for an avalanche worker is to assess an unstable slope as stable since the consequence of such an assessment is that you, your clients or the public could be caught in an avalanche. Thus, a primary goal in avalanche forecasting is to minimize such “false-stable” errors. In this paper we analyze the first season of data from the SnowPilot database. Starting with nearly 1,000 snowpits and 3,500 stability tests, we use stability test scores, shear quality, and weak layer depth to identify what we term the “critical weak layer” in each pit. We also divide the pits into “stable” and “unstable” categories based on the assessed snow stability and observations of obvious signs of instability (collapsing, cracking and recent avalanche activity). This filtering leaves us with 289 compression, rutschblock and stuffblock stability tests that fractured on the critical weak layer on unstable slopes. Of those 289 tests, 38 of them (13%) presented “false-stable” results, which we define as CT21 or greater, RB5 or greater, or SB drop heights 40 cm or greater. If we include shear quality and consider strong test results with a Q1 shear to be unstable, we decrease our false-stable cases to around 9% of the total. This implies that – if we use only stability test results – around 1 in 10 times we assess unstable slopes we will conclude that it is stable, which is unacceptably high. Recently spatial variability research has led some to argue that digging snowpits is unnecessary or futile, but we believe our data reinforce the idea that the key to analyzing snow stability lies in digging more rather than fewer pits, and using a holistic approach that considers much more than simple stability test results. Though our dataset is limited, it suggests that digging multiple pits might be an effective strategy for minimizing false-stable situations. In fact, having stability tests and associated shear quality from two different, but representative locations on the slope might decrease the chance of a false-stable error from around 10% to closer to 1%.
This research documents two cases where field workers unintentionally fractured a snowpack weak layer, but no avalanche released. Measurements from before and after the fractures provide unique data sets on the temporal change of snow stability. Shear strength decreased immediately after fracture on both slopes. Subsequent strengthening occurred in both cases, though the rates differed presumably due to the characteristics of the weak layers. Our results have two important implications. First, they suggest the sub-critical weak layer fractures assumed as a prerequisite in some snow slab avalanche release models are transient features, and future modeling efforts must take this into account. Second, they provide insights into interpreting snow stability tests and assessing the stability of slopes with fractured weak layers.
While doing avalanche mitigation work or traveling in the backcountry, occasionally a sizable part of a slope fractures without triggering an avalanche. An example is when a weak layer fractures with a characteristic “whumpf” sound and tensile cracks open up, but no avalanche releases. Disagreement exists among avalanche professionals about the immediate safety of these slopes. Many assume that if the slope does not slide during initial fracture propagation then it is unlikely to slide and is probably safe. Others treat the slopes with extra caution, especially immediately following the event. This paper provides a synopsis of recent research and two case studies that provide insight into this problem. Research shows that shear strength decreases immediately after a collapse, followed by differing strengthening rates. In both case studies, avalanche mitigation work with explosives resulted in the fracturing of some slab boundaries, as evidenced by tensile cracks visible on the surface. Additional explosives applied to the slopes shortly following the initial fractures resulted in sizable avalanches, casting doubt on the idea that fractured slopes are necessarily safe. Over many years and a handful of such experiences, an unofficial policy at Big Sky Ski Area has evolved whereby the snow safety group typically will not open slopes that have deep fractures until the following day. Our paper does not provide definitive answers about the safety of fractured slopes. However, it does point out uncertainties in our knowledge and, as a result, suggests taking a cautious approach toward such slopes.
Throughout the winter, avalanche forecasters issue bulletins to help the public and the managers of public facilities make avalanche safety decisions. These bulletins typically describe important snowpack features and current weather events before rating the avalanche danger on a scale of one through five. Although the character of avalanche conditions may vary between regions, the physical processes that form avalanches are universal. In addition, the methods used to forecast avalanche activity are similar throughout North America and Europe. We use the distribution of fatal avalanche accidents with respect to forecasted avalanche danger level to examine how effectively avalanche forecast groups communicate with the public and how consistent these groups are within countries and internationally. The results show that avalanche forecast groups are effectively communicating with the public when relatively benign or very dangerous conditions exist.
For dry slab avalanches, fractures initiate and propagate in a weak layer or along an interface. Current field tests like compression or stuffblock tests are designed for assessing fracture initiation; however, these tests may not be as useful for assessing fracture propagation. Furthermore, in some cases these test may identify layers that are more likely to initiate a fracture under stress, but not necessarily those layers that initiate and propagate the fracture as well, thereby occasionally “masking” those layers of real concern in a weak snowpack. This paper describes the development of a new field test that aims to assess both fracture initiation and propagation in an isolated column. Tested during the winters of 2005-06 in Colorado and 2006 in New Zealand, this test is a variation of the compression test and can be used in the same manner with the stuffblock. By tapping on one side of an extended column of 30 cm downslope by 90 cm in the cross slope direction, the extended column allows a slab to transmit stress across the width of the column. The fracture initiation results are collected as well as the results of the fracture propagation across the extended column. Out of 68 tests of unstable snowpacks (where avalanches recently occurred, or there was whumphing or shooting cracks) the fracture propagated across the entire block in 1 or 2 loading steps every time (100%) and 63 times (93%) it fractured with a compression test load of easy to moderate. Conversely, out of 256 pits where the snowpack was stable, only 4 cases (1.6%) propagated across the entire extended column through a single layer or interface. Thus, in stable snowpacks a fracture may be initiated, but it typically does not propagate across the column. For the snowpacks we tested the extended column test effectively discriminated between stable and unstable slopes.
Wet loose snow avalanches are a significant hazard within many ski areas. Wet snow stability changes dramatically over short time periods which typically coincide with operating hours, and few quantitative tools exist for avalanche workers attempting to predict the onset of wet snow avalanching. This study documents changes in surficial shear strength during melt freeze cycles and relates these changes to observed wet loose avalanche activity. We conducted field work at two study sites in southwestern Montana over the course of four April days in 2005 and 2006. We used a 250 cm2 shear frame to make as manay as 210 surficial shear strength measurements of melt-freeze snow per day, and adjusted our results for known shear frame size effects. We also collected SnowMicroPen penetrometer profiles in conjunction with shear strength during one melt-freeze cycle. Initial results are encouraging. Changes occurred rapidly within the melt-freeze cycle as shown by highly significant changes in shear strength within half hour intervals. SnowMicroPen data shows significant positive correlations between the microstructural hardness of snow and shear strength. Most importantly, our limited data shows an apparent association between surficial shear strength and avalanche activity. On 22 April 2006 when our shear strength measurements dropped below 250 Pa we observed, and triggered, wet loose avalanches in the immediate vicinity of study slopes. Conversely, surficial stability on our study slope improved when shear strength values exceeded 300 Pa.
Snow avalanches are a major mountain hazard that kills hundreds of people and causes millions of dollars in damage worldwide annually. Yet, the relationship between the well-documented spatial variability of the snowpack and the avalanche release process is not well understood. We utilize a cellular automata model to show that the spatial structure of shear strength may be critically important for avalanche fracture propagation. Fractures through weak layers with large-scale spatial structure are much more likely to propagate over large areas than fractures through weak layers with smaller-scale spatial structure. Our technique of integrating spatial structure into the model can improve many cellular automata models that aim to explain and predict other natural hazards, such as forest fires, landslides and earthquakes.
Avalanche forecasting involves the prediction of spatial and temporal variability of the stability of the snowpack. Greater spatial variability increases the uncertainty of forecasts and reduces the ability of a forecaster to extrapolate snowpack stability reliably. A greater understanding of the spatial patterns of stability, and how they change through time, could improve avalanche forecasting.
I examined temporal changes in shear strength and stability of three persistent weak layers at three different sites. Sites were located on uniform slopes to minimize factors that introduce variability or large-scale trends in the snowpack. At each site, shear strength and stability of the same persistent layer were measured in adjacent plots, sampled at intervals of one to eight days apart. Experimental variograms and pit-to-plot ratios provided measures of the spatial variability. Because adjacent plots began with similar conditions, differences between the plots were attributed to temporal change.
Shear strength of two buried surface hoar layers increased through time and became more variable. As the layers aged, the rate of strengthening decreased. Stability indices initially increased, then decreased as snowfall increased the slab stress. Changes in the spatial structure were most apparent when the layers were younger and gaining strength most rapidly. As the layers aged, the spatial measures provided less information.
Strength of depth hoar increased initially, then decreased as the depth hoar grew and bonds weakened. Spatial correlation increased over time between the first three plots. A strong wind event and warm weather led to considerable change to the snowpack between the third and fourth samples, complicating comparisons.
On these three weak layers, shear strength could be reliably extrapolated over a distance of at least 17 m on 86% of the days sampled, provided a sufficient number of tests were conducted to characterize the statistical distribution. The optimal spacing of tests changes as the autocorrelation length of shear strength changes. The number of tests required increases as the overall variability of shear strength increases. This suggests that test spacing is less important on older layers because the autocorrelation length is short, but more tests are required to characterize the slope statistically.
The U.S. Forest Service initiated avalanche control and forecasting in the United States. The Forest Service manages large tracts of publicly owned land called National Forests and they permit certain activities and businesses on those lands. The Forest Service began permitting ski areas on National Forests in 1938 when they issued a special use permit to Alta Ski Area in Little Cottonwood Canyon near Salt Lake City, Utah.
Almost immediately the Forest Service realized that avalanches threatened the public both while they traveled the Little Cottonwood Canyon Road to reach Alta Ski Area and while they actually skied. The Forest Service hired Douglass Wadsworth to help mitigate the threat and he became the first Forest Service Snow Ranger.
The Forest Service Snow Ranger Program grew to become one of the most effective and innovative avalanche control and forecasting programs in the world. Forest Service Snow Rangers pioneered the use explosives for avalanche control in the US, developed the first effective avalanche forecasting programs in the US, and initiated the use of military artillery for avalanche control in the US.
My paper will trace the development of the Military Artillery for Avalanche Control Program in the United States from its inception to today. It will explain how military artillery work, examine alternatives to military artillery, discuss three 106mm RR accidents and discuss the future of military artillery for avalanche control in the U.S.
We investigated a buried surface-hoar layer using the SnowMicroPen (SMP), an instrument designed to measure detailed snowpack profiles. We collected data from two adjacent parts of a slope 6 days apart. In addition, one manual snowpack profile was sampled each day, as well as 50 quantified loaded column tests (QLCTs) which provided an index of shear strength. For the SMP data, a 900 m2 area was sampled on both days in a grid with points 3mapart, with some sub-areas of more closely spaced measurements. We collected 86 SMP profiles on the first day and 129 on the second day. Our analyses involved manually locating layer boundaries and calculating statistics for the force signal through the surface-hoar layer. The shear strength index increased by 40% between the two sampling days, but the SMP data show no statistical difference in layer thickness, and the mean, minimum, median, and a variety of percentile measures of the SMP force signal through the layer also do not change. Interestingly, the maximum hardness, and the variance and coefficient of variation of the SMP signal, increased. Since the small SMP tip might only break one or a couple of bonds as it passes through the weak layer, we interpret these changes as being indicative of increasing bond strength. Though we cannot specifically tie the increasing maximum hardness of the SMP signal to our QLCT results, our work suggests that the maximum SMP signal within buried surface hoar layers may be useful for tracking increases in the shear strength of those layers.
This paper compares the spatial structure of the compressive strength of slab and weak layers from two different snow climates. Our data come from arrays of SnowMicroPen (SMP) measurements in southwestern Montana, USA, and near Davos in eastern Switzerland. In both cases, buried surface hoar comprised the critical snowpack weakness. We analyzed the SMP data by manually delineating the surface hoar layers and a number of layers within the slabs above the weak layers. Using the log-10 transformed mean of the penetration resistance for our layers of interest, we investigated the spatial structure of the data by looking for linear slope-scale trends and assessing the residuals of any trends with semivariograms. Our results demonstrate that the layers investigated had variable spatial structures, both in terms of their linear trends and their semivariograms. This suggests that each snowpack layer has a unique spatial structure possibly arising from its depositional pattern and the subsequent changes to the layer when buried. We also demonstrate that the specific layout used for the measurements strongly influences the observed spatial variability. The complicated spatial structures of individual layers, and how they interact, likely contributes to the sometimes confounding overall patterns of spatial variability of stability observed on snow slopes.
SnowPilot (www.snowpilot.org) is a free software program that allows users to collect snowpit and avalanche occurrence data onto a Personal Digital Assistant (PDA or Palm Pilot). This data is then stored on a PC where it can be viewed in a graphical snowpit format. The data is also uploaded to a database on the web where it can be viewed and accessed.
This study presents evidence that slope aspect plays a significant role in the formation, size, type, and extent of surface hoar and near-surface faceted crystals. Experimental stations were placed on the north and south-facing aspects of Pioneer Mountain in southwest Montana to measure wind speed and direction, snow temperatures from 0.1m above the snow surface to 0.35m below the snow surface, incoming and outgoing shortwave radiation, and snow surface temperature. Each time a surface hoar or near-surface faceted crystal layer formed, snowpack and meteorological variables were gathered and snow crystals were collected, measured, characterized and photographed so that crystal size and structure could be compared between aspects. Persistent weak layers formed on January 9-10, 2004, and January 12-13, 2004. Results show a statistically significant difference in the size of surface hoar and near-surface faceted crystals based on slope aspect. During these two periods, better developed near-surface facets formed on south-facing slopes, while surface hoar crystals grew larger at the north-facing study site. These aspect dependent differences are important for assessing mountain range scale spatial variability and may also play a role at smaller scales due to subtle aspect changes. The results of this research may help us better understand some of the differences in weak layer formation on different aspects which, in turn, lead to different avalanche conditions.
For over 35 years avalanche workers in the United States used observations described on “the blue and green sheets” published by the USDA Forest Service for standard snow, weather and avalanche observations. Those observation outlines served the U.S. avalanche community well and produced a valuable long-term data set. However, a new and updated set of guidelines are now needed since the blue and green sheets are out of print, and the number, types and needs of avalanche programs in the U.S. have changed.
The American Avalanche Association and the USDA Forest Service National Avalanche Center joined forces to research, write and publish a new set of observational guidelines to support avalanche programs in the United States. This effort began by licensing the Canadian Avalanche Association’s Observational Guidelines and Recording Standards for Weather, Snowpack and Avalanches. This well established document along with the blue and green sheets formed a solid foundation for the new guidelines, which were specifically developed to fit the diverse needs of avalanche programs in the U.S. The resulting document, Snow Weather, and Avalanches: Observational Guidelines for Avalanche Programs in the United States, was published by the American Avalanche Association in the fall of 2004.
Information about existing snowpack weaknesses is essential for backcountry avalanche forecasting. However, the incorporation of detailed information about snowpack weaknesses significantly increases the complexity of the forecasting process. The goal of this research is to examine the scale characteristics of persistent snowpack weaknesses and related avalanche activity with respect to large-scale backcountry avalanche forecasting (≥ 1000 km2). The study focuses on the snowpack of the mountain ranges in Western Canada, namely the maritime Southern Coast Mountains, the transitional Columbia Mountains and the continental Rocky Mountains.
Scaling and scale issues are of fundamental importance in the avalanche forecasting process due to the multi-scale character of the avalanche phenomenon. Although professionals have developed successful strategies to use information across scales, scaling needs to be incorporated explicitly into formalized forecasting approaches. Hierarchy theory (Ahl and Allen, 1996) is used in this research as a conceptual framework for discussing scale issues in avalanche forecasting. The two-dimensional reference system consists of a temporal hierarchy of seven levels representing the main groups of factors contributing to avalanches. Within each temporal level, there is an embedded spatial hierarchy of processes.
In this research, the SNOWBASE database of Canadian Mountains Holidays (1996/97 – 2000/01) and the InfoEx dataset of the Canadian Avalanche Association (1991/92 – 2001/02) were used to examine the temporal and spatial characteristics of three main types of persistent snowpack weaknesses (weak layers of faceted grains, surface hoar layers and pure crust interfaces) and their related avalanche activity. While significant weaknesses of all types were often consistently observed across large parts of the study area, the related avalanche activity exhibited distinct smaller-scale patterns in space and time specific to the weakness types.
This research suggests using ‘Avalanche winter regimes’ as a new classification scheme for describing local avalanche characteristics with respect to forecasting. While existing snow climate classifications (see, e.g., Mock and Birkeland, 2000) focus only on the average winter weather characteristics, it is the comprehensive character of a winter, including the sequence of events that produce persistent weaknesses, which is of crucial importance for backcountry avalanche forecasting. The analysis of this research reveals three distinct initial avalanche winter regimes for Western Canada.
The spatial variability of snowpack mechanical properties strongly influences the fracture initiation and fracture propagation properties of the snowpack, thereby largely controlling the avalanche formation process. The slope-scale spatial variability of the snowpack was investigated on small potential avalanche slopes above timberline near Davos, Switzerland. The characterization of variability was accomplished by combining classical and new measurement methods. The sampling strategy was to optimize the measurement layout for geostatistical analysis. The fracture initiation properties of the snow cover were measured with stuffblock and rammrutsch point stability tests at 24 locations on each slope. High-resolution profiles of penetration resistance were recorded with a novel snow micro-penetrometer at more than 100 locations on each slope. On each slope, a classical stratigraphic profile was made and samples of weak layers were taken. The spatial structure of stability and penetration resistance was modeled as a trend plus residual variation that was described with the semivariogram. The trend was modeled as a linear regression on the measurement coordinates. A spherical semi-variogram model was used to describe the residual variation.
Twenty weak layers were identified by fractures in point stability tests. The spatial variability of point stability had a spatial structure, mainly in the form of slope-scale trends. The trends accounted for around 40% of the observed spread. The quartile coefficient of variation around the linear trend was around 20% with a maximum of around 50%. This variation around the trend had little spatial structure and was within the measurement error of the tests. The weak layer depth partly explained the variation in stability.
The snow cover stratigraphy was reconstructed from the snow micro-penetrometer profiles. The 21 investigated layers were both weak layers and wind-slabs. The weak layers were identified in all penetrometer profiles on a slope, i.e. they were spatially continuous. A significant spatial trend in penetration resistance was found in most layers. The quartile coefficient of variation around the linear trend was around 15% with a maximum of around 60%. Spatial structure around the linear trend was found in all layers except in a layer of buried surface hoar. The range of spatial auto-correlation varied between 2 m and more than 10 m. The layer depth partly explained the spatial variation in penetration resistance. Each layer had unique geostatistical properties with regard to penetration resistance. These were likely caused by the different depositional processes acting during each depositional event, thus emphasizing the sedimentary nature of the snow cover.
Comparing the results from the point stability tests and the micropenetrometer profiles, similar spatial trends at the slope scale were found.
For the first time, the three-dimensional variability of the snow cover was quantified. The results can be used to improve snow cover models that currently do not reflect the spatial variability observed in the field. With regard to the avalanche formation process, the study showed that the weak layers were continuous (through going) at the slope-scale, and that their penetration resistance had a spatial structure with typical length scales of a few meters. Likewise, results from snowpack stability tests had a spatial structure. Therefore, stability tests will not give random results and are useful to assess snow cover stability. The type of variability found suggests that once initiated, a fracture is likely to propagate through the weak layer without being arrested before reaching the critical size where a snow slab avalanche is released.
This research uses a two-dimensional cellular automaton model, with inputs taken from field data, to mimic snow slab avalanche release. The initial weak layer shear strength in the model follows a normal distribution with known mean and standard deviation. For each realization of the spatial distribution of weak layer shear strength, the model stresses all cells equally until the weakest cell fractures. The stress from the fractured cell is transferred to cells in the neighborhood of the fractured cell, possibly causing a propagation of the fracture and a model avalanche. The stochastic shear strength field makes the model avalanche size stochastic for statistically constant initial conditions. The standard deviation of shear strength strongly affects the proportion of model avalanches covering nearly all cells in the model, with low variability leading to a high proportion of large model avalanches, a result that supports previous conceptual models. Stress transfer properties in the model are also important for the proportion of large model avalanches, but no field data exist to constrain these values. Spatial auto-correlation of shear strength is likely to be important for the size of model avalanches, and will be built into a future version of the model.
This research investigated whether single snowpits could reliably represent snowpack strength and stability conditions throughout apparently ‘uniform’ slopes. Seven slopes were selected by experienced avalanche forecasters, three each in the Bridger and Madison Ranges of Southwest Montana (USA), and one in the Columbia Mountains near Rogers Pass, British Columbia (Canada). Teams performed 10 quantified loaded column tests in each of five snowpits within a 900 m2 sampling site on each ‘uniform’ slope, measuring shear strength in a single weak layer. Collection of slab shear stress data enabled the calculation of a stability index, SQLCT. Altogether, eleven trials were performed during 2000/2001 and 2001/2002, testing several weak layer types exhibiting a wide range of strengths. Weak layer strength and slab stress conditions varied widely across the sampling sites, with coefficients of variation in weak layer shear strength ranging from 10% to 50%, coefficients of variation in shear stress from 2% to 48%, and stability indices ranging from 1.8 to 5.7. Of the 54 snowpits completed, 10 pits were empirically rejected as unrepresentative of the stability index at their sampling sites. Of the remaining 44 statistically analyzed pits, 33 pits were statistically representative of their site-wide stability index, and the other 11 pits were found statistically unrepresentative of their site. All five snowpits at a site were statistically representative of their site-wide stability index in three of the eleven trials. The frequent inability of single pits to reliably represent stability on those eleven 900 m2 sampling sites, located on apparently ‘uniform’ slopes, highlights the importance of improving our understanding of the processes affecting the variability of snowpack stability on any given day and the uncertainties associated with ‘point’ stability data.
Avalanche forecasting involves the prediction of spatial and temporal variability of the snowpack. To predict avalanches with more accuracy it is important to determine whether the snowpack is becoming more spatially variable or more spatially uniform. Greater variability increases uncertainty in extrapolation and prediction. Our results offer a look at the evolution of the spatial variability of shear strength and stability of a buried surface hoar layer in southwestern Montana, USA, from shortly after burial until it was no longer the weakest layer in the snowpack. We selected the study site for its 27-degree planar slope, uniform ground cover, and wind-sheltered location. This simplified the comparison of the plots by minimizing initial spatial differences so we could focus on temporal change. Within the site, we sampled four 14 m x 14 m arrays of more than 70 shear frame tests in a layout optimized for spatial analysis. Over a three-week period, the sampling of the four adjacent arrays showed temporal change. The variability of the shear strength of this layer initially decreased then became increasingly variable through time. This suggests that extrapolating test results to other locations becomes increasingly unreliable as layers age, a result that matches practical experience. The data also provide indications that shear strength has a correlation length, the distance at which test results are related, of just a few meters. This short correlation length demonstrates quantitatively why stability tests that are relatively close together can be quite different.
Many ski areas, backcountry avalanche centers, highway departments, and helicopter ski operations record and archive daily weather and avalanche data. The objective of this thesis is to present probabilistic techniques that allow avalanche forecasters to better utilize weather and avalanche data by incorporating a Geographic Information System with a modified meteorological nearest neighbors approach. This nearest neighbor approach utilizes evolving concepts related to visualizing geographic information stored in large databases. The resulting interactive database tool, Geographic Weather and Avalanche Explorer, allows the investigation of the relationships between specific weather parameters and the spatial pattern of avalanche activity. In order to validate these new techniques, two case studies are presented using over 10,000 individual avalanche events from the past 23 years that occurred at the Jackson Hole Mountain Resort.
The first case study explores the effect of new snowfall, wind speed, and wind direction on the spatial patterns of avalanche activity. Patterns exist at the slide path scale, and for groups of adjacent slide paths, but not for either the entire region as a whole or when slide paths are grouped by aspect. Since wind instrumentation is typically located to measure an approximation of the free air winds, specific topography around a given path, and not aspect, is more important when relating wind direction to avalanche activity.
The second case study explores the spatial variability of hard slab and dry loose avalanches, and characterizes these avalanche types with respect to their geographic location and associated weather conditions. I analyzed these data with and without the incorporation of three weather parameters (wind speed, 24-hour maximum temperature, and new snow density). Slide paths near each other often had similar proportions of hard slabs and a higher proportion of hard slabs occurred on exposed ridges. The proportion of loose avalanches also was similar for adjacent slide paths, and these paths were typically sheltered from strong winds. When I incorporated the three weather parameters I found significant increases in the average proportion of hard slabs with increases in new snow density, but not for changes in the 24-hour maximum temperature or wind speed. When I analyzed the proportion of loose avalanches associated with the three weather parameters I found a more direct relationship than with hard slabs. Changes in both wind speed and density significantly changed the average proportion of loose avalanches, with low wind and low density resulting in higher proportions of loose avalanches. My results quantify what operational avalanche forecasters have long known: Geographic location and weather are both related to the proportion of hard slab and dry loose avalanches.
A study on the formation of radiation recrystallized near-surface facets in snow was performed experimentally in an environmental chamber. This recrystallization occurs when surface snow metamorphoses into faceted crystals that result from absorbed solar radiation coupled with cooling effects from longwave and turbulent fluxes. The environmental chamber utilized a metal-halide lamp to mimic solar radiation, which penetrates the snow adding thermal energy at depth. In addition, the ceiling was cooled to simulate a cold sky, thus inducing a net longwave radiation loss at the snow surface. Turbulent flux parameters, including relative humidity and wind velocity were measured. Forty-centimeter thick snow samples with insulated sides were placed in the -10 C chamber on a constant temperature plate also at -10 C. The study focused on the significance of radiation balance and snow density on the recrystallization of snow near the surface. Imposed constant boundary conditions led to formation of facets of varying size at and near the snow surface. Faceting was observed when applied solar flux between 350 – 1100W/m2 was combined with longwave and turbulent exchange for snow with densities below 300 kg/m3. To better understand the governing processes and to extend the number of scenarios a thermodynamic model was used to extrapolate upon the experimental results. The model incorporated meteorological inputs and calculated a snowpack temperature profile based on relevant snow parameters. Conclusions from both experimental and model analysis show radiation and snow density to be significant factors in radiation recrystallized near-surface facets.
Few avalanche forecast models are tailored specifically for wet avalanche forecasting. Bridger Bowl (intermountain climate) is a good area to develop a wet avalanche probability model. The primary archived data consists of eight variables. The archived data for March from 1968 to 2001 (1996 data unavailable) were used to develop 68 predictor variables related to temperature, snowpack settlement, and precipitation. The original dataset was divided into days with snowfall in the past 48 hours (new snow) and days without (old snow). There were 33 significant old snow variables and 22 significant new snow variables. Six variables are common to both old and new snow. The best predictor variables for old and new snow are different. The variables were analyzed with binomial logistic regression to produce probability models for old snow and for new snow wet avalanche conditions. The old snow model uses the prediction day minimum temperature and the two-day change in total snow depth as predictor variables and has a 89% overall success rate. However, the majority of this success is due to correct prediction of days without wet avalanches (96% of all correct predictions). The new snow model uses the prediction day minimum temperature and three-day cumulative new snow water equivalent as predictor variables, but is less useful. The models are applicable only to Bridger Bowl. The numerical forecast models can be used as one of the tools in the forecasting toolbox but limited data and complexity of process require that the decisions about closure remain in the hands of the ski patrol.
Wet avalanches are a safety concern for all ski areas because they are difficult to control artificially and the shift from safe to dangerous wet snow conditions can happen very quickly. Forecasting for wet avalanche conditions in intermountain ski areas, such as Bridger Bowl, Montana, can be especially difficult because intermountain snow climates can exhibit wet avalanche characteristics of either maritime or continental snow climates. Various statistical models have been developed for avalanche prediction; however, most are tailored specifically for dry avalanche forecasting. Archived meteorological, snowpack and avalanche data for the month of March from 1968 to 2001 (1996 data unavailable) were used to develop 68 possible predictor variables related to temperature, snowpack settlement, and precipitation characteristics. The original Bridger Bowl dataset was divided into a ♯new snowα and an ♯old snowα dataset. A ♯new snowα day has newly fallen snow that is less than 48 hours old; an ♯old snowα day has newly fallen snow that is more than 48 hours old. The two datasets were used to determine whether the factors that influence ♯old snowα and ♯new snowα wet avalanche occurrence differ. Hypotheses were developed and tested to determine which ♯old snowα and ♯new snowα variables behaved significantly different on days with wet avalanches compared to days with no wet avalanches. The 33 ♯old snowα significant variables and the 22 ♯new snowα significant variables were analyzed with binomial logistic regression to produce one prediction model for ♯old snowα wet avalanche conditions and another prediction model for ♯new snowα wet avalanche conditions. The ♯old snowα model uses the prediction day minimum temperature and the two day change in total snow depth as predictor variables. This model has a 75% success rate for calculating accurate wet avalanche probabilities for ♯old snowα days. The ♯new snowα model uses the prediction day minimum temperature as well as the three day cumulative new snow water equivalent as predictor variables. This model has a 72% success rate for calculating accurate wet avalanche probabilities for ♯new snowα days.
The majority of slab avalanche accidents occur when the victim triggers the slide. Slab hardness is an important property affecting skier-triggered avalanches because hardness partially determines whether sufficient stress reaches the weak layer to cause failure and/or fracture. This study examines how new and old snow layer hardness varies with aspect and which meteorological variables most influence those changes. Slab hardness was measured with a ram penetrometer on north and south aspects from January through March, 2000 at Jackson Hole Mountain Resort and Grand Teton National Park, Wyoming. Continuous weather data were obtained from weather stations at Jackson Hole Mountain Resort. Analyses were carried out on new and older near surface snow layers. New snow layer hardness increased most rapidly on the south aspect due to accelerated settlement and densification from warming by incoming shortwave radiation. With the exception of the surface layer, old snow layers, 2 months after deposition, became harder on the north aspect in comparison to the south aspect. A temperature index was calculated for the south and north aspects to describe the delayed effect of increasing temperature on increasing hardness through sintering, settlement, and densification. The south temperature index, maximum daily temperature, and the interaction between maximum daily temperature and incoming shortwave radiation were the most significant predictors of new snow layer hardness on the south aspect. The north temperature index, maximum daily temperature, and the previous day’s wind speed were the most significant predictors of new snow layer hardness on the north aspect. The temperature index was the only significant predictor of old snow layer hardness on both the north and south aspects. The results of this research suggest that it may be possible to use meteorological factors to predict changes in snow hardness, which is an important component in predicting skier-triggered avalanches.
Many ski areas, backcountry avalanche centers, highway departments, and helicopter ski operations record and archive daily weather and avalanche data. This paper presents a probabilistic method that allows avalanche forecasters to better utilize historical data by incorporating a Geographic Information System (GIS) with a modified meteorological nearest neighbors approach. This nearest neighbor approach utilizes evolving concepts related to visualizing geographic information stored in large databases. The resulting interactive database tool, Geographic Weather and Avalanche Explorer, allows the investigation of the relationships between specific weather parameters and the spatial pattern of avalanche activity. We present an example of this method using over 10,000 individual avalanche events from the past 23 years to analyze the effect of new snowfall, wind speed, and wind direction on the spatial patterns of avalanche activity. Patterns exist at the slide path scale, and for groups of adjacent slide paths, but not for either the entire region as a whole or when slide paths are grouped by aspect. Since wind instrumentation is typically located to measure an approximation of the free air winds, specific topography around a given path, and not simply aspect, is more important when relating wind direction to avalanche activity.
A helicopter ski guide died in an avalanche at 3:00 pm on February 10, 1996 while guiding clients on a large west-facing slope in the Smoky Mountains of Central Idaho. The ski guide, his clients, and several other guides and their clients had skied over one hundred runs in the vicinity and on the same slope prior to the accident. The avalanche measured nearly one half-mile wide and involved three chutes that are separated by counter ridges. Three other guides and sixteen clients were on an unaffected counter ridge when the avalanche occurred.
The avalanche fractured on buried surface hoar and near surface facets. The avalanche varied from one foot to an estimated five feet deep and ran on 28 to 38 degrees slopes. Forty-eight hours of extreme avalanche hazard including many natural and human triggered avalanches followed the fatal accident.
Prior to the February 10 avalanche, significant WNW winds on January 27 rapidly loaded slopes throughout the area and numerous large natural and artificially triggered new snow avalanches occurred. No avalanches were reported from February 4 through February 9. An “inverted” storm dropped several centimeters of snow and three centimeters (1.2 inches) of water equivalent from February 4 – 8. Temperatures rose from –16 degrees C (2F) February 2 to –1 degree C (30F) February 9 and then spiked to 5 degrees C (41F) at 2:00 pm February 10.
Our paper will examine possible reasons why the deadly cycle occurred despite no overly dramatic weather event immediately preceding it, how the accident affected the local community including avalanche education, and possible ways to deal with similar events in the future.
This paper presents evidence of frequency-size power-laws in several groups of snow avalanche paths. Other natural hazards, such as earthquakes and forest fires, exhibit similar power-law relationships. In addition, an analysis of the response of one group of snow avalanche paths to storms through time demonstrates a power-law between the response of the system and the binned frequency of those responses. Our results, as well as our experience with these complex, non-linear systems, are consistent with self organized criticality. The practical implication of this work is that the frequency-size relationship for small and medium sized avalanches may be useful for quantifying the risk of large snow avalanches within a group of avalanche paths.
The spatial variability of snow stability presents a challenge to avalanche workers and scientists attempting to assess avalanche danger. Recent research has demonstrated, and attempted to explain, snowpack and snow stability variations over individual slopes. However, these studies have provided single snapshots of an exceedingly dynamic system instead of investigating stability variations over time. This research begins to address how spatial variability at the local slope scale changes through time and the some possible mechanisms for those changes. We performed ten Quantified Loaded Column Tests (QLCTs) in each of five snowpits within a 900 m2 plot on a relatively uniform slope on three different days. We make a case that the behavior of the snow avalanche system is consistent with complex, nonlinear Earth Surface Systems (ESSs), and that research into ESS behavior in other fields might prove to be useful for examining changes in spatial variability through time. In particular, ESSs are typically characterized by sensitivity to initial conditions, which leads to increasing spatial variability through time. Our series of three QLCT trials exhibited changing stability patterns, suggesting that the spatial variability on a slope may increase through time in the absence of external forcing, but that variability may then decrease when additional load is added to the slope. This research provides some interesting initial insights into changes in spatial variability for practitioners and provides some baseline data for future scientific work in this area.
Our poster summarizes the results of Birkeland and Landry (2002), where we direct the interested reader for more details on this research. Our results show scale-invariant relationships between avalanche frequency and size in several groups of avalanche paths. Specifically, data from Gothic, Colorado, Bridger Bowl, Montana, Jackson Hole, Wyoming, and Snowbird, Utah demonstrate loglinear relationships, or power-laws, for both natural and artificially released avalanches. We also analyzed a group of snow avalanche paths at Yule Creek, Colorado for their response to 104 storms, showing that a power-law exists between the magnitude and frequency of the resultant avalanche cycles. Recently, other researchers have also documented a number of power-laws associated with snow avalanches, including acoustic emissions, crown crack heights, and crown crack lengths (Louchet et al., in press; Dendievel et al., 2002; Faillettaz et al., 2002a; Faillettaz et al., 2002b), as well as presenting evidence that slab avalanching is a chaotic process (Rosenthal and Elder, 2002). Several other natural hazards, such as earthquakes and forest fires, exhibit similar power-law relationships (Drossel and Schwabl, 1992; Olami et al., 1992; Malamud et al., 1998). Our results, as well as our avalanche forecasting experience, suggest that snow avalanches may exhibit self-organized criticality (Bak et al., 1987; 1988; Bak and Chen, 1991), an observation recently backed up by the research of others (i.e., Louchet et al., in press; Dendievel et al., 2002). Frequency-size relationships for small and medium sized avalanches, and avalanche cycles, may be useful for quantifying the risk of rarer large snow avalanches, and avalanche cycles, within a given group of avalanche paths.
The objective of this study was to investigate the importance of topography in controlling the geographic patterns of deep snow temperature gradients within a seasonal snowpack. Demonstration of the relative importance of topography in influencing spatial snowpack temperature gradients could aid future modeling of snow layer development and behavior, with benefits for avalanche and snowmelt modeling. This spatial, or geographic, analysis of the relationship of snow temperature gradient patterns to topography utilizes landscape-scale modeling in an attempt to identify responses in complex, mountainous terrain. During the snow season of 2001-2002, 30 temperature profiles were sampled on each of nine sample days. Profiles were collected through the use of a portable snow temperature profile probe (Deems, 2001). These data were used to calculate temperature gradients for each profile. Topographic attributes were derived from a digital elevation model (DEM) using a Geographic Information System (GIS). Linear regression models quantified the relationships between the topographic variables and snow temperature gradient patterns in our spatially distributed dataset, and demonstrate the relative importance of the terrain variables in determining spatial patterns of temperature gradients.. Analysis shows a complex pattern of relationships between temperature gradients and the static topographic variables. A qualitative assessment of weather variables recorded onsite suggests the utility of using more dynamic variables such as weather data in future research.
The objective of this study was to investigate the relative importance of topography in controlling the geographic patterns of snow temperature gradients within a seasonal snowpack. Regression models quantified relationships between topographic parameters and temperature gradient statistics for our spatially distributed dataset. Demonstration of the relative importance of topography in influencing spatial snowpack temperature gradients could aid future modeling of snow layer development and behavior, with benefits for avalanche and snowmelt modeling. This spatial, or geographic, analysis of the relationship of snow temperature gradient patterns to topography, utilizes landscape-scale modeling in an attempt to identify responses in complex, mountainous terrain.
During the snow season of 2001-2002, 30 temperature profiles were sampled on nine sample days. Profiles were collected through the use of a portable snow temperature profile probe (Deems, 2001). These data were used to calculate temperature gradients for each profile. Topographic attributes were derived using a Geographic Information System (GIS) and a Digital Elevation Model (DEM). Linear regression assessed the relationship between the topographic variables and snow temperature gradient patterns, and demonstrates the relative importance of the terrain variables in determining spatial patterns of temperature gradients. Analysis of the regression models shows a complex pattern of relationships between average temperature gradients and topographic variables. A qualitative assessment of weather variables suggests the utility of weather data in future modeling efforts.
This study investigates whether collecting shear quality data in conjunction with stability test results improves snowpack evaluations. Over the past six seasons we have consistently evaluated shear quality when evaluating snowpack stability. Shear quality is subjectively evaluated on a 3-tiered scale from Q1 (clean, fast shears) to Q2 (average shears) to Q3 (irregular or dirty shears). Our method is a formalization of what ski patrollers and others have been doing in the U.S. and elsewhere for at least several decades. We used a dataset of nearly 700 individual stability tests (rutschblock, stuffblock and compression tests) collected by seven observers on slopes from Alaska to Chile. In addition to stability test results, observers noted whether slopes they felt were similar to their snowpit location had avalanches, or collapsing or cracking snowpacks, on that day. Results suggest that shear quality provides important stability information, especially when stability test results appear to indicate relatively stable conditions, but the shear quality is rated Q1. This might be because stability test results are often spatially variable, while our experience indicates that shear quality is more homogeneous. Given these results, we believe formally integrating some description of shear characteristics into stability assessments may be important for avalanche workers and backcountry enthusiasts.
The majority of slab avalanche accidents occur when the victim triggers the slide. Slab hardness is an important property affecting skier-triggered avalanches because hardness partially determines whether sufficient stress reaches the weak layer to cause failure and/or fracture. This study examines how new and old snow layer hardness varies with aspect and which meteorological variables most influence those changes. Slab hardness was measured with a ram penetrometer on north and south aspects from January through March, 2000 at Jackson Hole Mountain Resort and Grand Teton National Park, Wyoming. Continuous weather data were obtained from weather stations at Jackson Hole Mountain Resort. Analyses were carried out on new and older near-surface snow layers. A temperature index was calculated for the south and north aspects to describe the delayed effect of increasing temperature on increasing hardness through sintering, settlement, and densification. The south temperature index, maximum daily temperature, and the interaction between maximum daily temperature and incoming shortwave radiation were the most significant predictors of new snow layer hardness on the south aspect. The north temperature index, maximum daily temperature, and the previous day’s wind speed were the most significant predictors of new snow layer hardness on the north aspect. The temperature index was the only significant predictor of old snow layer hardness on both the north and south aspects. The results of this research suggest that it may be possible to use meteorological factors to predict changes in snow hardness – an important component in predicting skier-triggered avalanches.
This research investigated whether single snowpits can reliably represent snowpack stability on uniform slopes. The study utilized seven carefully selected slopes, three each in the Bridger and Madison Ranges of Southwest Montana, and one in the Columbia Mountains near Rogers Pass, British Columbia. Teams performed ten Quantified Loaded Column Tests in each of five snowpits within a 900 m2 plot at a slope, measureing shear strength in a single weak layer. Collection of slab shear stress data enables the calculation of a strength/stress stability ratio. Altogether, eleven stability sampling trials were performed during 2000/01 and 2001/02, testing several weak layer types exhibiting a wide range of strengths. Of the 54 snmowpits completed, 26 pits (48%) represented plot-wide stability and 28 pits (52%) did not one plot collapsed prior to completion of a 55th pit. Two of the eleven plots did contain full complements of five representative snowpits. The results of this study suggest the importance of improving our understanding of the processes affecting the variability of snowpack stability on any given day.
Avalanche forecasters frequently perform field tests at study plots or other representative sites to reduce uncertainty regarding snowpack stability. This research investigated whether single snowpits represented stability throughout a carefully selected plot. The study utilized seven relatively uniform 900 m2 plots, three each in the Bridger and Madison Ranges of Southwest Montana, and one in the Columbia Mountains near Rogers Pass, British Columbia. Teams collected systematic samples from five snowpits, each containing ten 0.25 m2 stability-sampling cells, at each plot. Quantified loaded column stability tests measured strength in a single weak layer. Collection of in-situ slab shear stress data enabled the calculation of a stability ratio. Altogether, eleven stability sampling trials were performed during 2000/2001 and 2001/2002, testing several weak layer types exhibiting a wide range of strengths. Of the 54 valid snowpit results, 28 (51.9%) represented plot-wide stability, 16 did not, and the remaining 10 pits were empirically unrepresentative of their plot. Three of the eleven plots sampled contained full complements of five representative snowpits. As an additional component of this study, a GIS-based model extrapolated Bridger Range plot stability data onto avalanche starting zones, with poor results. The results of this study provide sufficient evidence of local spatial variation in snowpack stability within relatively uniform plots to reject the hypothesis that stability at a single snowpit will reliably represent a plot. However, these results do not suggest that information from snowpits is not important. Experienced forecasters interpret study plot stability data conservatively and are capable of utilizing “targeted sampling” and a variety of other data to effectively reduce uncertainty about slope stability.
Many ski areas, backcountry avalanche centers, highway departments, and helicopter ski operations record and archive daily weather parameters and the resulting avalanche activity. This paper proposes a new probabilistic method to allow avalanche forecasters to better utilize their historical weather and avalanche data by incorporating a Geographic Information System (GIS) with a modified meteorological nearest neighbors approach. This approach utilizes concepts from Geographic Visualization (GVis) and Knowledge Discovery in Databases (KDD). The resulting interactive database tool allows avalanche forecasters to visually explore regional spatial patterns of avalanche activity at multiple scales. This technique allows the correlation between weather parameters and the spatial pattern of avalanche activity. An example of this method was implemented using 23 years of historical avalanche data from the Jackson Hole Ski Area with over 10,000 avalanche events to analyze the effect of new snow, wind speed, and wind direction on the spatial patterns of avalanche activity. Patterns were found at the slide path scale and for sub-regional groups, but not for the entire region as a whole, or when slide paths were grouped by aspect categories.
Snow microstructure significantly influences the mechanical, thermal, and electromagnetic properties of snow. The microstructure is constantly evolving from the time it is deposited on the surface until it sublimates or melts. The resulting time variant material properties make the study of snow metamorphism of fundamental importance to a wide variety of snow science disciplines. Dry snow metamorphism has traditionally been classified by the thermal gradient encountered in the snowpack. Snow experiencing a predominantly equi-temperature environment develops different microstructure than snow that is subjected to a temperature gradient. As such, previous research has evaluated snow metamorphism based upon select thermal gradient dependent processes, when in reality, there is a continuum of physical processes simultaneously contributing to metamorphism. In previous research, a discrete temperature gradient transition between the two thermal environments has been used to activate separate morphological analyses. The current research focuses on a unifying approach to dry snow metamorphism that is applicable to generalized thermal environments. The movement of heat and mass is not prescribed, but is allowed to develop naturally through modeling of physical processes. Heat conduction, mass conservation, and phase change equations are derived in a simplified two-dimensional approach. Each differential equation is non-linearly coupled to the others through phase change. The microstructural network is then discretized into elements and nodes. Finite difference equations are developed for the network, and numerically solved using iterative techniques. The finite difference model provides a unique platform to study the influence of numerous geometric and thermodynamic parameters relating to dry snow metamorphism. Numerical metamorphism studies in an equi-temperature environment agree well with established trends and published experimental results. A smooth transition between equi-temperature and temperature gradient environments is defined and influencing parameters are examined. In the temperature gradient environment, a dominant grain theory based on crystallographic orientation is postulated through numerical modeling, and is supported by experimental observation. Several specific metamorphism applications, ranging from avalanche debris sintering to model integration in a full scale snowpack, are presented. The microstructural model has proven to be capable of evaluating metamorphism for a broad range of geometric parameters and thermal environments, yet is flexible enough to accommodate additional scenarios.
This research investigates snow stability on the eastern side of a small mountain range in southwest Montana, U.S.A., on one mid-season day and one late-season day during the 1996/97 winter. Although previous research has addressed snow stability at smaller spatial scales, this is the first field-based study to investigate snow stability (as measured by stability tests) over a mountain range in order to better understand its spatial distribution and the implications for predicting dry-slab avalanches. Using helicopter access, six two-person sampling teams collected data from over 70 sites on each of the two sampling days. Variables for terrain, snowpack, snow strength and snow stability were generated from the field data, and analyzed using descriptive statistics, correlation analysis and multiple regression. Results from the first sampling day show stability is only weakly linked to terrain, snowpack and snow-strength variables due to consistently stormy weather conditions leading up to that day. The second field day’s results demonstrate a stronger relationship between stability and the other variables due to more variable weather conditions that ranged from periods of sunshine to storms. On both days stability decreased on high-elevation, northerly-facing slopes. The data-structure complexity provides insights into the difficulties faced by both scientists and conventional avalanche forecasters in predicting snow avalanches.
Snow avalanches are a significant hazard in mountainous environments around the world. This paper investigates the major February 1986 avalanche cycle that occurred in the western United States, and broadly analyzes the avalanche, snowpack, and weather conditions at twenty sites. These analyses suggest that the avalanche cycle resulted from the interaction of a relatively ‘normal’ snowpack with an exceptional storm event, which was particularly noteworthy for the amount of precipitation it produced. Composited 500-hPa anomaly maps show the event resulted from an uncommonly persistent blocking pattern that resulted in a strong zonal flow and copious moisture being funneled over the western United States. Understanding severe and widespread avalanche cycles may improve our long-term forecasting of these events, and help mitigate the resulting avalanche activity.
Avalanche forecasters can better anticipate avalanche extremes if they understand the relationships between those extremes and atmospheric circulation patterns. We investigated the relationship between extreme avalanche days and atmospheric circulation patterns at four sites in the western United States: Bridger Bowl, Montana; Jackson Hole, Wyoming; Alta, Utah; and Taos, New Mexico. For each site, we calculated a daily avalanche hazard index based on the number and size of avalanches, and we defined abnormal avalanche events as the top 10% of days with recorded avalanche activity. We assessed the influence of different variables on avalanche extremes, and found that high snow water equivalent and high snowfall correspond most closely to days of high avalanche hazard. Composite-anomaly maps of 500 hPa heights during those avalanche extremes clearly illustrate that spatial patterns of anomalous troughing prevail, though the exact position of the troughing varies between sites. These patterns can be explained by the topography of the western United States, and the low-elevation pathways for moisture that exist to the west of each of the sites. The methods developed for this research can be applied to other sites with long-term climate and avalanche databases to further our understanding of the spatial distribution of atmospheric patterns associated with extreme avalanche days.
Recently, a computer model has been developed by the Swiss Federal Institute for Snow and Avalanche Research that simulates the evolution of a natural snow cover. Using common meteorological parameters as input, SNOWPACK predicts characteristics such as snowpack temperature and density, in addition to snow microstructure and layering. An investigation was conducted to evaluate the effectiveness of SNOWPACK in a Montana climate. A weather station was constructed in the Bridger Mountains near Bozeman, Montana, to provide the meteorological parameters necessary to run SNOWPACK. Throughout the 1999–2000 winter, weekly snow profiles were performed in undisturbed snow to provide a benchmark for the model output. Density, grain size, and crystallography were recorded on 10-cm intervals over the full snow depth, and the temperature profile was monitored with a thermocouple array. Finally, the meteorological parameters were input into SNOWPACK, and a statistical comparison was performed comparing the predicted snowpack to the observational data. Snowpack temperatures are predicted reasonably accurately by SNOWPACK. The modeled and observed densities correlated well, but the model typically underestimates snowpack settlement. Comparison of grain size and shape was problematic due to different definitions utilized by the model and observer, but still demonstrated some agreement.
To assist local avalanche forecasters, Buser (1983) developed an avalanche prediction system based on the nearest neighbor method. NXD2000 is an improved and further developed version of the program, and has been installed at several places in Switzerland, Austria, Kazakhstan, and the USA. Improvements include the introduction of explanatory variables such as settlement and a mass of wind-transported snow and functions giving certain values of a variable more load, e.g. snow surface temperature which is important in the range right below O°C, but not far below at -20°C. We explored the influence of the functions and attempted to optimize the variable loadings at Parsenn Ski Area in Davos, Switzerland and Snowbasin Ski Area in Utah, USA.
A slab avalanche occurs when a weak layer or interface below the slab fractures, causing the slab to release. A large percentage of accidents associated with slab avalanches result when the victim triggers the release. While it is recognized that avalanches are caused by the unique interaction between the weak layer and the slab, most research has focused on the composition of the weak layer and how its strength changes over time. Relatively few studies have examined how slab properties change over time and space. We expect slab properties to change over time and space because of both the inherent variations in the energy balance associated with aspect and the dependency of snow metamorphism on the energy balance. This research attempts to address the following research question: to what degree does slab hardness vary with aspect and time when elevation and slope are kept relatively constant? Slab hardness was measured with a ram penetrometer on north and south aspects from January through March, 2000 at Jackson Hole Mountain Resort and Grand. Teton National Park, WY. While hardness increased more rapidly on southern than northern aspects due to settlement and densification, solar radiation on southern aspects also contributed to greater heterogeneity in the snowpack.
Recently, a computer model has been developed by the Swiss Federal Institute for Snow and Avalanche Research that- simulates the evolution of a natural snow cover. Using common meteorological parameters as input, SNOWPACK predicts characteristics such as snowpack temperature and density, in addition to snow microstructure and layering. An investigation was conducted to evaluate the effectiveness of SNOWPACK in a Montana climate. A weather station was constructed in the Bridger Mountains near Bozeman, Montana, to provide the meteorological parameters necessary to run SNOWPACK. Throughout the 1999-2000 winter, weekly snow profiles were performed in undisturbed snow to provide a benchmark for the model output. Density, grain size. and crystallography were recorded on 10 cm intervals over the full snow depth, and the temperature profile was monitored with a thermocouple array. Finally, the meteorological parameters were input into SNOWPACK, and a statistical comparison was performed comparing the predicted snowpack to the observational data. Snowpack temperatures are predicted reasonably accurately by SNOWPACK. The modeled and observed densities correlated well, but the model typically underestimates snowpack settlement. Comparison of grain size and shape was problematic due to different definitions utilized by the model and observer, but still demonstrated some agreement.
A computer model called SNOWPACK h as been developed by the Swiss Federal Institute for Snow and Avalanche Research that simulates the evolution of a mountain snowpack. Using meteorological parameters measured at mountain weather sites, a prediction of snowpack stratigraphy is made by modeling snow characteristics such as snow depth, temperature, density, grain size, and crystal type. In order to evaluate the accuracy of the model, SNOWPACK was run using meteorological variables measured at a mountain weather station near Bozeman, Montana, and weekly snow profiles were conducted to provide a benchmark for the model output. A statistical analysis was then performed in order to objectively compare the predicted snowpack to the snow profile data.
While kinetic-growth metamorphism has been investigated in the laboratory previously, new technologies allow similar experiments to be performed with greater accuracy and efficiency. A methodology was developed that utilizes a computed tomography (CT) scanner to obtain cross-sectional images over time of a snow sample under a large temperature gradient. Using innovative stereological software, the microstructural properties of the snow can then be measured from the two-dimensional CT images.
SNOWPACK predicts snowpack temperatures with reasonable accuracy, but is less effective at simulating density. Different definitions of grain size utilized by the model and human observers resulted in large variations between the modeled and observed grain size. Predicted and observed grain types also demonstrated low correlation. Other aspects of the analysis suggest that the manner in which the surface energy exchange, wet snow metamorphism, and new snow density are modeled need refinement. Despite these deficiencies, SNOWPACK still provides the snow practitioner with a useful tool for simulating the mountain snowpack. The laboratory experiments succeeded in quantifying the changes in snow microstructure during kinetic-growth metamorphism, but are also applicable to equilibrium conditions. The presented methodology demonstrates that CT technology and stereological methods are improvements over previous techniques for investigating snow metamorphism. Since the metamorphism laws in SNOWPACK are based on snow microstructure, the results of future experiments could provide data permitting validation and improvement of these theories.
Most historical avalanche data are stored in tabular form, and are difficult to access and visualize. This project utilizes a Geographic Information System (GIS) to spatially visualize historical avalanche data at Jackson Hole Ski Area in Wyoming, USA. Jackson Hole is an ideal location for this study because: 1) the ski patrol have been collecting avalanche and weather records for over 25 years, 2) the ski area is spatially diverse, with aspects ranging from north to east to sough, and 3) a three-meter Digital Elevation Model (OEM) and a geo-referenced orthophoto are available for the area. We have combined the OEM, orthophoto, and historical data to geographically display historical avalanche data, as well as to be used for future statistical modeling. The GIS allows greater three-dimensional visualization of historical avalanche data by displaying specific avalanche path activity and corresponding size, thereby enabling forecasters to better utilize their historical data. Additional weather information can be displayed simultaneously, and the entire picture can be viewed through time in a frame-by-frame movie format to help avalanche forecasters observe trends. The GIS can also be used to obtain data about specific avalanche paths such as slope, elevation, and aspect. We hope to use these data to train statistical models, and to eventually display historical avalanche activity that most closely resembles the current day in real-time for the ski area avalanche forecasters.
The snow avalanche climate of the western United States has long been suspected to consist of three main climate zones that relate with different avalanche characteristics: coastal, intermountain, and continental. The coastal zone of the Pacific mountain ranges is characterized by abundant snowfall, higher snow densities, and higher temperatures. The continental zone of the Colorado Rockies is characterized by lower temperatures, lower snowfall, lower snow densities, higher snow temperature gradients, and a more persistently unstable snowpack resulting from depth hoar. The intermountain zone of Utah, Montana, and Idaho is intermediate between the other two zones. A quantitative analysis of snow avalanche climate of the region was conducted based on Westwide Avalanche Network data from 1969 to 1995. A binary avalanche climate classification, based on well-known thresholds and ranges of snowpack and climatic variables, illustrates the broadscale climatology of the three major zones, some spatially heterogeneous patterns, and variations with elevation. Widespread spatial shifts toward more coastal conditions occurred during 1985/86 and 1991/92, and shifts toward more continental conditions occurred during 1976/77 and 1987/88. Height anomalies at 500 mb explain many of these shifts, but daily plots of climate and avalanche variables during seasonal extremes for sites in northern Utah also illustrate the importance of understanding snowpack and weather variations that occur at daily to weekly timescales. Data from several central Rocky Mountain sites indicate some relationships with the Pacific–North American teleconnection pattern and the Pacific decadal oscillation, illustrating the importance of applying long-term records in an avalanche hazard assessment.
The stuffblock is a new snow stability test developed by the Gallatin National Forest Avalanche Center and used operationally since 1993. The test involves stressing an isolated column of snow 0.30 m2 by dropping a nylon sack filled with 4.5 kg onto the column from 0.10 m increments until weak layer failure occurs. Results over several winters correlate the stuffblock test with the more widely used rutschblock test, and validate the usefulness of the test for evaluating snow stability in several different climates. Further, the test provides results that are consistent between observers, a favorable attribute for regional avalanche forecasting operations which use numerous observers.
Although recent observations indicate that weak layers of near-surface faceted crystals are widely associated with snow avalanches, little research has addressed these layers. Further, current research has been hindered by an absence of a framework with which to discuss their formation. This paper proposes terminology and describes three predominant processes observed in mid-latitude mountains which result in extreme near-surface temperature gradients, thereby forming near-surface faceted crystals: radiation recrystallization, melt-layer recrystallization, and diurnal recrystallization. It is hoped that this framework will improve scientific discussion and theory-building related to the formation and spatial distribution of near-surface faceted crystals.
Though avalanche workers and scientists recognize that distinct patterns of snow stability exist in mountainous terrain, no field-based research has been conducted to rigorously analyze those patterns at a scale of interest to backcountry avalanche forecasters. This research investigates snow stability throughout a small mountain range. Using helicopter access, sampling teams collected data from over 70 sites on each of two sampling days. Variables generated were analyzed using a variety of statistical techniques. Results demonstrate that links exist between terrain and stability, and that those relationships change over time. The data structure is complex, providing insights into the myriad difficulties faced by avalanche forecasters.
In the winter of 1995-96 we investigated the temperature and vapor pressure gradient conditions associated with the formation of faceted crystals that develop in the upper levels of the snowpack due to diurnal recrystallization. We used an array of six thermocouples connected to a data logger to continuously measure snow temperatures in the region from 0.005 m above the snow surface to 0.20 m below the snow surface; Measurements during clear sky conditions in March showed temperature gradients in excess of 200°C m−1 at night in the top 0.05 m of the snowpack, with the temperature gradient shifting direction and exceeding 100°C m−1through this layer during the day. These temperature gradients resulted in vapor pressure gradients which exceeded 25 mb m−1during the day and at night. During this time, a significant weak layer of 1 mm faceted snow formed within 36 h. Widespread avalanche activity occurred for up to 9 d after this layer was buried by 0.50 m of snow.
Determining the likelihood of an avalanche occurring on a given slope is a critical decision faced by both avalanche professionals and backcountry enthusiasts. An important part of that decision-making process is using stability tests and interpreting their results. Though a variety of stability tests are available, we use the rutschblock and stuffblock tests since they both identify weak layers in the snowpack and, to a limited extent, quantify the stresses necessary for weak layer failure. However, there are several limitations to these stability tests, including site selection and interpretation of results, both of which require experience. While site selection may be the most important limitation, we focus on the interpretation of results. Such interpretations can be problematic, since those without the benefit of avalanche experience often latch onto stability test values as an absolute indicator of stability. However, our experience indicates that occasionally dangerous avalanche conditions exist when stability tests show a stable snowpack and that sometimes the snowpack is more stable than , stability tests indicate. In order to partially address the difficulties in test interpretation, we have begun to qualitatively define the quality of the shear failure into three separate categories. This paper will focus on combining the interpretation of rutschblock and stuffblock tests, and describing the shear failure quality, thereby giving useful qualitative field information about the avalanche conditions in a given location. In the end, users must realize that stability tests do not provide a numerical description of the snow stability, but are instead just a piece of the puzzle for avalanche prediction.
Avalanches may be important hydrologically if avalanche deposits alter the timing and volume of runoff. Avalanches often fail from a weak layer such as surface hoar, graupel, or faceted crystals. One important type of faceted crystal. called near-surface faceted crystals. form in the near-surface layers. During the 1997/98 winter. A study of near-surface faceted crystals was conducted along the Red Mountain Pass corridor, Colorado. Six types of near-surface faceted crystals were differentiated: small faceted crystals, radiation recrystallization grains. Faceted precipitation crystals, near-surface hoar, faceted partly-decomposed precipitation crystals and needles. A particularly well-developed near-surface facet layer that evolved in December acted as the dominant weak layer in the study area for seven weeks. Observation of 14 avalanches showed that 79% of the failures occurred on near-surface faceted layers. The majority of stability test failures occurred on near-surface faceted layers. Understanding the growth of near-surface faceted crystals and their effect on snow stability is important for avalanche forecasting
This study examines the effect of forest clearing for ski runs and artificial snowmaking on snow water accumulation at of Big Sky recreation area in southwest Montana. Extensive research shows that clearing the forest canopy for timber harvest increases water stored in the snowpack, but no studies have compared water stored in the compacted, skied, and machine groomed snowpack of ski runs with the water stored under the natural forest canopy. Thirty six sampling locations were selected, terrain variables (elevation, aspect and slope angle) were surveyed, and snow depth, density and water equivalence were measured both under the forest canopy and on the adjacent ski run. Additionally, some of the sampling locations were in areas with artificial snowmaking. Snowpack on the run has a higher density and water equivalence, and a similar depth, to snow off the ski runs. Elevation is the primary terrain variable that affects snowpack characteristics, and increases in snow water equivalence (SWE) with increasing elevation are similar to previously published research. There was no significant difference between snow water storage on ski runs with snowmaking and those without snowmaking, although the snow density on the runs with artificial snow was higher. Total increase in snowpack water storage for all ski runs at Big Sky was greater than 700,000 m3. This number is close enough to values derived from a model designed to measure water yield due to forest cutting that it appears that the clearing of ski runs causes similar increases in snowpack water storage and water yield as forest clearing for timber harvests.
Predicting heavy mountain snowfall, which is critical for avalanche hazard forecasting, is difficult due to complex interactions between rugged topography and atmospheric circulation. In this study the relationship between atmospheric circulation patterns and extreme snowfall events is examined at Bridger Bowl, Montana, U.S.A. which has a 26-year winter record of daily snowfall. Five hundred millibar composite and anomaly maps were constructed for days of heavy snowfall (greater than 32.8 cm). These maps show that during and prior to heavy snowfall, Bridger Bowl is located beneath the back side of a upper-level trough, with predominant winds and storms coming from the northwest. This atmospheric circulation pattern differs from those for other high-elevation sites in the North American interior due to the surrounding regional topography. High mountain ranges to the southwest and west often block incoming moisture, while relatively lower topography to the northwest allows Pacific moisture to reach Bridger Bowl. The results of this study can be used to complement operational forecasting models for predicting heavy snowfall at Bridger Bowl, thereby facilitating snow avalanche forecasting in the region
Birkeland, K.W., R.F. Johnson, and D. Herzberg. 1996. The stuffblock snow stability test. U.S. Forest Service Missoula Technology Development Center publication 9623-2836-MTDC.
In the winter of 1995-96 we investigated the formation of faceted crystals that develop in the upper levels of the snowpack. We used an array of six thermocouples connected to a datalogger to measure hourly diurnal temperature changes in the region from 0.005 m above the snow surface to 0.20 m below the snow surface. Measurements during clear sky conditions in March showed temperature gradients in excess of 200 deg C/m at night in the top 0.05 m of the snowpack, with the temperature gradient shifting direction and exceeding 100 deg C/m through this layer during the day. A significant weak layer of faceted snow formed within 36 hours with a grain size of about 1 mm in the upper snowpack. Widespread avalanche activity occurred for up to nine days after this layer was buried by 0.50 m of snow.
Since snow avalanches are believed to release from zones of localized weakness, knowledge of snow-strength patterns is important for determining slope stability and for applying effective avalanche-control measures. In this study, the spatial variability of snow resistance (an index of snow strength) and depth were measured and compared with terrain features on two inclined slopes. A refined instrument allowed the strength of an entire snow slab to be characterized in a short time. The spatial pattern of trees appeared to affect the pattern of snow depth at one site, where a significant linear relationship was found between snow depth and average snow resistance. These results suggest that localized snow-depth variations may be important in snow-strength genesis. Although a linear relationship existed at that site, additional factors may be critically relevant. A second site with more complex terrain features and less localized wind drifting did not show a linear relationship between depth and average resistance. Instead, complex patterns of resistance demonstrated that many factors contribute to snow resistance. In particular, die snow overlying rocks was found to have significantly weaker resistance than that in adjacent areas not over rocks.
The stuftblock is a new snow stability test developed and used operationally by the Gallatin National Forest Avalanche Center during the 1993-94 winter. Shortcomings of other stability tests, including the inability to effectively communicate results, the complexity of the test, and the time necessary to collect a measurement make the development of a new test desirable. The stuffblock is performed on an isolated column of snow about 0.30 m (l ft) square which is cut out of the wall of a snow pit. A nylon sack (stuff sack) is filled with 4.5 kg (10 lbs) of snow, which is measured with a lightweight scale. An avalanche shovel blade is placed on top of the isolated column and the stuff sack is first placed, then dropped onto the shovel from increasing heights. The drop height is increased in increments of 0.10 m. When shear failure occurs, the drop height is noted. Initial results from the 1993-94 field season indicate that, for snowpack conditions found in southwest Montana, a positive relationship exists between stuffblock drop heights and rutschblock numbers. While the stuffblock is not perfect, it is inexpensive, quick, easy, and provides numbers that can be readily compared between observers. This latter attribute is especially useful for regional avalanche forecasters who must often compare the results of several different observers with differing avalanche skills. For the individual avalanche worker, the stuffblock provides one more useful tool for snowpack stability evaluation.
Birkeland, K.W.. 1993. Recipe for starting a regional avalanche center. The Avalanche Review 11(6), 4. [link coming soon].
Since snow avalanches are believed to release from zones of localized weakness, knowledge of snow strength patterns is important for determination of slope stability and for the application of effective avalanche control measures. No previous studies have mapped snow strength over an entire inclined snow slab. In this study, the spatial variability of snow resistance (an index of snow strength) and depth were measured and compared with terrain features on two inclined slopes in Montana during two winter field seasons. An instrument that indexes snow strength by measuring snow resistance was refined, allowing the strength of an entire snow slab to be characterized in a short time. Measurements of depth and resistance were taken at 1 m intervals across and down the slopes. The spatial pattern of trees appears to affect the pattern of snow depth at the first site, where a significant linear relationship was found between snow depth and average snow resistance during both years of study (p-values <1X10^-6). When data sets were reduced to lessen the effects of spatial autocorrelation, the relationship between snow depth and average resistance continued to be significant (p-values < 7 X 10^-3). These results suggest that localized snow depth variations may be important in snow strength genesis. Although a linear relationship existed at that site, low r values for the two years (r^2 < 0.357) indicate additional factors may be critically relevant. A second site with greater complexity of terrain features and less localized wind drifting did not show a linear relationship between depth and average resistance. Complicated patterns of resistance at that site demonstrate that many factors contribute to snow resistance. In particular, the snow over rocks was found to have significantly weaker resistance than adjacent areas which were not over rocks (p-value < 1 X 10-6). Results may provide predictive information of weak zone locations in snow slabs, which would improve avalanche forecasting and control techniques.
The digital resistograph (DR) is a microprocessor-based probe which is used to rapidly obtain strength index profiles similar to the profiles obtained with the ram penetrometer. The current version of the DR can collect as many as 500 strength profiles in a work day. In this study the DR is compared to the ram penetrometer for repeatability, learnability and ease of use. Additional studies were conducted to determine just how sensitive the DR readings are to penetration rate. The test results did show that the instrument has a repeatability at least as good as the ram penetrometer. The DR has a decided edge in terms of learnability. Two novices were found to give roughly the same DR profiles after a minute of instruction, whereas the ram penetrometer is widely known to show a wide range of results when used by novices. One of the the primary problems with the DR was durability, as the instrument frequently malfunctioned and failed to work properly in cold conditions. The DR shows much promise for becoming a substantial improvement over the ram penetrometer, provided a more reliable and durable version can be engineered.