In a static supercooled cloud dendrites hardly grow at about -15°C except at the pointed tip of a needle-like ice crystal or an isolated thin plate-like ice crystal. When ice crystals are moved slowly in a static supercooled cloud, dendrites grow at about -15° C and the α-axis growth rate increases as the velocity of the dendrites increases; at velocities higher than 20 cm/s, however, the a-axis growth rate decreases as the velocity increases due to the influence of heavy riming. The maximum a-axis growth rate in a supercooled cloud is observed at about -15°C in experiments growing ice crystals in free fall.

A new calorimeter was designed for measuring the free water content of wet snow, aiming at simplicity in mechanism and handling as well as high accuracy and speed in measurements. The mass of hot water to melt the sample and the mass of the mixture (snow sample and hot water) are measurable indoors with sufficient accuracy; therefore, in the field we need only two measurements of temperature (T and T ); this calorimeter reduces the time of measurement of one snow sample to four or five minutes. Using artificially made wet snow, having a known value of free water content, 121 tests were conducted. Standard deviation of the error was 0.82%

An estimation of average mass balance of a high hanging glacier in the Swiss Alps was made by measuring volumes of ice avalanches originating from this glacier. Ice avalanches are this glacier’s predominant form of ablation. Since the volume of the glacier has not noticeably changed over the past few years, the annual ice loss due to ice avalanches can be taken as an indication of average total net acumulation above the ice cliff where the avalanches originate. The mass balance value, as determined by recording ice avalanches, compares well with values obtained by independent methods (measurements of firn stratigraphy in the cliff, direct accumulation measurements in the vicinity). No seasonal variation in the frequency of ice avalanche occurence was detected.

Information on snow conditions in high mountain river basins is of vital interest for flood predictions and power production. Based on techniques derived for mapping of snow cover from digital NOAA-data, relations are established between snow covered area and remaining snow storage for three basins in southern Norway. Together with estimates of the precipitation and information on maximum accumulated snow, the relation can be useful in run-off predictions for the snow-melt period.

This paper describes a method of distinguishing between pack ice and sea clutter in radar echoes, an attempt to roughly estimate the thickness of sea ice from measurement of surface temperature by air-borne infrared radiometer, and an application of thermal images from satellite data to estimate the concentration of sea ice off the Okhotsk Sea coast of Hokkaido.

A flat condenser in strip-line technique has been developed to measure the gradient in liquid water content near the snow surface. This dielectric sensor is connected to a twin-T-bridge with an operating frequency of 27 MHz. The special design of the sensor allows the non-destructive measurement of the mean dielectric constant of underlying material at two different depths of 0 to 2 cm, and of 2 to 4 cm below the snow surface. Based on the model of Polder and van Santen, changes in the gradient of liquid water content can be calculated from measured changes in the dielectric constant in different depths; the density has to be measured separately. Results of measurements of wetness gradient with this flat capacitative sensor are given. Field measurements have been carried out in the Stubai Alps (3100 m a.s.1.) and at the Hafelekar (2256 m a.s.l.) near Innsbruck.

The glide of a snow cover on a slope covered with bamboo bushes was simulated so that the bushes’ resistance to glide was quantitatively related to the glide distance and speed. In one case their stalks and leaves were kept in the snow without falling down, in a second case they fell down and lay under the snow. Using the relations obtained, a description was given as to the behavior of a snow cover on a uniform slope as well as the distribution of stress and glide speed immediately before crack formation. Then, a criterion of crack formation was given in terms of glide speed.

Studies on the snow-layer stratigraphy were carried out at the Gyajo glacier, Khumbu region, East Nepal from 1973 to 1978. The ages of the snow layers were determined by stratigraphy and tritium analyses, and the fluctuations of sedimentary environments of the Gyajo Glacier, the lower type of a glacier located below 6,000 m in the Nepal Himalayas, were clarified from 1940s to 1975 in relation to the climatic changes. The results are: (1) It is likely that one snow layer per year is preserved divided by dirt bands: there are depositional periods ranging for 4-9 years, interrupted by depositional gaps (lacuna) lasting several years. Such stratifications contrast with such higher glaciers as the Khumbu Glacier, where two snow layers per year are formed with continuous deposition. (2) The ages of snow layers were determined by variation of tritium concentrations due to the thermonuclear tests series of 1950s-1970s. From 1940s to 1960s, the net annual balances have a decreasing trend from 450 to 45 mm, while annual balances are relatively large at the start of the 1970s, ranging from 540 to 270 mm, due to the increased precipitation.

Observation on the isotopic composition of falling snow particles was carried out at two points 30 km apart in the coastal area of the Sea of Japan, between 20 January to 25 February 1983. Sampling periods of snow particles were from 10 minutes to several hours. Time variation of 6 hours running mean of δ18O values showed good correlation with the variation of surface air temperature. The δ18O value is maximal at the beginning of a snowfall, except when large graupels were observed at the beginning. Comparison of the results between the two observation points showed that the differences in the δ18O values are great when precipitation is plentiful and precipitation cells are small.

A repeated compression and annealing experiment was conducted using polycrystalline ice samples. The sample was compressed in steps of 10% at a constant low deformation rate at 272 K., and annealed at the same temperature, thus allowing the sample to recrystallize. As a result, the ice gradually changed to a fabric pattern of the 4-maximum or 3-maximum type. The result indicates that repetitive recrystallization plays an important role in forming the preferred orientation of ice in glaciers.

The effect of surrounding luminance on visual range in blowing snow condition was investigated in daytime, by comparing visibility derived from luminance contrast with that derived from transmissivity observed simultaneously. A volume of air sampled for monitoring may not accurately represent the atmosphere from which it was taken. Differences in measurements through two types of visibility meters increased from daytime to twilight, and with cloudy or lightly blowing snow conditions in daytime. In heavy blowing snow, visual range measured by transmissivity coincided with that by luminance contrast.

We was concluded that visual range obtained by transmissometer was inclined to be over-estimated in conditions of cloud and light blowing snow.

Dependence of Kinosita’s hardness R of wet snow on its dry density ρd and free water content W was obtained in the following form where a, b and c are constants. When wet snow was qualitatively classified into two types, I (coarse-grained granular) and II (new and fine-grained compact) and the constants were determined separately for each type, the correlation coefficients of the above relation were high except between In R and W for snow I. For further investigation of the effect of W on R, a dry snow sample was prepared and R was measured before and after water addition. Aging effect on R was also investigated.

Relations between the structure and falling motion of early snowflakes, ie those composed of a few crystals, were investigated to clarify their growth process.

The following results were obtained: (1) Two crystals of the same shape in a snowflake are similar in size, but those of different shape are considerably different in size. (2) In snowflakes having two crystals of the same shape, one crystal is attached near the center of the other; where the two are of different shapes, a smaller crystal is attached near the tip of the other. (3) The nondimensional amplitude of the spiral falling motion of snowflakes composed of two crystals of different shape, is slightly larger than for those with crystals of the same shape. This suggests that snowflakes composed of different shapes are more likely to aggregate to other snowflakes or crystals. (4) In snowflakes consisting ?G three crystals, the centers are usually arranged in a straight line. (5) The nondimensional amplitude of the spiral motion of snowflakes consisting of three crystals increases with departures from the straight-line arrangement of their centers.

Macropores - large open channels that can conduct water - have been found within snowpacks of the Sierra Nevada of California. Extensive networks of channels as well as apparently isolated single channels have been documented. These openings, which are usually found near the base of the snowpack, were recognized at several widespread sites. Such basal channels may be an important means of draining liquid water from the snowpack. The presence of open conduits in snow helps to explain such phenomena as rapid streamflow response to percolating rain and melt water.

An experimental study on the preferred growth of grains in sea ice was performed. As a seed an artificial bi-crystal was prepared by welding two single crystals; the c-axis of one grain was in a horizontal plane, that of the other made an angle of α with the horizontal plane, and they intersected at right angles. An angle between a resulting grain boundary and the vertical line, Θ, was measured on thin sections of sea ice after it grew about 10 cm downward in sea water. The relation between a and ? was given by ? = 0.4α + 6, when a was varied from -25º to 25º. When α = 0º , that is c-axes of both the grains are at right angles on a horizontal plane, encroachment still occurs, which is contrary to the earlier studies. These results can be explained by the difference in the horizontal component of growth rate between neighbouring grains at the grain boundary groove.

The melting rate of the bottom of a snow cover during the winter and its variation with time were studied in relation to heat exchange at the ground-snow interface. Continuous observations were made of bottom-melt, using a lysimeter, during the past four winters beginning December 1980, at our test field in Moshiri, Hokkaido, which is known for severe coldness and deep snow.

During the four months in this area the amount of bottom-melt ranged from 45 - 70 mm in total and from 0.4 - 0.6 mm per day on average. In winters with much snow the daily amount of bottom-melt decreased gradually from a maximum of about 1.5 mm.d-1 in early December, reached its minimum of 0.2 - 0.3 mm d-1 day in February, and then increased again. In a winter with a smaller amount of snow it varied greatly with time during the early period of the winter and became zero temporarily. Such variations of the melting rate could be explained by conductive heat fluxes observed in the ground and snow. The daily average of the former was about 5 W.m-2 at a depth of 0.1 m in early December and decreased to about 2.5 W.m-2 in early April.

This paper deals with compressive strength vs temperature characteristics (-40 ºC to -2ºC), size effect (cross-sectional area of test specimen) and probability distribution of compressive strength and fracture toughness KIC (corresponding to notch effect in bending strength), of first year sea ice sampled from Saroma Lagoon. The main experimental results are as follows. (1) Temperature dependent compressive fracture modes: at temperatures of -20 ºC to 0ºC, crush fracture is dominant, and at temperatures of -40 ºC to -20 C, brittle fracture is dominant. (2) The larger the cross-sectional area of a test specimen, the lower the compressive strength becomes. It is inferred that the number of weak spots increases with increase in the cross-sectional area of the test piece, which thus becomes more susceptible to fracture. (3) The scatter of compressive strength can be approximated by a normal distribution curve. (4) The KIC values of sea ice are below 10 kg/cm2cm, though they change slightly depending on the dimensions of test pieces, the relative angle between crystal growth direction and load direction, environmental condition (air or water), and testing method.

An automatic ice fabric analyser has been developed which precisely determines both c- and a-axes of the individual grains in a thin polycrystalline ice specimen by the use of X-ray Laue method (eg Cullity 1956). The method of analysis is computer-aided indexing of the zone-axes in X-ray Laue patterns. Crystal orientations can be automatically determined by the analyser with an error of approximately 0.6_. Time for analysing one specimen containing 100 grains is approximately three hours at present.

Settlement force of a snow cover acting on horizontal beams O.8 m high on the ground has been measured using electric force gauges since winter 1974/75, to understand mechanisms of settlement force generation. Shapes of cross-sections of the beams were rectangular, square, equilateral triangle and triangular with a 37_ vertical angle. From the measurements, it was found that the magnitude of the settlement force acting on a beam did not depend upon the shape of cross-section when snow depth was smaller than 1.3 m, but in deeper snow it was smaller on the triangular than on the rectangular beam. When the surface of the snow cover turned into a crust, the settlement force increased from about 0.3 kN m-1 to 1.1 kN m-1. Physical properties of snow cover are important for quantitative estimation of the settlement force on a beam.

Sorting effects of snow particles during deposition were studied in field observations of snow cornices and wind-tunnel experiment of snow drifts. Grain size, density and hardness were larger at the upper part (root) of a cornice than at the lower part (scarp). Experiments with the use of a horizontal step in a wind-tunnel revealed the importance of redistribution of snow particles on the formation of drifts: at wind speed of 5 m/s, about 40% of particles fallen onto the surface behind a 0.1-m high step were removed by erosion, rebound and creep. The surface mass balance controlling the growth and shape of a drift are briefly discussed.