Snow avalanches have long been an active area of applied glaciology. Empirical and traditional methods of forecasting and controlling avalanches are gradually giving way to the application of modern science and engineering. There are four areas of this science where progress is being made through active interaction among scientists, engineers, and practical men. The artificial, release of avalanches as a control measure has seen a large body of research regarding effects of explosives on snow but still is handicapped by inadequate basic knowledge about avalanche release mechanisms. There is currently a surge of interest in testing numerical methods of avalanche forecasting and several sophisticated statistical techniques have been introduced, but operational forecasting still depends largely on empirical experience. The pressure of development on alpine lands has brought to the fore a number of problems associated with mapping avalanches, determining their return intervals and deducing their behavior from both observations of terrain and vegetation and calculating their behavior from theory. The ability to predict the characteristics of moving avalanches is advancing through a combination of theoretical insights and field observations, but is inhibited by the difficulty of the latter.

Relationships between snow structure, climatic factors, and the nature of avalanche events in the San Juan Mountains of Colorado, U.S.A., are analyzed. Physical and mechanical properties of the snow cover have been monitored on a continuous basis by conventional methods as well as through the application of a profiling snow-density gauge. Temperature-gradient metamorphism is the dominant mechanism which determines the character of the local snow structure. Properties unique to well-developed temperature-gradient snow (depth hoar) are discussed. Three specific metamorphic processes are described which contribute to a highly varied and mechanically weak snow structure. Analysis of fracture-line profile data allow general studies of the local snow climate to be compared with the stratigraphy of actual release zones. Fracture-line data support the conclusion that meteorological conditions prevailing at the time of new-snow deposition play a less significant role in the eventual release of slab avalanches than metamorphic processes which influence the snow cover between storm periods. Shear failure zones are consistently composed of weak layers metamorphosed weeks and even months prior to the precipitation event which triggers the release.

A co-ordinated study involving both a field investigation in Yellowstone National Park and a laboratory investigation was undertaken to evaluate the process of temperature-gradient metamorphism on the mechanical properties of snow. Both parts of the investigation showed that, when subjected to a negative temperature gradient, low-to-medium density snow metamorphoses first into a fine-grained anhedral depth hoar before finally acquiring a fully developed, and stronger, euhedral crystal structure. Measurements indicated that the subhedral snow was the weakest stage in the metamorphic process and that, while strength may drop to as low as 10% of the original strength, the material stiffness decreased by less than 50% Also, it was observed that the location of weakest snow was usually the point of a local maximum density and largest temperature gradient, thus suggesting a relationship between metamorphic state and thermal conductivity.

A powder-snow avalanche is a particular type of turbulent gravity current which it is proposed to treat using methods developed in contexts where the Boussinesq approximation is valid. Hence, it is assumed that the air entrainment rate E primarily depends on the internal Froude number (Tochon-Danguy and Hopfinger, [1975]) and that the large density variation in an avalanche reduces only the drag resulting from the air entrainment. In order for this drag to dominate over the wall stress it is necessary that the slope angle be greater than a certain minimum value which depends on the density ratio of the flow. For a low-density flow, for example, a minimum slope angle of about 5° is necessary. The more unusual aspect is that an avalanche may also entrain snow along its path which increases its mass and can produce in certain cases an accelerating flow. This aspect is included in the theoretical treatment. Experiments have been carried out in an inclined water channel using brine as dense fluid. Visual observations as well as measurements of the velocity and density profiles have been made. Various states of flow are proposed depending on whether the avalanche is “long" (having the structure of a usual gravity current) or “short" (having the structure of a thermal) or whether snow is entrained. The practical consequences which can be drawn from these results are discussed, in particular in the context of the design of defence structures, and flow visualizations of a gravity current impinging on an obstacle give an idea of the protected region.

Variation of flow depth of a flowing avalanche caused by a change in the slope angle is examined using Bernoulli’s energy equation, used in hydraulics to determine a non-uniform steady flow. This equation is modified for a material with internal friction and a strongly curved track. The calculated flow depths are compared with those obtained by tests with a snow slide. In the flow model dry and turbulent friction are taken into account. Friction coefficients are estimated comparing calculated and measured flow depths. It appears that in wet snow they differ from those of dry snow. Finally, this model is used to calculate the runout distance of a natural avalanche.

Equations are derived, using a total potential-energy formulation, which predict the wavelength of the dominant wave pattern of a flowing snow slab. Account is taken of the visco-elastic properties of an assumed surface layer of snow which overlays a distinct visco-elastic foundation layer of snow. The predicted wave pattern is recognized as a precursory mechanism which influences the characteristics of the sawtooth flank fracture pattern that may remain following a snow avalanche, particularly on broad, uniform slopes. The parameters that influence the wave pattern are determined by an order-of-magnitu de reduction of the terms in the governing differential equation for visco-elastic slab response.

Principal problems concerning the raw data and methodological limitations of statistical and conventional avalanche forecasting methods are summarized. The concepts of four statistical models based on multivariate data analysis, are outlined in a few words. In order to give an idea of the potential and quality of the different methods, test runs over two winters are discussed and a tentative store is established. Statistical models I and IV, together with the conventional forecast, attain a score of 70-80%, whereas statistical models II and III show a slightly poorer performance.

A series of devices have been developed and tested which are designed to initiate the release of snow avalanches by mechanically disturbing or altering the snow cover. These serve as substitutes for the conventional use of high explosives and artillery projectiles. Canister charged with an explosive mixture of gases can be fired electrically by remote control and recycled to fire again as needed. Air bags placed underneath the snow cover can be inflated with reasonably low pressure to disrupt and dislodge the snow cover. Energetic vibrators powered by compressed air can dislodge unstable snow under favorable circumstances. A polyethylene membrane placed on the ground surface of a sleep mountainside effectively inhibits the accumulation of snow by allowing new snow to slide off as fast as it accumulates. All of these methods have led to the practical release of avalanches but none appear a universal a universal solution to avalanche release in all circumstances. The gas exploder system is the most promising to date and is currently being installed for operational use.

Spontaneous emissions of noise were monitored in a snowpack on an avalanche track using three geophones. 30 Hz noise was recorded on magnetic tape for the whole avalanche season and 3 Hz noise for a short period of time. Periods before and during avalanches were associated with high noise, while periods of known stability were associated with low noise.

Over the past few years an investigation has been conducted to determine the extent of seismic signals associated with avalanching snow slopes. A number of the signals recorded have been examined and classified according to their origin. One of these signals, however, is not clearly defined in terms of an observable source mechanism. To obtain information regarding the origin of this signal we have compared the results from several investigations conducted to study the seismic activity associated with glaciers. A comparative analysis of the snow and glacier signals indicates that the high-frequency signals observed in snow fields are due to internal fracture within the snow.

The method of the artificial release of slab avalanches by explosives is investigated. The effect of a detonation on a seasonal snow cover is measured and discussed. The attenuation of the resulting stress waves as a function of the type of explosive used, the position of the charge relative to the snow surface, the snow-stratification profile, and the type of ground are determined in field experiments. The effective ranges of shots in potential fracture zones are estimated.

Results of ten-year measurements of avalanche impact upon a large-sized installation are shown. The installation is equipped with 20 pressure sensors using the principle of the intrusion of a steel cone into a duralumin plate. The impacts of 30 dry avalanches with volumes from 100 up to 50 000 m3 are analysed. A maximum pressure at “a point" of 1 100 kN/m2 and a pressure of 650 kN/m3 averaged over the area of the installation (24 m2) were recorded. It is found that 75% of the impact energy is concentrated in the lowest 2 m of the installation. Proceeding from these data, and also taking into account that the thickness of an avalanche body amounts to 10-20 m at the moment of impact, we may conclude, that there is a “dense core" and an upper ’"high aerosol" part in an avalanche body.

The use of slereo-photogrammetric surveys of moving avalanches revealed the “velocity depression" effect in the Khibin avalanches: Avalanche velocity decreases abruptly (by a factor 1.5 to 2) at the end of transit channel and then grows rapidly up to a maximum at the end of the avalanche cone. The greatest variations of velocity amount to 51 m/s.

The data obtained prove the inefficiency of anti-avalanche filling dams, especially because of their negative role in the generation of disasterous air waves, whose pressure varies from 20 up to 150 kN/m 2.

Snow-drifts have been studied by many researchers both in field and laboratory conditions, however these investigations have been carried out mostly at wind speeds up to 20 m/s whereas in many areas of our planet snow-storms occur at winds up to 40 m/s and more. During the winter seasons of 1972-76 the authors carried out a great number of experiments with an artificial snow-storm in a special wind tunnel 27 m long. The wind speeds reached 40 m/s (60-65 m/s at the 10 m anemometer height). The existing theories and hypotheses of snow-drifting, and in particular the “diffusion" model, were tested in a series of the experiments. These have not confirmed the assumption of the Australian scientists on the decisive role of diffusion in drift mechanism at large wind speeds. The problem of strong snow-storm effect on snow accumulation on avalanche-danger slopes, in particular, wind redistribution of snow is no less important.

The results obtained may be used for the determination of snow accumulation in avalanche starting zones due to deflation. This is especially important for forecasting very dangerous and frequently-occurring avalanches due to snow-storms. The investigations performed enable us to estimate the snow deposition produced by strong and superstrong snow-storms, to account for the peculiarities of such snow-storms and the means of protection, to forecast snow distribution in mountainous regions, and to define the role of snow-storms in glacier mass balance.

Three basic areas of ice accretion are discussed: (i) the motion of the (supercooled) droplets in air up to collision with the icing object, (ii) the net growth rates, and (iii) the heat balance of an icing cylinder. The weakness of presently used approaches to establish collision efficiency and possible improvements are outlined. Experimental evidence on the icing of cylinders is then reviewed which establishes that the growth rates by icing are farless than expected because up to 80% and more of the initially accreted water can bounce off or be lost by shedding at air temperatures of –20°C. The reduced growth rates have a considerable effect on the heat balance of the icing object, whose temperature and water-phase mixture are controlled by the temperature and phme of the lost water-substance. These latter properties are completely unknown and require further studies. In conclusion the scientific aspects of icing are far from being understood; empirical findings may have to provide guidance for applications.

A rough comparison between thermal and mechanical methods of de-icing indicates that mechanical methods could potentially de-ice with an order-of-magnitude less energy than that required to melt an ice accretion. Two applications of mechanical de-icing using pneumatically driven inflatable de-icers are described in this report. The first of these was the de-icing of a small cylindrical radome used for air navigational purposes. Two seasons of testing were conducted with a de-icer consisting of an inflatable-deflatable flexible plastic covering. The de-icer was driven by tanks with pressure and vacuum reservoirs that were recharged by an on-site air compressor in response to a pressure sensor. The de-icing cycle was activated by an ice detector so the system responded to icing events on a demand basis driven by the ice detector. The system proved successful in keeping the radome free of ice without manned operation and with small energy consumption in a mountain icing environment. The second application was an attempt to de-ice the walls of locks used in river navigational facilities. Ice usually formed at the high-water-mark by the freezing of the water exposed to low air temperatures or by the pressing of ice against the walls by ships using the locks. The de-icers consisted of air-driven hoses mounted on the wall covered by a thick flexible rubber mat and protected from ship damage by steel outer plates. This method was successful in removing ice accumulations up to 2 m long by 0.3 m thick over the area covered by the de-icer. Installation costs and the necessity for protection of the de-icer against abrasion by ships may make this de-icing method prohibitively expensive compared with methods which are not as susceptible to damage by ships (e.g. chemical coatings and electrical heating cables buried in the walls).

This paper presents a study on the accretion of wet snowflakes transported by a strong cyclone on electric wires by means of wind-tunnel experiments and field observations with the aim of successfully preventing them from accreting on and causing damage to power lines and high-voltage power-transmission towers in snowy areas in such countries as Japan, Canada, and the U.S.A. Although extensive studies have been made on snow accretion in Japan since the 1930's by many workers, their studies have been restricted to the coastal regions of central Honshu facing the Sea of japan, which are typified by heavy wet snowfalls of the monsoon type in a weak wind of less than 3 m/s at a temperature between –1° and +1.5+C continuing throughout the day. However, it has long been known that a cyclone characterized by a strong wind exceeding 10 m/s that carries heavier and wetter snowflakes can cause snow accretion on electric wires to grow to 20–30 cm in diameter. Because its detailed accretion processes and mechanism had not been clarified, studies have been made on the growth processes of snow accretion, the adhesive stress of snow to a wire, the structure and texture of the accreted snow, coefficients of collision and collection of snow on the wire, and trajectories of individual snow particles impinging on or passing by a wire at various wind speeds. It has been confirmed that wet snow can accrete on wires at any wind speed between 0 and 20 m/s. Moreover, new techniques have been developed to prevent wet snow from accreting on wires, and these have proved to be effective in situ for any type of snow accretion in Hokkaido.

Based on many years of measurements and observations on the frequency and intensity of deposition of various types of rime which have been conducted by the Department of Meteorology and Climatology of Wroclaw University at the top of Szrenica, Sudety mountains, a hundred meters above the tree line, the essential factors influencing this feature have been determined and are discussed. The upper parts of the Sudety mountains belong to the regions with the most frequent and heaviest rime deposits in Europe, which often appears to be hazardous to farming. It has been found that intensity of rime deposition depends clearly on air temperature and wind velocity, and it is related to certain types of weather situations characterized by types of atmospheric circulation delineated by Konček and Rein (1971).

As the continental shelves of the Arctic become important as source areas for the oil and minerals required by human society, sea ice becomes an increasing challenge to engineers. The present paper starts with a consideration of the different fields of engineering which require information on sea ice with the tasks ranging from the design of ice-breaking ships to Arctic drilling platforms and man-made ice islands. Then the structure of sea ice is described as it influences the observed variations in physical properties. Next the status of our knowledge of the physical properties important to engineering is reviewed. Properties discussed include mechanical properties (compressive, tensile, shear and flexural strengths; dynamic and static elastic moduli; Poisson’s ratio), friction and adhesion, thermal properties (specific and latent heats, thermal conductivity and diffusivity, density) and finally electromagnetic properties (dielectric permittivity and loss, resistivity). Particular attention is given to parameters such as temperature, strain-rate, brine volume, and loading direction as they affect property variations. Gaps, contradictions in the data, and inadequacies in testing techniques are pointed out. Finally suggestions are made for future research, especially for more basic laboratory studies designed to provide the data base upon which further theoretical developments as well as field studies can be built.

During March-May 1976, a combination of laser and radar ranging systems was used to study the motion of both the fast ice and the pack ice near Narwhal and Cross Islands, two barrier islands located 16 and 21 km offshore in the vicinity of Prudhoe Bay, Alaska. Laser measurements of targets on the fast ice near Narwhal Island indicate small net displacements of approximately 1 m over the period of study (71 d) with short-term displacements of up to 40 cm occurring over 3 d periods. The main motion was outward normal to the coast and was believed to be the result of thermal expansion of the ice. The radar records of fast-ice sites farther offshore show a systematic increase in the standard deviation of the displacements as measured parallel to the coast, reaching a value of ±6.6 m at 31 km. The farthest fast-ice sites show short-term displacements of up to 12 m. There are also trends in the records that are believed to be the result of the general warming of the fast ice with time.

Radar targets located on the pack ice showed large short-term displacements (up to 2.7 km) but negligible net ice drift along the coast. There was no significant correlation between the movement of the pack and the local wind, suggesting that coastal ice prediction models can only succeed if handled as part of a regional model which incorporates stress transfer through the pack. The apparent fast-ice-pack-ice boundary in the study area was located in 30-35 m of water.

The understanding of the horizontal movement of fast ice is important for applied sea-ice mechanics. A case study, carried out in conjunction with a polynya known as North Water, is presented in this paper. The displacements of the fast-ire arches which separate the polynya from the surrounding ice-covered sea, were measured and found to be small. It is, therefore, confirmed that these arches prevent the influx of large quantities of sea ice into the polynya. The results are then explained in terms of the external forces (wind and current), the stress- strain situations and some physical characteristics (temperature and thickness) which were measured simultaneously.