A subjective review and summary of the key ideas presented at the conference is given, with occasional indications as to which scientific steps might resolve specific queries that arose from the work. The intention is to encourage closer reading of the papers.

Observations of ice drift received from an array of ARGOS buoys drifting in the Weddell Sea in winter 1986 are described. Wind and current data are also available, permitting derivation of the complete momentum budget including the internal ice stress computed as residuum. It is shown that the variability of forcing both of the atmosphere and of the ocean is large, and that internal ice stress is not negligible; monthly vector averages amount to about half of the wind and water stresses. Coefficients of shear and bulk viscosity are derived according to Hibler's model of ice rheology; they turn out to be negative occasionally, in particular when small-scale forcing of the atmosphere is large.

A generalized numerical model which allows for a variety of non-linear rheologies is developed for the seasonal simulation of sea-ice circulation and thickness. The model is used to investigate the effects (such as the role of shear stress and the existence of a flow rule) of different rheologies on the ice-drift pattern and build-up in the Arctic Basin. Differences in local drift seem to be closely related to the amount of allowable shear stress. Similarities are found between the elliptical and square cases and between the Mohr-Coulomb and cavitating fluid cases. Comparisons between observed and simulated buoy drift are made for several buoy tracks in the Arctic Basin. Correlation coefficients to the observed buoy drift range from 0.83 for the cavitating fluid to 0.86 for the square rheology. The average ratio of buoy-drift distance to average model-drift distance for several buoys is 1.15 (square), 1.18 (elliptical), 1.30 (Mohr-Coulomb) and 1.40 (cavitating fluid).

The granular nature of the Arctic pack ice, and the plastic nature of the deformation of the pack due to ridging, has long been recognised. However, because of a lack of experimental data, assumptions must be made to define the shape of the yield curve and the associated flow rule which characterize the aggregate rheology of the ice pack. In this work, the results of numerical experiments with a simulated granular viscous-plastic material are presented. The experiments model the granular texture of the ice pack as a dense assembly of non-uniform diameter disks in a rectangular control area. Deformation of the control area is driven by constant strain rates. In the experiments, the ratio of the principal strain rates is varied from isotropic convergence through uniaxial divergence. The stresses computed in the experiments, at the various strain-rate ratios, define a yield curve in principal-stress space. The effects of different coefficients of friction and viscous damping on the shape of the yield curve are explored.

A two-level, viscous-plastic, sea-ice model is modified to allow ridged and unridged ice to be treated separately. This is accomplished by introducing a third continuity equation with additional terms to transform level ice into ridged ice during deformation. The standard model is run for three years using observed forcing from 1981–83, along with three sensitivity runs, to investigate the role of strength parameterization and energetically consistent deformation on ridge production. As expected, most of the ridging occurs near the coast, with the highest ridged-ice fraction produced off the Canadian archipelago and northern Greenland coast. The sensitivity studies indicate that, although the ridged-ice fraction is very sensitive to the strength parameterization, the lead fraction and total ice volume are much less affected. Requiring the deformation to be energetically consistent with the assumed yield curve almost doubles the amount of ridged ice produced.

A 40 km-resolution ice—ocean model of the Bering Sea is used to investigate the effects of ocean circulation and vertical convection on the seasonal and interannual ice extent variations in the Bering Sea. The model is driven with daily time-varying atmospheric forcing from 1981–83. A series of sensitivity studies is carried out to examine the effects of the vertical diffusion and precipitation on the ice margin and the effect of stratification on the ocean circulation. For comparison, an ice-only simulation, with a motionless oceanic boundary layer of fixed depth, is also carried out. In the Aleutian Basin, the ice-ocean model exhibits a cyclonic ocean circulation which consists mainly of a baroclinic current component. On the eastern Bering Sea shelf the flow is mainly barotropic, with a northwestern shelf flow along the Alaskan coast and a return southeastern flow along the shelf break. The seasonal and interannual variability of the ice margin is significantly better simulated by the ice-ocean model than by the ice-only model, especially when an enhanced vertical diffusion is used. However, the seasonal cycle of ice extent exhibits too little ice in the southeastern Bering Sea and excessive ice in the northwest. The advance and retreat of the ice edge also tends to lag behind the observed results by a few weeks. The inclusion of precipitation improves the ice extent in the southeast. The results suggest that an enhanced vertical resolution, together with a more complete boundary layer formulation, will be required to achieve realistic seasonal simulations of the Bering Sea ice–ocean system.

A prognostic ice-ocean model of the Arctic, Greenland and Norwegian seas with daily wind and atmospheric forcing is integrated for 30 years to quasi-equilibrium. Three simulations are carried out to investigate the role played by ice deformation and transport in baroclinic adjustment of the Arctic Ocean: a standard run with precipitation and ice transport, a simulation without precipitation and a “thermodynamics only” simulation without ice transport but including precipitation. A diagnostic model is integrated for five years to serve as a comparative control run. Comparison of the vertically integrated stream-function of each of the model runs indicates that the vertical density stratification needed to maintain the circulation of the Arctic Ocean is reduced excessively when precipitation is neglected and artificially enhanced if ice transport out of the basin is ignored. This effect is even more noticeable in the surface currents and is also apparent in a comparison of simulated and observed drifting-buoy tracks. An analysis of the salt budget of the Arctic Ocean indicates that the three main components, salt transport by the ocean, salt flux from the annual cycle of ice, and a fresh-water flux from precipitation and river runoff are approximately of the same magnitude. The main circulation deficiency identified in the simulations is an inadequate flow of Atlantic water into the Arctic Basin through the Fram Strait.

The circulations of the Arctic ice cover and ocean are investigated using a coupled ice-ocean model. The coupling is strong and two-way for synoptic time scales, but is limited on seasonal time scales: the geostrophic ocean currents are not changed by the computed heat and salt fluxes. The ice-drift motion, Ekman transports and the wind-driven part of the barotropic circulation are examined for the months of February and August 1986, representing different atmospheric forcing, ice-thickness and ice-strength regimes. Initial examination of the results revealed no significant seasonal dependence of ice-drift response on the synoptic time scale, other than larger velocities with larger wind stresses. Daily maximum ice-drift velocities range from 20-40 cm s−1 in February, and 15-30 cm s−1 in August. The corresponding mean monthly maximum drifts were 11 and 9 cm, respectively. The drag associated with the geostrophic currents plays a much bigger role in the summer because of the lighter atmospheric stresses. The well-known reversal of the normally clockwise Beaufort Gyre to a cyclonic system in August takes place in a few days and lasts well into September. In February, the Beaufort Gyre varies between a large, clockwise system covering all the Canadian Basin to a small, tight gyre centered over the southern Beaufort Sea, without any hint of reversal or disappearance. Large areas of strong divergence were found in the Ekman transport patterns, as well as the ice-divergence fields, indicating areas where ice thinning, openings and new ice formation might occur. In August this occurred in the Chukchi Sea, and in February just north of Novaya Zemlya.

The mechanical energy balance of sea ice provides information about ice dynamic behavior, driving forces and the constitutive law. The energy balance equation, formed as the product of ice velocity with the ice momentum balance equation, describes changes to the kinetic and potential energy densities as power is input to the ice by wind and current. The momentum balance equation may also be used to describe the ice-stress divergence, air stress, and water stress, but the scalar form of the energy balance is simpler to understand. This paper provides new interpretations of several terms in the energy balance equation, in particular power input by air and water stress and by sea-surface tilt. Barometric pressure fields and drifting buoys deployed on the Chukchi Sea ice cover during 1982 provide wind, ice motion and current measurements that allow each term in the energy balance equation to be evaluated as a function of time. Magnitudes of power input by wind and current show how the energy balance is decomposed and help describe the relative importance of these driving forces. In the nearshore Chukchi Sea during February, March and April 1982, both wind and current provided significant forcing of the ice. Ice stress was also important and, at times, dominated other terms in the mechanical energy balance.

The role of internal forces in determining the geometrical properties of local features such as leads and ridges in sea ice has been investigated. A description of the horizontal stress distribution in a granular material subject to two-dimensional deformations has been derived. The description yields expressions for the propagation of deformation characteristics in a granular deformation field with constant angle of internal friction. The characteristics are mathematical singularities interpreted as trajectories of constant state variables, e.g. ice concentration or ice thickness typified as leads, fractures, pressure and shear ridges. The derived descriptions are used to study the geometrical nature of curving and rectilinear lead patterns frequently observed in sea-ice deformation fields. An expression for the relationship between the arching (tangent angling) of characteristics and relative increase in shear (maximum shear stress) along them is derived. It is explained how the acute angle between large-scale rectilinear characteristics is related to the important mechanical property of the deformed material, the angle of internal friction. It is outlined how the derived results can be applied to study the horizontal distribution of internal forces in sea ice by using imagery data and stress measurements at one site only.

A two-dimensional particle simulation model of the sea-ice ridging process is developed. In this model, ridges are formed from an intact layer of newly frozen lead ice colliding with a thick multi-year floe. Blocks broken from the leading edge of the lead ice collect above and beneath the multi-year floe to form the characteristic ridge structure seen in the central Arctic. The total energy consumed in ridging ice, which is converted into the potential energy of the ridge structure and dissipated by the frictional and inelastic contacts between blocks of ice, is calculated explicitly. The results of preliminary numerical experiments using this model indicate that the amount of energy required to ridge ice may be much larger than previous estimates.

In this study, the collision of ice floes under the action of a monotonic wave is quantified. The lateral motion of an ice floe caused by the wave is modeled as the sliding of an object under gravity. In this case, the gravity component in the direction of motion varies with time and space as the wave progresses by the floe. Drag and added mass effects are included in the model. Two floes located at different positions are shown to have a net difference in their drift (caused only be repeated wave passages). In most cases, this differential drift eventually causes floe collision. When two floes collide, a spring and dash-pot model is adopted to calculate the contact force. A one-dimensional wave passing through a one-dimensional array of disc-shaped floes is examined. Two phenomena are apparent from the analysis. First, waves have a herding effect that forms bands of floes with the width equal to the wavelength. Secondly, the frequency of collision is sensitive to the elastic properties of the floes and the wave amplitude. With sufficient values of the damping constant, which operates when two floes collide, the floes stay in contact for prolonged periods, indicating the potential to freeze together and form composite floes, as was observed in the field studies.

The radiation stress at an ice edge due to incident sea and swell is reconsidered in the light of new theoretical results concerned with the calculation of the reflection and transmission coefficients at the edge of a thin elastic sheet. Both monochromatic seas and seas made up of a spectrum of periods are discussed, the latter invoking the Pierson-Moskowitz spectral model. It is found that the force per unit length due to radiation stress is comparable in magnitude with other driving mechanisms at the edge, i.e. with winds and currents.

The possibility of long-period ocean waves coupling to an ice shelf is investigated. A thick elastic plate model is used for the ice shelf with comparisons made to the simpler thin-plate model. The strain set up on the ice shelf by a normally incident single frequency ocean wave is calculated by completely solving the equations governing the velocity potential for such a system. In the absence of measurements on an ice shelf, existing measurements of long-period strain on an ice tongue are used to estimate the required incident amplitude in the open water to induce the observed oscillations. It is found that the height of seas required indicates that ocean wave driving is a plausible forcing mechanism for observed oscillations.

In this paper, the transient response of an ice sheet subjected to sudden uplift is investigated. The ice sheet is modeled as an infinite thin plate undergoing small deflections that is being pushed vertically upwards by a rigid flatended cylinder. The water underneath the ice sheet is modeled using incompressible potential flow theory. For a given motion of the cylinder, the current problem is reduced to a Volterra integral equation of the first kind for the indentation force. It is solved by transforming the governing equation into an integro-differential equation. For the physical quantities of interest, comparisons are made with related experiments.

An analysis of the relationship between iceberg calving rates and water depth has been completed for 22 tidewater glaciers. A linear relationship provides reasonable accuracy, with a correlation coefficient of 0.85, for all tidewater glaciers examined, whether they be polar or temperate. The polar glaciers have a slightly lower calving rate for a given water depth. This relationship indicates a lower calving rate for water depths over 50 m than determined by Brown and others (1982). It is based only on glaciers or ice streams and cannot be applied to ice shelves.

A temperate glacier ending in water advances by depositing a moraine shoal, which “dams” tidewater calving, and then recycling this shoal in conveyor-belt fashion. A simple model suggests that conveyor-belt recycling of the shoal must continue until calving stops and all ice flux is removed by surface ablation, or until the glacier retreats rapidly from the shoal; retreat can occur without external forcing if the glacier advances into deepening water and if sediment recycling is rapid compared to supply of new sediment to the moraine shoal.

Two sea-ice layers, one measured as 9 m thick, the other at least 12 m thick and estimated to be 24.5 m thick, have been located by ice core drilling in the west Ward Hunt Ice Shelf. To examine the preservation of physical-structural characteristics over long time intervals, the crystal structure and brine volumes in the sea ice, which possibly dates back to about 3000 BP, have been studied. The structural characteristics are immediately recognizable as those of undeformed congelation sea ice accreted by Stefan growth. Brine volumes in the ancient sea ice are higher than those in modern multi-year ice at the same temperature. The preservation of brine over a time span of hundreds to thousands of years is attributed to an absence of surface meltwater to effect brine flushing and the very slow, even negligible action of gravity drainage, brine pocket migration and brine expulsion. The congelation structures indicate that sea ice can grow by the Stefan accretion mechanism to thicknesses exceeding the equilibrium thickness (2.5–5 m) of most undeformed multi-year ice. The observed physical-structural characteristics of the Ward Hunt sea ice strongly suggest that many of the properties attained by sea ice are permanent and not affected by slow-acting physical processes.

Analysis of AVHRR data collected during the summer and winter over the Ross Ice Shelf reveals complex patterns of curvilinear stripes. In particular, a large, looping pattern of stripes is observed west of Crary Ice Rise in an area where conventional glaciological data collected with surface and airborne methods have been interpreted to suggest uncomplicated flow. On the basis of previous work using radar data to study ice flow downstream of Crary Ice Rise, we conclude that the stripes represent relict flowlines. The mechanism that produces these stripes is unclear, but we hypothesize that they are associated with subtle topography. Based solely on the patterns of stripes and their location in the outflow of major ice streams, we propose that they are related to an ice raft torn from the grounded ice sheet about 400 km upstream from its present position.

Generation mechanisms of mesoscale eddies in the ice-covered ocean are studied by using numerical ice–ocean models and discussed with reference to previous papers. The three possible mechanisms of eddy generation, with sea ice as a passive tracer, are current instability, current-eddy interaction and current–bottom topography interaction. The current instability, categorized into barotropic and baroclinic instabilities, may be responsible for eddies near the ice edge associated with a strong current. An eddy can interact with a current, producing additional eddies, where stability of the current is again an important factor for eddy formation. Eddies over bottom topography on the continental shelf are explained by current–topography interaction; i.e. anticyclones are produced over banks. The particular mechanism that includes ice as an active material is an ice–ocean interaction; i.e. a wind stress is larger over the ice than on open water and induces Ekman pumping and suction, which produce dipole eddy motions in the non-uniformly ice-covered ocean. The eddies are suggested to be important for cross-shelf exchanges of ice and heat as well as determining locations of deep convection.