Flow avalanches may be regarded as being composed of a granular fluid. When dislodged, the snow masses accelerate down a slope until the inclination of its bed tends towards the horizontal, at which stage bed friction eventually brings the snow to rest. We present a completely new analysis of the motion of a finite mass of granular material along an inclined base.

We regard a granular snow mass as an incompressible continuum to which a Coulomb-like basal friction law can be applied. Depth-averaged equations of motion are formulated in terms of a curvilinear coordinate system along a curved bed, and incorporate an averaged longitudinal velocity and a height distribution. A numerical finite-difference technique is employed to integrate these equations. We present numerical results obtained for motion along a curved bed and compare this with the solutions of the equations with results from laboratory observations. The experiments have been performed in order to monitor the motion of a finite mass of granules, either plastic particles or glass beads, along a chute consisting of both an inclined and a horizontal zone, the two zones being connected by a curved element. The particle spread along the chute and the mass distribution of the granules in the run-out zone, as obtained from these experiments, are compared with those derived from theoretical computations. The results show that the model used predicts the motion of a granular avalanche reasonably well.

Finally, it is indicated how the basal friction law may be extended or altered in order to reproduce the dynamic processes involved in causing the abrupt cessation of the snow mass-motion characteristic in the run-out zone.

A new visibility-range measuring system has been developed which uses a video camera, and with which visibility can be estimated by measuring the brightness contrast of a black target against a standardized background. This principle has been applied to development of the new visibility-range monitor described in this paper. In this monitor, a charge-coupled device (CCD) video camera measures the brightness of an arbitrary area in its range of vision, and visibility range measured by the video camera in blowing snow can be compared with that measured by the transmissometer-type visibility-range monitor and can also be calculated by direct observation using optical targets.

The new sytem can monitor visibility and record information about traffic on highways at 1 s intervals for a maximum of six separate positions. The vertical distribution of visibilities over a snow surface has been observed, and it has been found that the fluctuation in visibility is greatest at lower levels above the ground in blowing snow.

Laboratory and field experiments on hardness of snow have shown that the free water contained in snow decreases its hardness, and that solar radiation further decreases hardness down to a value below the limit of that which would result only from the influence of the water content of snow. A quantitative relationship between the amount of solar radiation absorbed by snow and decrease in snow hardness was derived. Thin-section analyses of snow were used to reveal the mechanism of decrease in snow hardness which had been caused by solar radiation.

Water flow through natural snow-packs is generally accepted to be extremely variable even over short distances. Differences in snow properties, and discontinuities in the horizontal and vertical structures of snow-packs, are responsible for variations in the effective area of the snow surface associated with outflow from a part of the base of the snow-pack. At the Central Sierra Snow Laboratory snow-pack outflow has been monitored using four snow-melt lysimeters of 6 m2 area each; two of these collectors were subdivided into three sections of equal area.

Substantial variations in daily and monthly outflow volumes were noted between the individual collectors and also to a greater degree between the 2 m2 sub-units of the divided containers. The relative collecting efficiency of the snow-melt lysimeters also changed over time. In the deep warm snow-packs of the Sierra Nevada forest zone, snow-pack water routing appears to be highly variable, and this variability is on a much larger scale than was previously recognized. The large differences in snow-pack outflow measured in this study suggest that structural discontinuities and microtopography of snow strata can redirect percolating water over an area of several square meters.

To make clear the structure and behaviour of a large-scale avalanche, the impact force-data obtained in the avalanche project of 1972–78 were analysed in detail. The wave forms of impact forces are classified into two types. Type 1 is composed of many separate spikes each of which represents the collision of a snow block. Type 2 has wider peaks, caused by collisons of snow blocks mixed with fluidized snow. Most of the type 1 peaks were in the width range corresponding to 0.005–0.01 s duration, and most type 2 peaks fell into the 0.02–0.1 s range.

The internal velocities of an avalanche were estimated by calculating cross-correlation spectra for a time series of impact-force records. It was discovered that these internal velocities varied from 10 to 50 m/s over time. The mean distance between snow blocks was found to be in the range 1.6–5.4 m in a type 1 avalanche, and between 0.7 and 3 m in type 2 avalanches. Sizes of snow blocks or snow clouds of type 1 and type 2 were in ranges 0.26–0.52 and 0.37–1.9 m, respectively.

This paper also reports on the project created to initiate artificial powder-snow avalanches in the Shiai-dani area and to make systematic observations of a variety of physical aspects. Results obtained in 1988 for both artificial and natural avalanches are given.

An array of radio echoes has been collected on Storglaciären in Sweden and used in an aperture-synthesis experiment. The digitized echoes were combined in a computer with suitable time shifts to focus on regions beneath the ice surface. The positions of six individual englacial cavities have been found, and the echoes from them isolated for interpretation. By performing the aperture synthesis with data sets from three different antenna configurations, information on the cavity shapes and orientations has been derived. At least five of the cavities are believed to be water-filled, and may be parts of an englacial drainage system. The technique could be applied to the problem of finding the principal water channel(s) in or under a glacier. Successful location would be of considerable importance in the planning of subglacial water intakes for hydro-electric power production.

The fracture of a slab avalanche is a multi-phase and progressive process. The different kinds of fracture and possible scenarios of avalanche release in the form of a zip effect are shown. In the course of investigations, most importance has so far been attached to shear failure along the sliding surface. Various cases of load and their effects on the stresses, on the changes of strength, and on the stability of the inclined snow-pack are discussed. The usual simple model of the shear-stability index is unsatisfactory. The present paper deals with the complex interaction of all supporting forces of a snow slab by means of simplified geotechnical considerations. For this purpose, the acting and reacting forces of a “standard avalanche” (i.e. dead load with driving and normal component, shear force, tensile force, compressive force, and flank force) are estimated from published boundary values. Using different combinations (e.g. hard slab with high circumferential forces on a weak shear surface with low shear force), it can be shown that suspension at the crown and lower and lateral support are of great importance. This applies especially to cases with low shear forces and, consequently, with low overall stability. Despite the fact that the circumferential area of the model avalanche is only 6% of the area of the shear surface, the circumferential force in this case is more than 150% of the shear force. In a parameter study with different avalanche sizes, these results are generalized and confirmed. For the assumed strength limits, critical areas and depths of possible slab avalanches can be derived. Although the supporting shear force is the major contributor to stability, particularly with larger slabs, it can be seen from the investigations that the redistributions of stress and spatial supports and suspensions of the whole slab avalanche must not be neglected in stability analyses.

Some preliminary experimental results are presented, which characterize the behaviour of small-scale artificial avalanches. The laboratory simulation consisted of releasing a finite mass of non-cohesive granular material to flow without restriction down an inclined surface which includes a curved part at the base where the run-out zone is encountered. Three contrasting materials, of 3 mm and 5 mm diameter size sortings, and also two different bed roughnesses were used.

Results indicate that a theoretical model which presumes a flow-regime-type avalanche may be too restrictive, in the sense that it describes a limiting case which does not often occur even under controlled conditions. Some suggestions are made for amendments to current theoretical concepts.

Snow-pressure measurements have been carried out on two masts at the NGI avalanche station in Grasdalen, western Norway. These two tubular masts have diameters of 0.22 and 0.42 m, respectively, and are situated on a 25° slope with a deep snow cover.

The most important conclusions are that within a homogeneous snow-pack there is a close correlation between snow-creep pressure and the product of acceleration due to gravity, g, density, ρ, and snow depth, H, that the highest pressures are recorded in late winter when the snow-pack is at the 0°C isothermal, and finally that a weak 0° C isothermal snow layer at ground level appears to increase snow pressure.

In 1980 the new advance of Glacier Blanc covered a trail leading to a mountain hut and threatened the footbridge that crosses its melt-water stream. This event prompted the Laboratoire de Glaciologie et Géophysique de l’Environnement (LGGE) to undertake a new study of the glacier. With the old Eaux et Forêts measurements, and the more recent LGGE ones, it has been possible to forecast the probable fluctuations in glacier length by two different methods: a statistical method based on a correlation between the variations in length and mass balance, and a one-dimensional numerical ice-flow model. Both methods show that a substantial glacier retreat is unlikely during the 1985–90 period.

Digital maps and terrain models are used to calculate maximum snow-avalanche run-out distance based on topographic parameters. Maps of 1:50 000 scale are found to be accurate enough for the purpose. 113 well-known avalanches are discussed in this paper. A computer system is used to calculate terrain parameters such as rupture area (A), avalanche-path length (L), avalanche-track lengths (L1L2), and run-out lengths (L3). Maximum run-out angle (α), avalanche-track angle (β). and average angle of rupture zone (γ) are also found by computer. The use of computer and terrain model reduces subjective judgement of parameters to a minimum. Run-out distance was found to be best expressed by the regression equation:

α = 0.91β + 0.08γ–3.5°. R2 = 0.94, S = 1.4°.

L = 0.93L1 + 0.97L2 + 0.61A + 182 m. R2 = 0.96, S = 137 m.

Mapping of areas exposed to avalanche hazard for the Norwegian Army was started by the Norwegian Geotechnical Institute in 1986. The background to this mapping is that large-scale military exercises are held annually in northern Norway, in terrain where there is a high danger of avalanche activity. Avalanche areas are divided into two zones: potential starting zones, and potential run-out zones. All potential avalanche areas are indicated on maps, and mapping is carried out by computer using a terrain model and digital maps. An interactive graphic work station is used to outline danger areas. Starting and run-out zones are identified by using terrain parameters which may be extracted from digital maps. The usual scale of the avalanche maps is 1 : 50 000, with 20 m contour intervals.

Structures placed in deep snow covers are subject to forces caused by interruption of the down-slope snow-pack deformation components. The resulting creep pressures are often the primary design consideration. In this paper, accurate field data (pressures) and theoretical analysis of the problem using a linear creep law to define snow deformation are presented. Results include analytical expressions for the pressures, and it is demonstrated that the resulting linear theory underestimates the mean pressures by about 20%. Higher accuracy will require that a nonlinear deformation law be formulated.

Avalanche run-out distances have a prominent statistical character which is not usually featured in calculations derived from avalanche dynamics models. In order to predict run-out distances in a rational manner, the statistical element should be included. Here, we present and analyze data on extreme run-out from four mountain ranges. When the data are fitted to extreme-value distributions, we are able to show how (and why) extreme-value parameters vary with terrain properties of the different ranges.

Slope instability and occasional devastating landslides are well-known hazards in high mountain areas. This paper describes and discusses an example of extensive and recurring damage associated with agricultural settlements around the lower reaches of the rapidly flowing Bualtar and Barpu Glaciers in northern Pakistan. These landslides occur over a zone about 20 km long in response to erosive processes at the ice-slope interface, and slowly descend 150–300 m from the edges of cultivation to the glacier margins. Damage is evident in the loss and/or abandonment of approximately 10 km2 of land, and in the destruction of dwellings and irrigation channels. The daily routine of local villagers is affected because alterations of both the slope and the ice surface destroy frequently used transport routes. Although the landslides have a history decades long, the landslide problem has more recently assumed heightened significance in relation to rapidly occurring economic and social change such as the introduction of wage labour and seasonal outmigration.

Hubbard Glacier, the largest tide-water glacier in North America, has advanced since it was first mapped in 1895 by moving a protective submarine moraine into the entrance of Russell Fiord. In May 1986, a weak surge of the Valerie tributary of Hubbard Glacier caused the glacier to block the fiord entrance, converting the body of water into a large glacier-dammed lake. This lake filled to a height of 25.5 m and stored 5.4 km3 of water before it burst out on 8 October 1986, producing a peak flow of 105 000 m3 s−1 averaged for 1 h.

Hubbard Glacier is expected to continue advancing because its accumulation area ratio (AAR) is 0.95, which is unusually large. Such an advance would undoubtedly block Russell Fiord again. If this happens, it is predicted that the lake will fill to a height of 39 m over a period of 1.1–1.5 years and then overflow into the Situk River near Yakutat. This, in turn, would increase the average flow of that small stream from ils present rate of between 10 and 15 ms−1 to an estimated annual average discharge of 230 m s−1. Such an increase in flow would be expected to flood and erode forest lands, fish habitats, subsistence fishing camps, archaeological sites, and roads. At the same time, the increased water depth in Russell Fiord could be expected to increase the calving rate of Hubbard Glacier, potentially threatening the stability of its calving terminus.

In a local range, crest winds were compared with winds at lower stations to make it possible to initiate a drift-transport model which would predict snow accumulation patterns on leeward slopes. Corrections to the model input were made after consideration of detailed drift-flux measurements in the lowest 2 m above snow surface. Good agreement was found between the total length of large avalanches in a path near the crest, the appropriate wind reading and the corrected snow-depth increments in the rupture zone. Control of medium-sized avalanches likely to cause injury to skiers can be improved with the proposed method.

Mini-avalanche systems were constructed both in a low-temperature laboratory and in a snowfield, and the behaviour of the flowing snow was observed in each case. Velocity profiles for the individual snow particles were determined and these implied that a viscous force, which has been neglected in most previous numerical simulations of snow-avalanche motion, needs to be taken into account for many avalanches. Kinematic viscosity coefficients for the fluidized snow were also measured using a modified Stormer-type viscometer. Substituting the dry-friction value and the kinematic viscosity coefficient for fluidized snow into the equation for avalanche motion, numerical simulation of natural events was achieved for the Shiai-dani region. Taking viscous resistance factors into account led to the conclusion that the magnitude of turbulent resistance of snow in avalanche systems is probably much smaller than that represented by the values previously in use.

The internal structure of powder-snow avalanches was investigated at Kurobe Canyon in the Shiai-dani area of Japan in 1988. Internal velocity was derived for avalanches of this kind by frequency analysis of impact-force data, and was found to undergo a remarkable change with time. The shear of the avalanche flow was estimated to range from 1 to 7 s−1. The front region of the avalanche wind was observed to precede the front of the avalanche by a distance of 17.3 m. The maximum wind velocity was comparable with the internal velocity of the front region of the avalanche.

In this paper a fundamental model is proposed to describe the time-dependent behaviour of uniform snow glide on an infinite slope. This model is composed both of a mechanical balance equation relating a driving force and a resistant force, and of a rate equation for a real contact area at the boundary. In the model, a steady motion of glide is described as a stable equilibrium of glide motion, and a non-steady motion is considered as a transient one not in equilibrium. Finally, the model is applied to a full-depth avalanche release caused by acceleration of snow glide. By comparing the model predictions with glide-velocity field data, a proper model of acceleration prior to full-depth avalanche release is determined. As a result, we can obtain the safety standards to be applied in the event of full-depth avalanches in terms of a glide velocity.