We have derived an analytical expression for the growth rate of sea ice by taking account of the processes relevant to the growth, eg heat conduction, diffusion of salt molecules, radiation, sensible heat transport, evaporation and so on. We discuss the role of each process as rate determining processes under various environmental conditions. It is shown that because of coupling of salt diffusion and heat conduction, the growth rate feeds back to the heat flux Qw from water to ice which controls the growth rate and that Qw decreases with the thickness I of sea ice, even if the environmental conditions are kept constant.

Heat budget studies of the sea ice cover near Pond Inlet, NWT, were made using data obtained at two locations in Eclipse Sound, one about 0.5 km from shore and the other about 7.5 km from shore. The observations at intervals of one week included ice temperatures at 10 cm separation in vertical profile, salinities of adjacent 2.5 cm-thick slices from vertical ice cores, and ice thickness. The time series analysed extend from three to six months in the six data sets obtained for three winters of observations. Values of oceanic heat flux have been determined as residuals in the energy balance equation applied to the ice cover. The results show that in Eclipse Sound the oceanic heat flux is a significant component of the heat budget of the ice cover. Its value over the winter is typically about 6 W m-2 about half as large as the average rate of release of the latent heat of freezing. There does not appear to be any systematic variation in value of the 4 week-average oceanic heat flux during the season. Nor is there any apparent correlation of oceanic heat flux with rate of release of latent heat (ie ice growth rate), or with the severity of the winter as measured by the magnitude of the conductive heat flux.

Each component of the heat balance equation was obtained independently for 24 days in winter and 8 days in summer in 1980 at Mizuho station, East Antarctica. In winter, net radiation QNR was -37.6W/m-2 , QS; +36.7W/m-2 and QC; +2.5W/m2 QL was three orders less than QS. In the summer, QNR; +l9.9W/m-2 QC;-7.4W/m’, QL ; -.SW/m-2 and Q C; .2W/m-2. In the winter cloud amount was an important factor determining the variation in heat balance components, but also variation in the strength of the katabatic wind had effect. Condensation of water vapor occurred in winter and sublimation in summer: summer sublimation had a significant effect on the heat balance. The small condensation may be due to the structure of the temperature inversion at Mizuho which is related to the katabatic wind. The present results show that at Mizuho, the radiation loss Q NR and Q S which compensate it is larger than at any other site on the continent excluding the coastal stations.

North Water is situated in the high Arctic between Greenland and Canada (74°-79°N, 70°-78°W) The sea surface of this region has no homogeneous ice cover during the winter and spring months, due to the occurrence of polynyas

The results of the heat balance and the airborne remote sensing over North Water during winter are reported The heat balance investigation was concerned with the identification of heat sources and their magnitudes The remote sensing was aimed at determining the sea surface temperature and clarifying the exact nature of the surface of North Water in winter

North Water was found to be mostly covered with new and young ice and probably devoid of first year ice until March The surface was about 20 °C warmer than that of the surrounding fast ice The warmest areas were frequently located in Smith Sound, Lady Ann Strait and around the Carey Islands Open water patches above freezing point, ranging from -1 5°C to -0 1 °C were often found along the west coast of Greenland

The surface of North Water receives heat from the sea water during winter between October and March at the rate of 15 MJm^d”1 of which 80% and 20% are from enthalpy contained in sea water and latent heat of fusion released as the surface water freezes, respectively. The basic difference in winter heat balance between North Water and other ice covered arctic seas is the magnitude of the enthalpy supplied by the seawater which is 9 MJn-2d-1 larger for the North Water This is the main heat source, which keeps the surface from becoming solidly frozen

Water temperature and salinity profiles were measured to a depth of 300 m below a fast ice cover near Mawson, Antarctica over a full annual cycle. Together with measurements of ice thickness and salinity, they are used to determine the heat and salt balance of the ice/ocean system at this site. The energy balance of the ocean is related to measured energy fluxes at the surface.

Throughout the winter there is a net advection of salty water to the site which enhances the salinity increase in the water due to brine ejected from ice. After the ice reaches its maximum thickness there is considerable advection of warmer water which both raises the water temperature at the site and provides heat for the large oceanic heat flux previously reported for Mawson. The rate of this heat advection increases as the ice extent around Antarctica decreases. The ice partially melts in situ and breaks out in mid January. This effective removal of fresh water is balanced by a large influx of melt water from the continental ice sheet. The fresh water, initially near the surface, becomes well mixed to depths of greater than 200 m by strong storms in the ice free period from mid January to early April.

We report new data on the position of ice edges in the eastern and southern Weddell Sea for the years 1983 and 1984. The data are derived from ship-borne radar measurements of individual points along the ice edge together with ship’s positions obtained by a satellite navigation system. They are accurate within 0.23 to 0.4 nm (426 - 741 m). Comparisons of ice shelf margins for the years 1980, 1983 and 1984 allow estimates of apparent ice advance rates during this period. Together with quantitative ice edge velocity estimates first conclusions about net changes along the ice front and the ablation along the margin of ice shelves in the eastern and southern Weddell Sea are derived.

Movement of pack ice off the Okhotsk Sea coast of Hokkaido was investigated using combinations of sea ice radar photographs and Landsat MSS imageries. The sea ice radar network, consisting of three C-band (5.54 GHz) radar stations, covers an area of about 60 km across and 250 km along the coast. As radar echoes display not the shape of ice floes but the roughness of the ice field, the shapes of floes were drawn on a radar photograph overlaid upon a simultaneous Landsat Fig. 1.The coverage of the sea ice radar network. imagery. Each floe was then traced on the sequential photographs of radar display. The path of each floe frequently indicated a trochoidal oscillation of 18-hour period which is close to the inertial period of this area-Such paths were examined as representing the motion of inertial circle transported upon a long-period movement. The parameter v/U indicates the magnitude of meandering movement of an ice floe within the inertial period, where v is the circumferential velocity of inertial circle motion and U is the average velocity of a main drift in the inertial period. Values of v/U were obtained in a wide range from 0.4 to 8.3 for 18-hour trochoidal paths sampled.

This paper describes the details of a field study on the deterioration of two icebergs grounded outside St John’s harbour in Newfoundland, Canada. Observational data was collected during the period 10-25 June 1983, and included berg-related, meteorological and océanographie data. The study indicated the need for a stable observation platform to enable accurate measurements of iceberg profiles.

The observed decay of the two icebergs is compared with simulations from a model that predicts mass losses due to insolation, buoyant vertical convection, forced convection in air and water, wave erosion and calving of the resulting overhanging ice slabs. There was good agreement between observations and model simulations with the model underestimating the mass losses by about ten percent. Other salient features noted during the field study are also discussed.

This paper describes characteristic features of brine channels observed both macroscopically and microscopically using samples of natural sea ice. Systematic observations of brine channels, including the effects of ice growth rate and ice thickness, provide knowledge of their spatial distribution in young sea ice.

The dielectric constants of alpine snow samples with different stages of metamorphism and with different liquid water saturations have been measured in the frequency range of 10Hz to 50MHz using a plate condenser and network-analyzer. The limiting static dielectric constant, €s’, has been derived from the measured frequency dependence of the complex permittivity of snow by a least-square-fit using the model of Cole-Cole. A strong dependence of e on porosity, liquid water content and on the shape of the snow grains was found. Calculations of shape factors from the measured static permittivity based on the model of Polder and van Santen are given, and are compared to shape factors derived from an analysis of photographs of the snow samples.

Snow stratigraphy was obtained in the laboratory and field, using a FM-CW radar system having microwave frequencies ranging from 2-8 GHz and from 6-12 GHz. A multi-layered model consisting of artificial snow such as glass beads or polystyrene plastics was used in the laboratory. In the field, the thickness of an individual layer within a snowpack was determined by analyzing the profile of responses from interfaces within the snowpacks. Large anomalies of the responses which produce misleading results, were found to have been caused mainly by multi-reflection between stratified layers as well as residual mismatch reflection in the system’s components.

The physical properties of a snowpack were measured by digging a snow pit. The comparison of the microwave responses in the profile with the visual stratigraphic layering of snow was made by inserting a reflection plate between individual layers within the snowpack,

Topography and snowdrifts may cause large variations in the snow cover as well as in snow depth from one year to another. A simple model is developed to study the influence of different snow distributions and the importance of this as a source of error. The ground surface “seen” from the detector will appear as a disc that can be divided into a number of small elements where it is possible to place the wanted snow distribution. Calculations of the gamma radiation field with different snow distributions show how small changes in the snow cover and distribution will influence the measurements as a function of the average snow water equivalent.

Principal stresses in a snow cover on a uniform slope were determined by two methods, each using thin pressure gauges to measure snow pressure in the snow. These snow pressures were principal stress σ2 on a vertical plane perpendicular to the contour lines and normal compressive stress σ θ on a plane perpendicular to the vertical plane. In addition, plastic Poisson’s ratio v was estimated in a snow cover on level ground. Estimates of principal strain rates were used to calculate principal stresses and viscosity by two different methods, using estimates of v and the constitutive equations of Yosida (1980) and the derived values of σ2and σθ. For dry and compact snow, σ1 and σ3 calculated by both methods agreed well with each other, and also with values obtained by the hole-mark method reported by Shimizu and Huzioka (1975).

When structures such as oil drilling rigs are constructed on or through the ice plate in coastal and offshore regions, the shear strength of sea ice must be estimated to determine the loading on these structures. Testing methods for shear strength must be established so that shear strength of sea ice in various conditions can be determined.

The authors have been conducting, for five years, direct shear strength experiments using sea ice samples from the Okhotsk Sea coast. Physical characteristics of sea ice, including shear strength, depend to a great extent on such properties as salinity, porosity, grain size, etc; thus, there is variation in the test results of five years experimentation since the samples obtained varied from year to year.

The following conclusions were drawn from this experiment:

• i) Under certain conditions the relation between shear strength and vertical stress can be represented by Coulomb’s equation of soil; ie, where Ts: shear strength, C* : apparent cohesion (shear strength at σv σv = 0), φ*: angle of internal friction, σv : vertical stress

• ii) The shear strength of sea ice increases, approaching a constant, with decreasing ice temperature.

• iii) The shear strength decreases with increasing ice shear area; an analogous relation exists in concrete.

• iv) The shear strength is not greatly dependent on either the shear velocity or stress rate.

• v) The shear strength is greater and generally increases more rapidly with decreasing ice temperatures in planes perpendicular to the ice growth direction than in planes parallel to it.

• vi) Two types of failure occurred in the sea ice samples. In the case of (sample diameter)/(sample length) less than 2 , the failure was induced by shear only. With d/l>2, the failure was not solely caused by shear since the existence of a small gap between the ice sample and shear box introduced a bending moment.

The morphology of snow crystals growing at a low temperature has been experimentally studied. The habit and the morphological instability of the crystals vary remarkably with air pressure. In addition, the morphological instability of the crystals depends not only on air pressure but also on supersaturation, crystal size, the ratio of growth rates and the ratio of axial lengths. It is supposed from the experimental results that long prisms with small skeletal structures forming at low supersaturation are precipitating in polar regions.

To compare results of icing studies conducted in wind tunnels with natural icing conditions, a series of rotor icing studies were made on top of Mt. Washington, New Hampshire. The results indicated that considerable differences exist between the two under conditions of similar liquid water content and temperature. The wet-to-dry growth transition temperature, for instance, with comparable temperature and liquid water content, may be more than 10°C higher under natural conditions than in wind tunnel studies. The possible cause of such discrepancies was found to be the vapor saturation existing in most laboratory experiments. The transition temperature of ice accretion measured in natural fog on board an aircraft agreed better with the results of the Mt. Washington study.

This paper describes the results of snow drift measurements made on a strong katabatic wind slope at Mizuho Station (70°42.6’S, 44°18.9’E; 2230 m above mean sea level) in East Antarctica. From the vertical profile of the mass flux of blowing snow up to 28 m above the snow surface under conditions of snow fall, the snow fall densitities have been estimated as asymptotes of the profile (Fohn 1980). Snow fall densities as asymptotes were estimated between 8x10-6 and 8xl0-5 kg/m3. Assuming a fall velocity of blowing snow particles as 0.5 m/s, above values correspond to values of the vertical flux of snow fall between 2x10-3 and 1.5xl0-1 mm/h.

To study crystal shapes, and formation and growth mechanisms of snow crystals formed below -20 °C, a new type of diffusion chamber was constructed. Using this chamber, different kinds of “peculiar shaped” crystals previously observed in nature have been produced, together with normal types of snow crystals. Gohei twins, one of the most typical polycrystalline shapes in nature, have been produced artificially. The vapor pressure was at or near water saturation at the time of nucleation. Analysis of photomicrographs and replicas of Gohei twins that were replicated in the polar regions show that the number frequency of the tip angle has a maximum frequency at about 77° and a minor one at about 54°.

On the basis of these results, a formation mechanism for some Gohei twins is proposed in this paper.

Observations were made on thermal modification due to cooling when air flows from a grass-covered area onto a melting snow surface. To clarify the relation between such modification of air and snowmelt, the downwind variation of temperature, humidity and wind speed, together with net radiation and spatial variation of snowmelt were observed at a small snow patch measuring 70 m long and 30 m wide. When air temperature was between 10 and 20°C, with nearly neutral stratification over the upwind grass-covered area, air temperature at 0.1 m level decreased by 4 to 9°C over the downwind distance of 48 m above the snow patch, and so the sensible heat flux decreased in the downwind direction. As a result of such cooling, snowmelt in the central part of the snow patch was found to be about 25% smaller than near its edge. This value can be explained by the decrease of the total heat flux toward the snow surface due to the air modification.

Effects of drifting snow are examined from measurements of radiation fluxes at Mizuho Station in the katabatic wind zone, Antarctica. A good correlation is found between the difference of downward longwave fluxes measured at two heights and wind speed used as an index of drifting snow. The wind increases the downward flux at a rate of 2 W m-2/m s-2 when wind speed is higher than 13 m/s. Drifting snow suppresses the net longwave cooling at the surface. Direct solar radiation is depleted greatly by the drifting snow; however, the global flux decreases only slightly, compensated by the large increase of the diffuse flux, at a rate of about 1% for each 1 m/s increase in wind speed. At Mizuho Station, the effect on longwave radiation prevails throughout the year. The relation between snow drift content and wind speed is obtained from shortwave optical depth measurements as a function of wind speed. A simple parameterization of radiative properties is given.