A new method for visualizing snow stability profiles

Snow stability assessment by interpreting snow profiles is a time consuming and fairly subjective process, especially when snow stratigraphy was recorded without performing a stability test at the same time. Snow stratigraphy is clearly related to snow stability, as had been shown by various studies that linked specific snowpack properties such as grain size and type to instability. We suggest a new method to visualize snow stratigraphy in regard to stability based on six structural variables (also known as the threshold sum approach). Each snow layer is represented by the number of variables that are not in the corresponding critical range. This approach has not only been implemented for manually recorded snow profiles but also – after adapting the threshold values – for simulated snow stratigraphy provided by the numerical snow cover model SNOWPACK. The new visualization method, applied both to the manually observed and simulated profiles, was tested by analyzing the most critical avalanche situations of the winter 2008–2009 in the Dolomites (north-eastern Italian Alps). Results indicate that the new visualization method is well suited to quickly and intuitively derive snow stability, in particular from simulated snow stratigraphy. Stability information derived from simulated profiles was clearly related to the independently estimated degree of avalanche danger. Supplementing the snow cover model SNOWPACK with the adjusted threshold sum approach increases its usefulness for avalanche forecasting purposes.     

On forecasting large and infrequent snow avalanches

Snow avalanches that threaten a highway or a residential area are often large avalanches that have a return period > 1 year. Danger assessment strongly relies on precipitation data since these avalanches are typically triggered by major snow storms. Given the extensive protection work that is in place in the European Alps, the avalanche control service (also called avalanche commission) responsible for danger assessment will usually monitor the avalanche situation throughout the winter, but only become active in case of a major snow fall. Related safety concepts describing the procedures and measures to be taken in a given danger situation are therefore often based on threshold values for new snow. By analysing the avalanche occurrence of a major avalanche path, we show that forecasting based on new snow amounts involves high uncertainty. Whereas the return period of an avalanche to, for example, the road was about 5 years, the return period for the corresponding new snow depth was substantially smaller, in our case slightly less than 2 years. Similar proportions were found for a number of other avalanche paths with different snow climate. The return period of the critical new snow depth was about 2–5 times smaller than the return period of the avalanche. This proportion is expected to increase with increasing return period. Hence, based on the return period of an avalanche path a first estimate for the critical new snow depth can be made. With a return period of the critical new snow depth of 1–2 years, avalanche prediction for individual avalanche path becomes very challenging since the false alarm ratio is expected to be high.     

Comparison of snow stability tests: Extended column test, rutschblock test and compression test

Several field tests have been proposed in the past for evaluating snow stability. However, few comparative studies have been performed so that presently the advantages and disadvantages of the various tests are partly unclear. During winter 2007–2008 we have collected a dataset of 146 snow profiles, consisting of snow stratigraphy, a rutschblock test (RB), one to two extended column tests (ECT) and in most of the cases also one to two compression tests (CT). Study slopes were classified in regard to stability as either rather stable or rather unstable, based on signs of instability or profile classification. We then studied whether the various tests were able to predict the slope stability class. The CT had an almost perfect probability of detection, but — as the structural stability index (threshold sum) — the CT largely overestimated instability (high proportion of false alarms). Of the small scale tests the ECT was best suited to differentiate between stable and unstable situations. By including the ECT score (number of taps), the number of false alarms was slightly reduced. The performance was similar to the RB which is, however, not independent of the stability classification we used. With two adjoining ECTs it was possible to classify 87% of our test slopes with an accuracy of about 90% into rather stable or rather unstable. Comparing two adjacent stability test results showed that only in about half of the pairs the same weak layer showed up as the most critical one. The snowpack properties (weak layer and slab) that favoured unstable test results for the ECT were associated with whole block releases in the rutschblock test. Thus, the two tests seem to provide similar information possibly related to fracture propagation propensity.     

Cross-comparison of meteorological and avalanche data for characterising avalanche cycles: The example of December 2008 in the eastern part of the French Alps

In December 2008, an intense avalanche cycle occurred in the eastern part of the southern French Alps. Southerly atmospheric fluxes that progressively evolved into an easterly return caused important snowfalls with return periods up to 10 years. Cold temperatures and drifting snow had important aggravating effects. The return period for the number of avalanches was above 50 years in two massifs and some of the avalanche had very long runouts that exceeded historical limits recorded in the French avalanche atlas. Using this case study, this paper illustrates and discusses how avalanche reports, snow and weather data and results from numerical modelling of the snow cover can be combined to analyse abnormal temporal clusters of snow avalanches. For instance, it is shown how statistical techniques developed in other fields can be used to test the significance of different explanatory factors, extract spatio-temporal patterns, compare them with previous cycles and quantify the magnitude/frequency relationship at different scales.     

Using GPS for snow depth and volume measurement of centennial avalanche field in High Tatras

Heavy snowfall in the High Tatras at the end of March 2009 was the cause of several avalanche falls in the ?iarská valley of the Slovakian High Tatras Mountains. The resulting avalanche field was almost 28 ha. The event was classified as a centennial avalanche, one of the biggest in modern history in Central Europe. Snow pack was measured by accurate differential GPS technology. Snow depth and avalanche field volume were calculated using a comparison between avalanche and terrain surfaces. Considering the inaccuracies of the photogrammetric digital elevation model, most likely caused by vegetation, new terrain surveying was required after the snow pack melted in October 2009. The results confirmed that deep snow packs even in rugged terrain can be accurately surveyed by GPS technology.     

Multi-basin avalanche simulation: A model

A process oriented avalanche prediction model for wide-area use was developed using five winters' data from Berthoud Pass, Colorado. The model was successfully tested on independent data from the Colorado Front Range and San Juan mountains. This research model is driven by temperature, precipitation, wind, and radiation; it simulates snow transport and deposition in starting zones and the development of a layered snow cover in forest sheltered clearings. Probabilities of avalanche occurrence are estimated according to recent loading and simulated regional snowpack stratigraphy.     

Study of snow climate in the Japanese Alps: Comparison to snow climate in North America

In the case of the Japanese Alps, it is experientially known that there is a notable snow climate difference between the Japan Sea side mountains and the Pacific Ocean side mountains. For the purpose of improving avalanche safety, we studied the snow climate characteristics using meteorological and snow pit data collected from two study plots in the mountain regions. Ten years of meteorological data and 4–10 years of snow pit data were employed in the study. A snow climate classification scheme proposed in North America was used to determine the snow climate of these study plots. The general snowpack characteristics for each snow climate presented in previous studies were used in the present study to determine the snowpack characteristics of the study plots. Both meteorological and snow pit data suggested that the Japan Sea side mountains have the same characteristics as the maritime snow climate in North America. On the other hand, the Pacific Ocean side mountains have unique characteristics caused by a combination of continental and maritime climate influences. The Pacific Ocean side mountains have similar characteristics to the continental snow climate of North America, however, that climate is different in that it is characterized by a large amount of rainfall and a high predominance of faceted crystals and wet grains. We identified a new snow climate for the Pacific Ocean side mountains of the Japanese Alps, a “rainy continental snow climate.”     

Rutschblock-scale snowpack stability derived from multiple quality-controlled SnowMicroPen measurements

Stability prediction from SnowMicroPen (SMP) profiles would support avalanche forecasting operations, since objective stability information could be gathered more quickly than with standard tests, thereby allowing sampling at higher resolution and over larger spatial scales. Previous studies have related the snow properties derived from the SMP to observed snow properties at Rutschblock (RB) and compression test failure planes. The goals of this study are to show to what extent snowpack stability for artificial triggering, based on RB, can be derived from SMP measurements and how multiple measurements at the RB scale improve the results. Measurements at 36 different sites were used for the development of a classification scheme. Each site included a RB test, a manual profile, and 6 to 10 adjacent SMP measurements, for a total of 262 SMP profiles. A recently improved SMP theory was applied to estimate the micro-structural and mechanical properties of manually defined weak layers and slab layers. SMP signal quality control and different noise treatment methods were taken into consideration in the analysis. The best and most robust predictor of RB stability was the weak layer micro-scale strength. In combination with the SMP-estimated mean density of the slab layer, the total accuracy of predicting RB stability classes was 85% over the entire dataset, and 88% when signals with obvious signal dampening (11% of the dataset) were removed. The total accuracy increased when multiple SMP measurements at the RB scale were used to calculate the mean weak layer strength, when compared to using just one SMP measurement at a site. The analysis was robust to trends and offsets in the absolute SMP force, which was a frequent signal error. However, it was sensitive to dampened or disturbed SMP force micro variance. The sensitivity analysis also showed that the best predictor of instability, the weak layer micro-scale strength, was robust to the choice of SMP signal noise removal method.     

Observations and simulations of snow surface temperature on cross-country ski racing courses

Fast skis are essential for an Olympic cross-country skiing athlete. Accurate and timely estimates of the snow surface conditions on a race course are needed to prepare race skis. For training purposes prior to the 2010 Winter Olympics, snow surface and snowpack observations were collected on the cross-country racing track at the Whistler Olympic Park, British Columbia during February 2008 and 2009. During periods with clear skies, snow surface temperatures varied by more than 10 °C around the course while temperatures in the stadium area increased by more than 16 °C from morning to early afternoon. Simulations using the SNOWPACK model of snow surface temperature were within 1 °C of those measured during a four day observation period in the stadium area. Idealized simulations were completed varying cloud cover, slope and aspect. These simulations provided realistic appearing changes in snow surface temperature as a function of time of day.     

Measurements of localized dynamic loading in a mountain snow cover

In the majority of fatal avalanches, skiers and snowmobilers apply load to the snow cover which triggers the initial failure in a weak layer. Understanding how the stress due to the dynamic surface load transmits through the snow cover can help people avoid situations where they can trigger avalanches. Capacitive sensors were used to measure this stress within the mountain snow cover. The three main variables affecting stress transmission through the snow cover investigated in this paper are the type of loading, depth and snow cover stratigraphy. At specific depths, snowmobiles added more stress than skiers did, thus increasing the probability of initiating a fracture in a weak layer and releasing a slab avalanche. The increased penetration depth of snowmobiles into the snow cover compared to skiers was the primary reason for this increase in stress. A decrease in stress was observed with increasing depth. A decrease in stress was observed with increased snow cover hardness. Supportive surface layers created a ‘bridging effect’ that spread stress out laterally and decreased the depth to which it penetrated.     

An adjustment for kinetic growth metamorphism to improve shear strength parameterization in the SNOWPACK model

Shear strength is an important parameter for interpreting the stability of simulated snow covers. In the SNOWPACK model, snow shear strength is estimated as a function of snow density using expressions for different grain types. In the model, shear strength changes discontinuously as grain type changes from rounded to faceted grains (and vice versa), but in nature, shear strength changes take place more gradually. An experiment on the growth of depth hoar allows a new parameterization of continuous changes in shear strength. A parameter is induced in the expression for shear strength as a function of water vapor transport. It ranges from 0 (rounded grains) to 1 (depth hoar) depending on the metamorphic stage of the snow. The parameterization is incorporated into the SNOWPACK model to calculate the progressive change in shear strength during snow metamorphism. The calculated shear strength using the improved SNOWPACK model agreed well with that measured in cold-room experiments using artificial snow. This model, which can calculate gradually changing shear strength, is expected to improve the accuracy of avalanche forecasting.     

An adjustment for kinetic growth metamorphism to improve shear strength parameterization in the SNOWPACK model

Shear strength is an important parameter for interpreting the stability of simulated snow covers. In the SNOWPACK model, snow shear strength is estimated as a function of snow density using expressions for different grain types. In the model, shear strength changes discontinuously as grain type changes from rounded to faceted grains (and vice versa), but in nature, shear strength changes take place more gradually. An experiment on the growth of depth hoar allows a new parameterization of continuous changes in shear strength. A parameter is induced in the expression for shear strength as a function of water vapor transport. It ranges from 0 (rounded grains) to 1 (depth hoar) depending on the metamorphic stage of the snow. The parameterization is incorporated into the SNOWPACK model to calculate the progressive change in shear strength during snow metamorphism. The calculated shear strength using the improved SNOWPACK model agreed well with that measured in cold-room experiments using artificial snow. This model, which can calculate gradually changing shear strength, is expected to improve the accuracy of avalanche forecasting.     

Snow Cover Mapping for Mountainous Areas by Fusion of MODIS L1B and Geographic Data Based on Stacked Denoising Auto-Encoders

Snow cover plays an important role in meteorological and hydrological researches. However, the accuracies of currently available snow cover products are significantly lower in mountainous areas than in plains, due to the serious snow/cloud confusion problem caused by high altitude and complex topography. Aiming at this problem, an improved snow cover mapping approach for mountainous areas was proposed and applied in Qinghai-Tibetan Plateau. In this work, a deep learning framework named Stacked Denoising Auto-Encoders (SDAE) was employed to fuse the MODIS multispectral images and various geographic datasets, which are then classified into three categories: Snow, cloud and snow-free land. Moreover, two independent SDAE models were trained for snow mapping in snow and snow-free seasons respectively in response to the seasonal variations of meteorological conditions. The proposed approach was verified using in-situ snow depth records, and compared to the most widely used snow products MOD10A1 and MYD10A1. The comparison results show that our method got the best performance: Overall accuracy of 98.95% and F-measure of 73.84%. The results indicated that our method can effectively improve the snow recognition accuracy, and it can be further extended to other multi-source remote sensing image classification issues.     

On forecasting large and infrequent snow avalanches

Snow avalanches that threaten a highway or a residential area are often large avalanches that have a return period > 1 year. Danger assessment strongly relies on precipitation data since these avalanches are typically triggered by major snow storms. Given the extensive protection work that is in place in the European Alps, the avalanche control service (also called avalanche commission) responsible for danger assessment will usually monitor the avalanche situation throughout the winter, but only become active in case of a major snow fall. Related safety concepts describing the procedures and measures to be taken in a given danger situation are therefore often based on threshold values for new snow. By analysing the avalanche occurrence of a major avalanche path, we show that forecasting based on new snow amounts involves high uncertainty. Whereas the return period of an avalanche to, for example, the road was about 5 years, the return period for the corresponding new snow depth was substantially smaller, in our case slightly less than 2 years. Similar proportions were found for a number of other avalanche paths with different snow climate. The return period of the critical new snow depth was about 2–5 times smaller than the return period of the avalanche. This proportion is expected to increase with increasing return period. Hence, based on the return period of an avalanche path a first estimate for the critical new snow depth can be made. With a return period of the critical new snow depth of 1–2 years, avalanche prediction for individual avalanche path becomes very challenging since the false alarm ratio is expected to be high.     

Large mobility of dry snow avalanches: Insights from small-scale laboratory tests on granular avalanches of bidisperse materials

This paper reports small-scale laboratory tests on granular avalanches of bidisperse materials made of fine particles and larger ones. These experiments were motivated by a recent study on the rheology of dense flowing snow which provided evidence for relevant similarities in flow behavior between bidisperse granular materials and dry cold snow [Rognon and others, J. Rheol., 52, 3 ([32])]. The mass proportion of fine particles in the initial binary mixture was systematically varied at constant initial released volume, and we measured the resulting flow depth, the avalanche front velocity and the final avalanche runout. In particular, we show that the avalanche mobility is largely increased, about 40% in our tests, when the mass proportion in fine particles reaches a critical value, around 0.25 in our tests. The avalanche deposit is shallow and lengthened for this critical mass proportion in fine particles. The experimental results are interpreted by the existence of different avalanche mobility regimes on the basis of a heuristic model previously reported in literature. Finally, we discuss their possible implications for the dynamics of full-scale dry snow avalanches.     

Spatial estimation of snow water equivalent at different dates within the Adamello Park of Italy

Spatial estimation of snow water equivalent SWE at six different dates from February 1st to June 1st is tackled using Kriging from a sparse network of 14 snow stakes with density within the Adamello Natural Park of Italy. Therein, SWE is measured at these six dates for the period 1967-2009. Second order statistics of SWE are evaluated and linked to physiographic features. The covariance of the SWE field within the Park is studied, necessary for Kriging, and its regularization provided based upon geomorphic attributes. Seasonal dependence of the covariance of the SWE field is observed, and taken into account for optimal estimation. Then, a Kriging procedure based upon the so obtained covariance fields is developed and cross-validated. The accuracy of Kriging estimates is then compared against that of other commonly adopted methods for spatial interpolation. Kriged SWE maps are then produced at the six dates for two sample years, to demonstrate use of the method. Snow Cover Area SCA from the MODIS® satellite is used to constrain Kriging procedure upon snowed areas. The procedure provides well estimated, least variance SWE values and it is relatively simple and fast because it uses only information of physiography of the area. The so obtained maps can be used for spatial estimation of SWE within the investigated region for water availability conjectures, for constraining hydrological models simulating runoff at thaw, for ecological conjectures upon snow cover related species within the Park, and to evaluate snowpack dynamics for avalanche risk assessment.     

Snowpack properties associated with fracture initiation and propagation resulting in skier-triggered dry snow slab avalanches

This study investigates snowpack properties associated with skier-triggered dry slab avalanches, with a particular view on snowpack conditions favoring fracture propagation. This was done by analyzing a data set of over 500 snow profiles observed next to skier-triggered slabs (including remotely triggered slab avalanches and whumpfs) and on skier-tested slopes that did not release a slab avalanche. The relation of the snowpack variables with fracture initiation and fracture propagation, both of which are required for skier-triggering, was investigated. Specific snowpack characteristics, including hardness difference and difference in crystal size across the failure layer, associated with skier-triggered dry slab avalanches were identified and the frequency of skier-triggering was determined. In order to assess snowpack variables favouring fracture propagation, variables from failure layers associated with skier-triggered slabs that were not remotely triggered and relatively small were contrasted with snowpack variables from failure layers of remotely triggered slab avalanches, whumpfs and relatively large slab avalanches. The properties of the slab overlying the weak layer, as well as the layer above the weak layer, were found to affect fracture propagation. Stiffer slabs were associated with large avalanches as well as whumpfs and remotely triggered avalanches. Furthermore, a correlation analysis of snowpack variables with the size and width of the investigated slab avalanches further accentuated the importance of these slab properties with regards to fracture propagation.     

Predicting the fracture character of weak layers from snowpack penetrometer signals

Digital penetrometers provide reliable assessments of snow penetration resistance with depth. However, extracting useful information from the signals relating to snow stability has proved to be challenging. In this study, penetrometer profiles were collected in close proximity to compression tests. A scheme for predicting the fracture character of weak layers in the compression tests from the penetrometer signals is presented. When a two-group classification between sudden (Q1) (an indicator of instability) and other fracture character groups was performed, potential failure layers were correctly classified 80% of the time. The variables offering the best discrimination between sudden and other categories were weak layer thickness, average force gradient above the weak layer, and both the average and the maximum force gradient below the weak layer. The effect of introducing randomly selected layers into the prediction scheme was also investigated. When such layers were introduced, the classification rate dropped to 67%, indicating that more effective fracture character prediction occurred when weak layers were manually pre-identified. This suggests that this scheme should be used in conjunction with a weak layer detection model rather than as a stand alone analytical technique for the purpose of critical weak layer identification. The classification rate dropped further to 55% when a more detailed, four-group classification scheme was used.     

Assessing the applicability of terrestrial laser scanning for spatial snow depth measurements

Snow depth observation in potentially dangerous avalanche-starting zones is important for avalanche prediction and dimensioning of permanent protection measures. The possible danger of avalanches complicates measurements of snow depth in the field (e.g. by probing). Therefore, the applicability of terrestrial laser scanning to measure the depth of the snow cover was analysed. Different long-range laser profile measuring systems were used carrying out numerous field campaigns (Vorarlberg, Austrian Alps). The objective of the study was to discover under which meteorological and snowpack conditions the measurements must be taken in order to provide accurate results (< 10 cm). For the first time a detailed investigation focusing on the limitations and properties of different terrestrial laser scanning systems for application in snow and avalanche research is presented and discussed. Results suggest that under adequate measurement conditions the distance between the scanner position and the surface of the snowpack can be measured with an accuracy < 10 cm. Poor weather conditions such as snowfall or fog preclude the collection of reliable data. If the snow surface was wet and the snow grain size was large (> 1 mm) only 50% of the emitted signal was received, depending on the angle of incidence. In any case, the accuracy decreases with increasing distance to the target. For distances to the object of more than about 500 m, the accuracy that can be achieved with the used TLS measuring systems is rather low and the errors can be significantly larger than 10 cm.     

The use of microwave FMCW radar in snow and avalanche research

This paper describes a frequency modulated, continuous wave (FMCW) microwave radar system used for different types of investigations in snow and avalanche research. Different semi-empirical equations describing transmission and backscatter of electromagnetic energy in snow are compared and applied to analyse the frequency domain spectra of the backscattered radiation. The FMCW scatterometers are either buried in the ground looking upward into the snow cover or are towed on skis looking downward into the snow. The backscatter of electromagnetic radiation from avalanche snow moving perpendicular to the radar beam is analysed to estimate the height of dense flow in the avalanche. The geometrical layering, density, water equivalence, settlement, total snow height, percolation of water through the snow cover and moisture content of the snow are determined from the backscatter of the stratigraphy of a static snow pack.