An implicit, analytical model for momentum, heat and salt flux within a sea-ice/upper-ocean system is developed. The model comprises three parts: (a) an equation for turbulent stress in the oceanic boundary layer and upper pycnocline, from which turbulent scalar fluxes are derived; (b) a model for heat and mass transfer in a thin sub-layer near the ice/ocean interface; and (c) a model for momentum flux lost to the internal wave-field if the ice under-surface has large-scale (pressure-ridge keel) relief. Features of the model are demonstrated by simulating response of the ice drift and upper-ocean temperature and salinity structure to constant heating and variable surface stress.

Seasonal simulations with large-scale coupled ice–ocean models have reproduced many features of the ice and ocean circulation of the Arctic Ocean and the Greenland and Norwegian seas (e.g. Hibler and Bryan, 1987; Semtner, 1987). However, the crude resolution and high lateral eddy viscosity used by these models prevent the simulation of many of the smaller-scale seasonal features and tend to produce sluggish circulation. Similarly, the use of a single year’s atmospheric forcing prevents the simulation of features on an interannual time-scale. As an initial step towards addressing these issues, an 80 km diagnostic Arctic ice–ocean model is constructed and integrated over a three-year period using daily atmospheric forcing to drive the model. To examine the effect of topographic resolution and eddy viscosity on model results, similar simulations were performed with a 160 km-resolution model. The results of these simulations are compared with one another, with buoy drift in the Arctic Basin, and with observed ice-edge variations. The model results proved most sensitive to changes in horizontal resolution. The 80 km results provided a more realistic and robust circulation in most areas of the Arctic and improved the modelled ice edge in the Barents Sea, while also successfully simulating the interannual variation in the region. Although it performed better than the 160 km model, the 80 km model still produced too large an ice extent in the Greenland Sea. No significant improvement in the ice-edge prediction was observed by varying the lateral eddy viscosity. The results indicate that problems remain in the vertical resolution in shallow regions, in treating penetrative convection, and in the simulation of inflow into the Arctic Basin through the Fram Strait.

For the winters of 1985 to 1989, ice-edge melt rates off Labrador and eastern Newfoundland were calculated from ice charts and satellite-tracked ice beacon data. The ice charts provided ice-edge motion and ice thicknesses, while the beacons provided ice-drift rates and rates of ice-edge retreat. Over the continental slope, ice-edge melt rates, ice-drift rates and heat fluxes required to melt the ice were all higher than those over the continental shelf. Ice-edge melt rates had a mean value of 18.0 km d−1 over the slope but reduced to 4.8 km d−1 over the shelf. Similarly, the mean ice-drift rate was higher over the slope (41.5 km d−1) than over the shelf (12.7km d−1). Ice-edge melt rates increased linearly, as ice-drift rates perpendicular to the edge increased. Estimates of heat fluxes (±30%), required to melt the ice at the ice edge, reached values of 700 W m−2 for 3-day periods and had a mean of 340 W m−2 over the slope and 130 W m−2 over the shelf.

Oceanographic, meteorological and sea-ice data were obtained from the northern Bering Sea and Chukchi Sea during the autumns of 1987 and 1988. Ice-edge location was observed from ships and via AVHRR satellite data, and ice-drift information was obtained from ARGOS-tracked drift buoys. Meteorological data were obtained from ships, from an ARGOS-tracked meteorological station and from synoptic charts. The ice edge was significantly farther south in 1988 than during other years and impacted the Alaskan coastline. In 1987, the ice edge was, conversely, anomalously far north. Ice melt-back in certain regions, such as along the Alaskan coast and in Herald Canyon, was due to input from warm ocean currents. The larger-scale interannual differences in ice extent were, however, due to interannual differences in the regional winds. In particular, the anomalous and extreme southward extent of the ice edge during 1988 was due to northerly to northwesterly winds, which held the summer pack ice against the the beach. Meltwater from this ice salt-stratified the upper water column, so that the ice eventually became effectively insulated against vertical flux of heat from the underlying warm water in the coastal current.

This paper discusses in detail the energy flux density estimation at the ice/atmosphere interface over the entire North Water region, based on the ice-type distribution during the winter (1978/79 to 1985/86). The analysis is based on passive microwave measurements of the scanning multichannel microwave radiometer (SMMR) onboard the Nimbus 7 satellite. Based on the relation of increasing polarization ratio (18 or 37 GHz) with decreasing ice thickness, a threshold technique was applied to classify the different young ice types. The polarization ratio method seems feasible for climatological applications, such as energy flux density estimations, over large areas with homogeneous young ice. More than 20% young ice (<0.3m) was found in the North Water area (100,000 km2) throughout most winters, with maximal values of 60% and more, and only a few percent of open water were classified. In general, the young ice cover in the North Water was decreasing towards the end of the winter. The mean energy loss for the entire North Water region was found to be 77 W m−2 for the months of November to March. Considering an energy supply by refreezing of 38 W m−2, the remaining 39 W m−2 must be withdrawn from the enthalpy in the sea water.

Several recent studies have suggested that much of the winter-time Antarctic ice is thin (<0.3m). The presence of extensive areas of thin ice has a significant effect on ocean-atmosphere energy exchange. This is investigated using the Maykut (1978) thin-ice energy-budget model in a study for typical September Antarctic ice and climatic conditions. To study the sensitivity of turbulent heat loss to ice concentrations, the Maykut model is combined with an empirical parameterization for the turbulent fluxes from leads (Andreas, 1980) which takes account of the non-linear relationship between heat loss and lead width (or ice concentration). In this one-dimensional sensitivity study, a constant floe-size is assumed, and ice-concentration variations are simulated by changing the width of the leads between floes. The modelled results, for the floe size considered, indicate that at 80% ice concentration the turbulent heat loss through the thin ice component can be greater than that from leads. As concentration decreases, however, the fractional loss through the ice, and, hence, the ice-thickness distribution, becomes less significant. For the concentrations lower than 50%, there is little change in turbulent loss with further decrease in ice cover, as the atmosphere effectively “sees” open ocean.

A coupled sea-ice-oceanic mixed-layer model for the Southern Ocean is forced with daily atmospheric variables from the global analyses of the European Center for Medium Range Weather Forecasts (ECMWF). Using the analyses of the lowest level in the calculations of the turbulent heat fluxes and stresses with the appropriate bulk formulae, the simulation results resemble earlier ones with climatological forcing. In order to avoid a predetermination of the simulated sea-ice conditions from the (climatological) specification of the surface boundary conditions in the atmospheric general circulation model (AGCM), the sea-ice model is coupled additionally to a one-dimensional atmospheric boundary layer model. Using the global ECMWF-analyses as before, the coupled model is now forced from the geostrophic level (850 hPa). Without any changes of the original model parameters and physics, the results are rather poor in that the ice extent as well as the ice velocities are generally too low and that the ice thickness distribution resembles the results of a pure thermodynamic sea-ice model. The results with the forcing from the higher level are more realistic when snow and mixed-layer effects are neglected, thus resembling those of Koch (1988) in the Weddell Sea. This indicates that the parameterizations in the atmospheric boundary layer model have to be readjusted in order to interact realistically with the snow-sea-ice-oceanic mixed-layer model. Additionally, it will be demonstrated that the pattern of the wind field, whether from the geostrophic or the surface level, has a significant influence on the sea-ice model results.

An observed ocean-drift data set is used as the basis of a wind-driven coupled ocean-sea-ice-atmosphere model including interaction and feedback. The observed characteristics of the Antarctic sea ice are described including the ice thickness, ice concentration and horizontal advection. The atmospheric model computes heat fiuxes, sea-ice growth, changes in concentration and advection. Sensitivity studies show reasonable and stable simulations of the observed sea-ice characteristics for the present mean Antarctic winter climate. The response times and feedbacks of the ice-atmosphere system as represented by the model appear to allow scope for the development of some persistence of anomalies.

The texture of a material is closely connected to both its development and its properties. Aided by fully automated image-processing techniques developed for a simple PC-based system, we extract textural information from sea-ice thin sections. As a quantifier of mean size and orientation of grains, the “contrast gradient” has been defined. The skewness of the chord-size distribution aptly describes the grain-size distribution. Through a combination of these two parameters, granular and columnar textures can be distinguished. Samples of mixed columnar/granular texture may be further specified through their fractal properties. Fully automated techniques may be of use due to their objective scrutiny and their ability to draw information from large databases. In addition, they allow for highly specific descriptions of texture and may further the understanding of its development.

We report on studies of sea-ice texture conducted during a number of expeditions into the Weddell Sea. Sea ice in the Antarctic is dominated by granular ice of frazil origin in floes of all ages, in contrast to ice in the Arctic, which consists predominantly of columnar ice of congelation origin. The large fraction of granular ice in first-year sea ice is a result of the dominant ice-formation process in the advancing ice edge, the pancake cycle. The dominance of granular over columnar ice in second- and/or multi-year ice is a result of the large degree of deformational activity in the Southern Ocean.

Granular, fresh-water ice was deformed at −40°C at 10−3 s−1 under confinement. Loading was applied such that confining stress, σ2 and σ3, were equal to each other and proportional to the highest stress, σ1; i.e. σ2 = σ3 = Rσ1. Two regimes were evident. At lower confinement, the fracture stress, σ1f, increases markedly as R increases. At higher confinement, σ1f increases less sharply. The transition occurs at about R = 0.15. Wing cracks were seen in the ice fractured within the low-R regime. The data are analyzed in terms of the frictional crack sliding-wing crack mechanism of brittle compressive fracture.

Uniaxial compression tests were carried out on fresh-water, isotropic, granular ice at a strain rate of 6 × 10−4 s−1. We investigated the effect of temperature (between −2 and −39°C) and grain-size (1 mm–8 mm) on the critical stress and strain at the initial crack nucleation. The amount of non-elastic strain at this event was estimated. The critical stress for initial crack nucleation increased strongly with decreasing temperature, following an Arrhenius law. It also exhibited a linear increase with dg−1/2 (where dg is the average grain-size). It is shown that the results cannot be explained by the purely brittle model of Sunder and Wu (1990). The results are interpreted in terms of grain-boundary sliding, controlled by the intrinsic viscosity of the boundary.

Dynamic ocean-induced loading of ice has recently prompted interest in the fatigue strength of ice and, consequently, aspects of the fracture testing of ice need to be examined. One aspect in fracture testing, the effects of notch acuity on the apparent fracture toughness (KQ) of saline ice, was investigated in this study using four groups of single-edge notched-bend specimens with different crack-(or notch-) root radii, fabricated by four different methods; namely, drilling a small hole of two different radii at the crack tip, cutting a crack with a band saw and using a hand-held razor blade to scribe a sharp crack. For this study, tests were performed on saline ice using one crack length and orientation, constant thickness and constant temperature (−25°C). The mean values and the standard deviations of the apparent fracture toughness of the specimens with blunt notches were higher than those with sharp cracks. The results presented in a plot of fracture toughness versus where ρ is the crack-tip radius, provide a preliminary estimate of the required notch acuity for fracture-toughness testing of saline ice.

Very little data exist on the fatigue behavior of ice, yet fatigue may have an important role to play in a number of ice failure situations. Whilst reversed-bending tests and cyclic-compression tests have been performed on ice, no attempts have been made previously to measure fatigue-crack growth rates in ice. This paper presents the results of fatigue-crack growth experiments on S2 (i.e. columnar grained with the c axes randomly aligned within the horizontal plane) fresh-water ice. There is clear evidence that crack growth has occurred under cyclic loading but, as yet, the data are insufficient to determine the validity of da/dΝ versus δK type behavior. Preliminary results are presented and the implications of the results are discussed.

An experimental investigation was performed on the kinetic friction coefficient of laboratory-grown, columnar saline ice sliding against itself. Tests were performed on a dual-opposing load apparatus specially manufactured for attachment to an MTS testing system. The mean kinetic friction coefficient, μ, was measured for sliding velocities from 10−6 to 5 × 10−2 m s−1 at temperatures from —3° to —40°C under a contact pressure of about 20 kPa. The ice specimens were oriented with grain columns perpendicular to the sliding interface. At -3°C and at —10°C, three distinct regions were observed: from 10−6 to about 10−5ms−1, μwas nearly constant at 0.5; at velocities from 10−5 to 10−3 m s−1, μ began to drop rapidly to about 0.1; and, above 10−3 m s−1, μ began to level off at ~0.05. The velocity at which μ began to decline increased with decreasing temperature. At temperatures below —10°C, μ increased from ~0.5 at v =10−6ms−1 to a peak value of ~0.7 near a velocity of 5 × 10−5ms−1 and then fell rapidly to about 0.1 at 10−2ms−1. In general, μ increased with decreasing temperature and sliding velocity.

The data from a small-scale experimental study on ice-structure interaction are used to compute the energy exchanges that take place during creep deformation and intermittent and continuous crushing of ice. The energy supplied by the carriage is partly stored in the structural spring, partly converted to kinetic energy, partly dissipated in deforming and extruding the ice and partly dissipated as heat in the damping mechanisms of the structure. Except for the heat dissipation, all other forms of energy were computed from the experimental data, and the heat dissipation was computed from the energy balance using the first law of thermodynamics. Plots of all forms of energy are shown in graphical form, in which their relative magnitudes, times of occurrence and interplay can be seen. The main result of this study is the thesis that intermittent crushing or ice-induced vibration takes place whenever there is an imbalance between the rates of work done by the carriage and the indentor and that there are no vibrations when these rates of work are equal.

Indentation experiments have been performed on fresh-water ice at -9°C using an apparatus which permitted visual observation of the contact zone at the ice/indentor interface by viewing through the specimens. Analysis of the video records and test data indicated that at least 50% of the energy dissipated during the process of indentation was consumed by pressure melting and heat generation through the rapid viscous flow of the liquid. The thickness of the liquid layer was estimated to be 16/μm and the contact pressure in the melting zone was at least 90 MPa. Video records of small-scale impact tests on ice indicated that the same process of pressure melting and extrusion was taking place.

M.V. Nella Dan was beset in ice near 66°S, 51°Ε from 27 October to 15 December 1985. The scientific investigations planned for the voyage into the sea-ice zone included ice thickness, concentration and extent measurements, aerial photography of the ice, a core drilling project and meteorological observations. Once beset, the programme was expanded to include repeat measurements of a small strain grid and measurements of the sea-ice drift rate. Strongly convergent local conditions had led to the initial besetment of the ship. Analysis of measurements of the strain grid area over an 11 d period shows that, although there is some indication that the ice field may have been divergent, opening by approximately 3.4% over this period, there are large errors in the measurements and some doubt must be placed on the reliability of this estimate. The drift speed and direction were found to be highly dependent on wind speed and direction, the drift rate being approximately 2.7% of the wind speed at an angle of about 29° to the left of the wind direction.

Local ice strains and in situ ice stresses were simultaneously measured on the Coordinated Eastern Arctic Experiment (CEAREX). The experiment took place in the fall of 1988 and was centered about an ice-strengthened ship moored to a multi-year floe in the pack ice northeast of Spitsbergen. During the period of data collection, which extended from early October to late November, the ship and the ice surrounding it drifted from 82°40′N, 32°32′E to 78°54′N, 31°27′E.

As soon as ice temperatures were low enough to permit installation, stress sensors were placed at four sites, two sites on each of two adjacent multi-year floes. Principal stress components and the principal stress direction were determined at each sensor. At the same time, microwave transponders, capable of measuring ice deformation to accuracies better than 1 m, were positioned within 1 km of the stress sensors and provided an approximation of the local strain field.

What makes this joint dataset particularly interesting is that it includes some large ridging events and a particularly large event which terminated the experiment when the multi-year floes in the local area were broken into small fragments. A wide range of ice stresses was measured during the period. The largest compressive stresses, about 250 kPa, were measured by the near-surface sensors. Although sensors in different locations responded differently to ice movement, the large events were common to all shallow sensors.

Ice-surface geometry at the shear edge of landfast ice has been examined. An aerial survey was conducted at the intersection of Lancaster Sound and Navy Board Inlet during March 1984. Surface profiles were then obtained in three representative areas and used to evaluate height statistics. A region, 150 m wide, at the ice edge was distinct from the rest of the ice cover. In the region at the shear edge, the surface height followed lognormal distributions. The average and maximum heights were 2.3 m and 5.6 m, respectively. In the region inside the shear edge, the height followed negative exponential distributions, with average and maximum values of 1.5 m and 4.2 m, respectively.