IP16b

Seawater Flooding, Snow Ice Formation and the Mass Balance of First–Year Ice Floes in the Ross, Amundsen and Bellingshausen Seas:
I. Observations.

Martin Jeffries, Barbara Hurst–Cushing, Ted Maksym
and Ricardo JaÒa

Geophysical Institute, University of Alaska Fairbanks, USA Instituto Antartico Chileno, Santiago, CHILE

A knowledge of the thickness distribution and mass balance of the Antarctic sea ice cover is essential for understanding the role of the pack ice in modifying atmosphere–ocean interactions and exchanges of heat, mass and momentum, and their influence on ocean and climate variability and the biological productivity of the ice and ocean. Fundamental to the understanding of the ice thickness distribution is a knowledge of the dynamic (ridging and rafting) and thermodynamic (freezing and melting) processes that contribute to the development of the sea ice cover under the influence of atmospheric and oceanic forcing. One of the key thermodynamic processes that sets Antarctic sea ice apart from most Arctic sea ice is the formation of snow ice after seawater has flooded the snow/ice interface.

In the austral winters of 1993, 1994 and 1995 we completed four cruises aboard the R. V. Nathaniel B. Palmer in the first–year pack ice of the Pacific sector of the Southern Ocean. One of the objectives of each cruise was to improve our understanding of the role of the snow cover in the mass balance of ice floes, i.e., to determine the extent of seawater flooding at the snow/ice interface, to identify the amount of snow ice that had formed, and to calculate the fractional contribution of snow to the total ice mass. This objective was achieved as follows: (1) snow depth, freeboard and ice draft were measured along 100m transects on floes, (2) ice cores were obtained for analysis of stable isotopic composition and crystal texture, and identification of ice types and formation processes; and, (3) a simple model was used to determine the snow fraction of the snow ice layers in each core (f_s) and of the entire length of each core (F_m). Currently, complete data sets are available for cruise #1 (August/September 1993, Bellingshausen & Amundsen seas), cruise #2 (September/October 1994, Amundsen & Ross seas) and cruise #3 (May/June 1995, Ross Sea).

Seawater flooding of the snow/ice interface, i.e., negative freeboard values, occurred at 18%, 51% and 29% of the drill holes during cruises 1, 2 and 3 respectively. Snow ice occurred in the majority of cores, and the amount of snow ice, as a function of the total length of core examined, was 24%, 37% and 25% during cruises 1, 2 and 3 respectively. The extent of seawater flooding and the amounts of snow ice are greater than most values reported elsewhere in the Antarctic pack ice, and indicate that flooding and snow ice formation play an important role in the thermodynamic thickening of first–year floes in the Pacific sector of the Southern Ocean. The significance of snow ice formation for thermodynamic thickening is emphasised by the fact that individual snow ice layers are generally thicker than either frazil or congelation ice layers, which thicken primarily by dynamic processes, i.e., rafting and ridging.

The Ross Sea data from cruise #3 indicate that snow ice formation is important even in the early winter development of the ice cover. However, there are strong spatial variations in the amount of snow ice and its contribution to the thermodynamic thickening of the ice cover. In the inner pack ice on the continental shelf, snow ice contributes to 15% of the total ice mass compared to a 65% contribution from congelation ice. In the outer pack ice on the deep ocean, snow ice contributes to 38% of the total ice mass compared to a 22% contribution from congelation ice. In both regions, frazil ice comprises much of the remainder of the ice mass. The preponderance of congelation ice in the southern Ross embayment might be due to lower snow loads, plus a low oceanic heat flux and a colder, less stormy environment on the continental shelf, which favour the more frequent and prolonged calm conditions necessary for significant congelation ice growth.

The mean f_s values for cruises 1, 2 and 3 are 9%, 14% and 10% respectively. The mean F_m values for cruises 1, 2 and 3 are 2.5%, 5.5% and 3.6% respectively. Differences in F_m values between cruises reflect a combination of differences in f_s values, snow ice layer thickness and ice core length. Regardless of these differences, the f_s and F_m values are similar to those reported in the Weddell Sea. Although the contribution of snow to the snow ice layers and the total ice mass is quite modest, these values are significant in view of the large amount of snow ice that contributes to the thermodynamic thickening of the ice cover.

IP16c

All-weather satellite monitoring of ice-covered ocean in Russian Arctic

Yu. A. Kravtsov, E.B. Kudashev Yu. G. Trokhimovski
and P. Myasnaikov

Space Research Institute, Russian Academy of Science Profsoyuznaya 84/32 Moscow, RUSSIA
Institute of Automatization and Processes Control Russian.Acad.Sci., Far-East Branch Radio 5 Vladivostok, RUSSIA

The Study of Arctic is a difficult and costly task. All weather satellite monitoring is the only reasonable way of obtaining reliable information on ice motions, ice cover and ocean pollution, and technogenic stress on the Arctic environment. There are several serious reasons to support development of satellite monitoring of the Russian Arctic areas:

the increasing role of ecological monitoring as an important factor in Russian economic development;

increasing industrial activity, connected with the development of Arctic natural resources:

remoteness and difficult accessibility of Arctic regions, severe climatic conditions, and the necessity to cover large areas:

increased attention of North European countries, as well as USA and Canada, to the ecological state of the Northern regions of Russia

the great influence of the Arctic areas on Global Climate Change

Information obtained presently from Russian satellites ("Meteor", "Resurs", "Okean") as well as from European (ERS-1,2) Canadian (Radarsaf) and American (NOAA series) satellites requires integrated analysis for ecological monitoring. The problem under discussion here is how to develop the methodology and to form an interface to an experimental airspace system for observation of the Arctic environmental state .

The system considered has to be able to:

Measure sea surface temperature by IR and microwave radiometers;

Reveal oceanic currents and fronts;

Compare visual, IR and radar images of the ice cover and show correlations between them, which may increase the information content of available data;

Reveal ocean surface pollution, oil and organic films (visual, IR and microwave ranges);

Detect pollution of the ice cover (optical spectral measurements);

Retrieve oil pollution and ecological change information on the land, especially in tundra areas (spectral measurements);

Measure the speed and direction of sea ice drift

Determine atmospheric processes which impact on ocean and marine ice (from observations of cloud cover);

Reveal seasonal changes in reflection characteristics of the oceanic and land ice;

Determine the ocean colour for estimation of ocean bioproductivity;

Reveal and monitor areas with anomalous characteristics;

Analyse surface manifestations of the ocean bottom relief ad other characteristics of shallow waters;

Register oceanic internal waves in the shelf zone, analyse the internal wave influence on mixing processes in the upper ocean, on the bioproductivity of the shelf areas and on pollution spread;

Generally analyse the ecological situation in Northern Seas with the purpose of predicting the situation in the most unfavourable regions;

Analyse the ecological situation in areas of radioactive and chemical waste storage;

Maximise use of radar data ("Okean", "Radarsat", ERS- 1,2");

Obtain information necessary for oil platforms;

Generally monitor the ice cover in Polar areas, including boundaries of "water-ice", "sea-ice-land’; detect leads and polynyas in ice;

Determine the ice concentration, ice age and thickness and estimate spatial-temporal distribution of ice forms.

To understand and to forecast the ecological situation in the Arctic one has to understand that the processes interact with each other, and that this interaction is of a complicated nonlinear nature. For that reason modern technologies of data processing and analysis are of high priority.

IP16d

Influence of broken ice on the propagation of surface waves

A.E. Bukatov, A.A. Bukatov and V.V. Zharkov

Marine Hydrophysical Institute of Ukrainian Academy of Sciences, UKRAINE

The influence of broken ice on planar propagating waves in an inviscous incompressible fluid filling a horizontally unbounded basin is considered.

An analysis of the influence of ice on the wave disturbances caused by propagation of surface waves of small amplitude from an infinitely deep area of the basin, over a bottom step into an area of finite depth, is made out on the basis of ‘wavemaker’ theory. An integral equation defining horizontal velocity of fluid particles over the step is obtained matching ‘wavemaker’ solutions for the considered basin areas. An unknown velocity function is expanded in terms of the complete orthogonal set of eigenfunctions for the area of finite depth. This allows us to reduce the integral equation to a system of algebraic ones. The system is solved numerically using the conjugate gradient method. As a result, the dependences of the amplitudes of transmitted and reflected waves on the incident wave period, ice thickness, and step submersion depth, are obtained. The width of the zone of disturbances caused by near-step evanescent modes is estimated.

In the case of a basin of constant finite depth, the evolution of surface waves of finite amplitude is considered. In this case, the non-linearity of vertical accelerations of ice floes is taken into account in the dynamic boundary condition at the basin’s surface. An asymptotic expansion, containing values up to third power of smallness for the fluid velocity potential and basin’s surface elevation, is obtained by using the multiscale method. Basing on this, an analysis of the dependence of elevation profile, phase shift and wave height on the ice thickness, length and amplitude of the initial basic wave harmonic is carried out. It is shown that the non-linearity influences the ice even at long waves, although in the case of the linear problem statement, the ice does not affect them. Under the conditions of non-linearity, the influence of the ice amplifies with time. In this case, it effects mostly the phase shift of the wave disturbance spatial distribution. The shift is a result of the damping effect of the ice. With short waves, the ice causes decreases not only the phase velocity but also the amplitude.

The evolution of the basin’s surface elevation profile is analyzed under conditions of non-linear interaction of traveling waves of the first and second harmonics. Amplitudes of initial harmonics have the same power of smallness. The dependence of the characteristics of the basin’s surface elevation on the amplitudes of interacting harmonics under ice conditions is obtained.

The contribution from the horizontal gradient of the first power approximation of the basin’s surface elevation into non-linear additions to the horizontal component of the wave disturbance velocity is estimated. It is shown that introduced corrections appear in the second power approximation via the dynamic boundary condition only. At the same time they appear in the third power approximation via both dynamic and kinematic conditions.

IP16E

Evaluation of thermodynamic parameterizations of sea ice for applications with general circulation models

T.E. Arbetter and J.A. Maslanik

Program in Atmospheric and Oceanic Sciences, University of Colorado, Boulder, Colorado, USA

Treatment of sea ice in GCMs has historically been lacking, and it is clear that improvements can be made. Contrary to this, however, is the desire to simplify sea ice parameterizations in order to maintain computational efficiency. As new and better thermodynamic parameterizations of physical processes of sea ice are developed, it is important to evaluate their relative impact on model simulations of sea ice. Here, the impact of thermodynamic parameterizations on simulations of sea ice in a two-dimensional dynamic-thermodynamic model is tested. These enhancements are then ranked based on their relative ease of inclusion into a GCM versus the magnitude of the enhancements' effects on ice simulations in the 2-D model. Sensitivity studies include more detailed treatments of such factors as spectral albedo, snowfall, approximations of ice thickness distributions, penetrating solar radiation, and brine pockets. Results of GCM runs incorporating a subset of these enhancements are summarized, and aspects of existing GCMs that affect the ease and effectiveness of including enhancements to sea ice processes are discussed. This discussion of the effects of including such sea ice enhancements relative to the cost required to implement them within a GCM should help guide future efforts to improve climate simulations.

IP16f

The Response of a Three Dimensional Coupled Ice-Ocean Model to Daily Varying Atmospheric Forcing from a Global Model

Ruth H. Preller and Pamela G. Posey

Naval Research Laboratory, Stennis Space Center, USA

A coupled ice-ocean model has been developed by the U.S. Navy for the purpose of sea ice forecasting in the Northern Hemisphere. The model consists of the Hibler Ice model coupled to the Bryan and Cox ocean model and has been called the Polar Ice Prediction System 2.0 (PIPS 2.0). The PIPS 2.0 extends from the North Pole south to approximately 30^o N latitude to include the ice covered seas of the western Pacific. The model uses a horizontal resolution of 0.28 degrees and 15 levels to define the vertical structure of the ocean. In a forecast mode, the coupled model is driven by the surface stresses and heat fluxes from the Navy Operational Global Atmospheric Prediction System, (NOGAPS). Daily fields from NOGAPS for the period 1992-1996 have been archived and used to drive the coupled model in a test mode. The atmospheric model fields as well as the resultant sea ice and ocean fields are examined to determine the interannual and seasonal variability in the atmospheric forcing and it's affect on the coupled model results. Results are presented as a time series for several different locations within the model domain over the 5 year simulation. Model results are compared to observations such as Arctic buoy data and ice concentration data derived from the Special Sensor Microwave Imager (SSM/I). These results show that a trend in NOGAPS, for initially warmer summers and consistently warmer winters, creates an increasingly thinner ice cover in the coupled model. Although some of these trends are observed in the ice concentration data, the coupled model results appear to exaggerate them. Reasons for these consistencies/ inconsistencies between model and data will be presented and discussed.

IP16g

Characteristics of Large-Scale Antarctic Sea-Ice Dynamics from Satellite Microwave Radar Data

Mark R. Drinkwater

Jet Propulsion Laboratory, California Institute of Technology, Pasadena,, USA

Ocean-atmosphere exchanges are exaggerated when the Antarctic sea-ice cover is parted and the ocean exposed to brisk winds. Relative sea-ice motions and lead formation occur under tidal shear or conditions of high wind-stress divergence, and large resulting fluxes of sensible and latent heat cause rapid ice growth. Resulting production of cold salty shelf water participates in Antarctic bottom-water mass formation, and to some extent in driving horizontal and vertical thermohaline circulation.

The majority of mixing and heat exchange in the ocean boundary layer is induced by momentum transfer to the sea ice surface during frequent storm bursts, and especially during the passage of fast-moving low pressure systems. Such noteworthy periods of dynamic-thermodynamic changes in the ice cover, however, are more often than not accompanied by a blanket of cloud due to atmospheric radiation-feedback mechanisms. In these cases, the atmosphere is inherently more electromagnetically opaque, and retrievals of ice characteristics information using satellite passive microwave algorithms are called into question. Recent studies indicate that traditional ice concentration estimates can be in error by values exceeding 10% under typical storm conditions.

Microwave radar satellites, with wavelengths in the range 2-6 cm are the only other uninterrupted source of weather-independent images. Since 1991, an array of international satellites have been acquiring microwave radar data over the Southern Ocean sea-ice cover. The C-band Active Microwave Instrument on board ERS, for instance has two modes; (a) Synthetic Aperture Radar (SAR) and (b) Scatterometer. Together, these combined microwave data enable uninterrupted imaging of this geographic region with (a) mesoscale coverage and high resolution (25m); or (b) global coverage with medium-scale (~12 km) enhanced resolution. Together with the more recent additions of Ku-band data from the NSCAT scatterometer, and the C-band RADARSAT SAR, these combined radar datasets are now routinely used to measure ice kinematics and surface conditions in response to meteorological forcing.

Together, these new datasets yield a variety of important new information on the seasonal to interannual variability in ice dynamics, and the processes linking storms with mesoscale ice divergence. 3-day gridded motion fields are validated using buoy drift trajectories, and combinations of entire years of ice motion used to derive climatological mean ice motion around Antarctica. These data indicate a large interannual variability in both seasonal and mean annual drift, which in turn appear linked to climatic anomalies in sea-level pressure and meridional wind stress derived from ECMWF analysis fields. Similarly, periods of high wind-stress divergence result in openings of the ice cover and notably a brief reappearance of the winter Weddell Polynya in 1994, which was first discovered in the Scatterometer images.

Active microwave radar observations in the Southern Ocean are actively contributing towards revolutionizing the study of Antarctic sea-ice geophysics. Seasonal patterns of global sea-ice drift together with the accompanying transitions in sea-ice characteristics can now be well characterized. The longer these high-resolution satellite radar records are extended, the better the chances of measuring climate-related timescales of Antarctic sea-ice variability.

This work was conducted at Jet Propulsion Laboratory, California Institute of Technology, under contract to NASA. Funding support was provided by Robert Thomas (Code YSG) in the NASA Office for Mission to Planet Earth.

IP16h

Sea Water Flooding, Snow Ice Formation and the Mass Balance of First–Year Ice Floes in the Ross, Amundsen and Bellingshausen Seas: II Modeling

Ted Maksym and Martin Jeffries

Geophysical Institute, University of Alaska Fairbanks, USA

The thickness distribution and mass balance of the Antarctic sea ice cover is governed by a variety of thermodynamic and dynamic processes. Key among these is sea water flooding of the ice surface and the subsequent formation of snow ice. Observations from four cruises in the austral winters of 1993, 1994 and 1995 aboard the R.V. Nathaniel B. Palmer indicate the snow ice comprises a substantial portion of the total ice thickness in the Pacific Sector of the Southern Ocean and may be the dominant thermodynamic mechanism for ice thickening. Flooding and snow ice formation have a complex and profound impact on sea ice development in that they affect the mass balance of both the ice and the snow, dramatically alter the thermal regime, and represent a significant transport of brine through the ice, all of which substantially affect the thermophysical properties of the ice and the snow. Therefore an understanding of the processes involved in snow ice formation and their impact is crucial to understanding the heat and mass balance of the Antarctic ice pack and the salt flux from the ice to the upper ocean.

A one dimensional numerical sea ice growth model is developed, which examines the role of flooding and snow ice formation in governing sea ice development. The evolution of the flooded layer is examined in detail in order to evaluate the effects of variation in the thermal regime due to snow cover variations, effects of ice permeability, and brine exchange processes on the development of snow ice and the overall heat and mass balance. In order to investigate the influence of processes within the snow cover, such as depth hoar formation, a simple parameterization of snow pack evolution is developed based on observations of the structural and physical properties of the snow.

Simulation results are presented for current climatic conditions during the growth season for the Pacific sector of the Southern Ocean. Observations of snow and ice thickness and structure from the four cruises are used to constrain the model. Snow accumulation rates are estimated from observations of snow depth and structure from the four cruises and from observed accumulation rates on the Antarctic coast. Model results of snow ice mass balance and salinity characteristics are compared to the observed data. Sensitivity of snow ice formation to variations in the timing and nature of snow cover accumulation and to variations in climatic forcing are examined. Heat and salt fluxes are evaluated to investigate the importance of snow ice formation in atmosphere-ice-ocean climatic interaction.

IP16i

Spatial and temporal wave scattering in the marginal ice zone

Michael H. Meylan

Dept. of Mathematics, University of Auckland, NEW ZEALAND

Ocean wave action is believed to play a critical role in determining the structure of the marginal ice zone. Two processes need to be understood for an effective model to be developed; the breaking of a continuous or broken ice cover by ocean waves and the scattering and dissipation of wave energy by a continuous or broken ice cover.

A model for wave scattering by a broken ice cover has been developed by the author. This model is based on a solution for the wave scattering by an individual ice floe where the floe flexure is included. This solution provides a kernel for a Boltzmann scattering equation which models the spatial and temporal evolution of ocean wave intensity. Solution of this equation for temporally or spatially independent solutions is mathematically straight forward and these results have already been presented. Extending the solution to the fully spatial and temporal solution remains a significant mathematical challenge.

IP16j

Study of small scale ice dynamics due to ocean wave action

Igor Lavrenov, Viacheslav Polnikov and Viacheslav Alekseev

Arctic and Antarctic Research Institute, St.Petersburg, RUSSIA

Ice cover dynamics research is one of the most urgent problems of modern oceanography. It is of a great importance both for analysis and forecast of ice cover state and for the climatic problems’ solution. A large number of results important from scientific and practical view points has been obtained at solving this problem. On the background of these results the lag in studying the small scale dynamic processes is especially noticeable.

One of the most important factors causing the small scale dynamics of ice floes is wave perturbations connected with sea wave motion. This changes many ice cover parameters during a short period of time.

In this article the interaction between ice cover and sea waves is considered on the basis of Hamiltonian formalism (Hasselmann 1968, Zakharov 1968, 1972; Krasitsky, 1990, 1993) for describing the non-linear wave processes in the sea covered by ice. The definition of the Hamiltonian function for the system "layer of liquid - ice floes", finding the canonical variables and the Hamiltonian function in the Fourier form for the canonical variables, derivation of the kinetic equation and integral of collisions, calculation of collisions integral are carried out.

By using these theoretical results we create a spectral model of wave energy evolution in the marginal ice zone. The model takes into account the refraction and transformation of the waves as well as non-linear redistribution of energy within the wave spectrum. A phenomenon of ‘ice storm’ is modelled on the basis of the ice floes motion on the waves.

IP16K

Percolation in Sea Ice

K. M. Golden, Stephen F. Ackley and Victoria I. Lytle

Department of Mathematics, University of Utah, Utah, USA

USACRREL, Hanover, N.H., USA

Antarctic CRC, University of Tasmania, Hobart, AUSTRALIA

The 1994 ANZFLUX winter cruise to the Eastern Weddell Gyre was designed to investigate air-sea-ice interactions in this region, which sustained a vast polynya during the early 1970’s. On this cruise we found a "stable", relatively thin ice pack (typically 30-60 cm), as opposed to a large polynya. However, large vertical oceanic heat fluxes were measured as well, giving rise to bottom melt rates of up to 3 cm per day. We have found that the persistence of the pack depends on freezing of a slush layer on top of the ice due to surface flooding, so that the ice thickness is in a dynamic rather than static equilibrium.

Furthermore, a predominant mechanism behind the flooding is a percolation phenomenon in the sea ice, which renders the ice porous and allows fluid and heat transport through the ice. In particular, the sea ice is a composite of pure ice with brine and air inclusions, and the brine phase becomes connected, or percolates, above a critical temperature around -5ƒ C (for 5 ppt salinity). Through analysis of the temperature profile over the course of the drift camps, we have discovered that while most of the sea ice layer remains above the critical temperature, the top 10 cm or so of the ice acts as a "faucet" which turns on during the typically warm storms, allowing flooding, and turns off during colder periods, where the resulting slush layer freezes. This transport of fluid through the ice enhances the transfer of heat from the ocean to the atmosphere, primarily through the release of latent heat during freezing on the surface, and has been found previously to play a major role in autumnal algae blooms in the ice.

We have also found that the cyclic flooding/freezing process influences the evolution of the microwave backscatter signature of the ice. Our findings have led us to model the evolution of the temperature profile with a nonlinear heat equation, incorporating a source term arising from the upward percolation of warmer brine. Furthermore, due to the central role that the percolation theshold plays, we have used the similarity of the sea ice microstructure to that of compressed powders to predict the -5ƒ C critical temperature. Our application of percolation theory to sea ice represents a new example in a broad array of disordered materials whose transport properties have been studied with percolation models, including sandstones, doped semiconductors, smart materials such as piezoresistors and thermistors, radar absorbing composites, and thin metal films.

IP16l

Some Sources of Acoustic Emission in First Year Sea Ice

P.J. Langhorne and T.G. Haskell

Department of Physics, University of Otago, Dunedin,
New Zealand
Industrial Research Limited, Gracefield, Lower Hutt,
New Zealand

Acoustic emissions have been routinely measured during flexural strength, single load cycle and fatigue experiments on the first year sea ice of M^cMurdo Sound, Antarctica. These emissions provide insight to the types of processes taking place during application and release of stress within the ice. The experiments, which took place in approximately the same location and span three field seasons in 1992, ‘94 and ’96, employ in situ, cantilever beams of sea ice nominally 10m long by 1m wide by 2m thick. This paper will present on overview of the numbers, types and implied sources of emissions measured under a variety of stress and stress history states. The emissions are also related to the measured physical properties of the sea ice sheet.

IP16m

The New Zealand Sea Ice Programme

Vernon A. Squire

Department of Mathematics and Statistics, University of Otago, Dunedin, New Zealand

The first stage of a multifaceted New Zealand Programme to study sea ice breakup in relation to its physical and mechanical properties nears completion in the Ross Sea, Antarctica. Sea ice fatigue, acoustic emissions due to microcracking, physical properties via coring and nuclear magnetic resonance, thermal structure and wave-related experiments have been done, each ultimately being drawn together to give an overarching view of how the sea ice of the Southern Ocean changes with time due to fracture by ocean waves. Remote sensing imagery is used to make the final step between process-oriented studies and the larger picture.

This talk will introduce the Programme and will be followed by focused presentations on its various components.

IP16n

Response of First Year Sea Ice to Cyclic Loading

S. Frankenstein, T.G. Haskell and P.J. Langhorne

Cold Regions Research and Engineering Laboratory, Hanover, USA

Industrial Research Limited, Gracefield, Lower Hutt,
New Zealand

Department of Physics, University of Otago, Dunedin,
New Zealand

Two series of in-situ measurements of the response of first year sea ice to cyclic loading were made during October and November, 1996 in the McMurdo Sound region of the Ross Sea, Antarctica. The first series consists of the mechanical load and resulting displacement histories of several 1m ¥ 10 m cantilevered beams. The second data set includes the strain history of the ice as a result of ocean wave forcing. In both instances, the thickness and temperature, density and salinity profiles of the ice were measured from cores taken at the time of loading. Attempts to fit various constitutive equations to the data are discussed as well as the comparisons of the equations between the two series.

IP16o

Fatigue Characteristics of First Year Sea Ice

T.G. Haskell and P.J. Langhorne

Industrial Research Limited, Gracefield, Lower Hutt,
New Zealand

Department of Physics, University of Otago, Dunedin,
New Zealand

Measurements of the fatigue characteristics of sea ice will give an indication of its possible failure mechanisms under repeated loading, such as occur with wave action and the operation of vehicles and aircraft. The result of a series of fatigue experiments, based on three field seasons in 1992, ‘94 and ’96, are presented here. The primary result is that the endurance limit of the sea ice appears to be approximately half the ultimate flexural strength.

IP16p

Measurement of Ice-Coupled Wave Coherence in McMurdo Sound

Colin Fox

Department of Mathematics, The University of Auckland, Auckland, New Zealand

When the break-up of fast ice is induced by ocean-wave flexure, quantative details of the break-up depend on the coherence length and coherence angle of the wave field. Measurements of the coherence lengths of the ice-coupled wave field have been made using an array of strain gauges deployed near the edge of the fast ice in McMurdo Sound, Antarctica. Estimates of the wave-field coherence based on those measurements will be presented along with some implications for modelling of fast-ice break-up.

IP16r

Multi-decadal ice-ocean-atmosphere interactions in the Arctic Ocean

Siobhan P. O'Farrell

CSIRO Division of Atmospheric Research, Aspendale, AUSTRALIA

A simulation with a global coupled ocean atmosphere model including dynamic thermodynamic ice provides a data source to examine the interaction of these three components of the climate system on a range of time scales. The focus of this paper is on the low frequency variability in the ice thickness and concentration distributions in the Arctic in a multi-century simulation of the climate model. The modelled ice responds to the variability in external forcing in both the ocean (e.g. changes in the SST pattern in the North Atlantic and Greenland seas which are linked to oscillations in the North Atlantic overturning streamfunction) and the atmosphere (e.g. location of Aleutian and Icelandic low pressure systems which determine storm track pathways).

The strongest signal in the variability of the ice volume and ice extent is on a 30 year time scale in both the Western and Eastern sectors of the Arctic and this signal is correlated strongly with changes in North Atlantic meridional overturning. The rate and location of ice production and loss in the Arctic and the flow rate of ice exiting the Arctic influences the surface salinity of the Greenland and Labrador seas. These changes in surface salinity represent part of the feedback loop that switch the magnitude of the oscillation in the North Atlantic overturning streamfunction and alter rates of North Atlantic deep water formation.

The anomalies in air temperature, surface pressure and barotropic ocean streamfunction, to match the maximum and minimum ice volume states, were calculated. The air temperature anomalies in the winter season were > 1 degree C over the Arctic basin and northern landmasses. The sign of the anomaly in the surface pressure systems and the strength of the sub-tropical gyres in both the Atlantic and Pacific oceans are also correlated with the ice volume/extent anomalies. However, the higher frequency signal of the North Atlantic Oscillation pressure index does not in this model have an impact on the ice extent in the North Atlantic as has been suggested by observations. As well as impacting on the North Atlantic circulation, the ice anomalies cause changes to the rates of brine rejection and freshening in the Arctic ocean basin and hence generate changes in the local ocean dynamics. In this paper the mechanisms involved in the coupling of this low frequency ocean-ice-atmosphere signal in the Arctic will be examined to see what time lag is involved in the growth of anomalies and whether the ice has an active role beyond being a source of fresh/saline water in integrating the processes involved and maintaining the signal over longer periods.

A further aspect of low frequency variability related to the Northern Hemisphere ice cover is seen in the coupled model simulations. Cool events occur in both the North Pacific and North Atlantic which persist for several years and often maintain ice for the winter months. These events will be investigated further, in particular to see if there is a link between the sea ice extent in the North Pacific and a Latif-Barnet coupled mode.

IP16s

A nonhydrostatic model of dense water formation in the partially ice-covered ocean

N. Okada, S. Minobe and M. Ikeda

Graduate School of Environmental Earth Science, Hokkaido University, Sapporo, Japan

Graduate School of Science, Hokkaido University, Sapporo, Japan

In the ocean partially covered by sea ice, nearly uniform atmospheric cooling can induce much more intense cooling through an open water area, and hence, more rapid brine rejection from ice formation than in the ice-covered portion. This nonuniform negative buoyancy flux produces a unique situation of convection: an area of convection is comparable with a convective plume.

A 3-dimensional ocean model is developed without a hydrostatic assumption. The model domain is a square in the plan view with the side lengths of 32km and a depth of 1000m. The spatial resolution is 250m in the horizontal directions and 100m in the vertical direction. The negative buoyancy flux is given to the initially homogeneous water for 6 days in the smaller square areas uniformly distributed in the model domain. A parameter study is carried out with their sizes varied from 250m x 250m to 16km x 16km. Here, the total area of forcing square is kept unchanged over the domain.

The following results are obtained:

- When each forcing square is large, the horizontally averaged density is larger near the bottom than at the surface. As the square size reduces, the surface water becomes lighter than the bottom water. The critical size between these two stratified states is similar to the size of the convective plume with a diameter of 2km.

- The larger each forcing square is, the heavier the heaviest water in the domain is. Thus, the dense water produced by brine rejection varies depending on the sizes of lead and polynia. This information is extremely useful for parameterizing dense water formation under the ice cover in a numerical model with a large (a few tens of km or larger) grid size. In order to simulate the ice cover ocean, we should include a size distribution of lead and polynia in addition to grid-averaged ice concentration.

IP16t

Snow cover on Arctic sea ice

S. Warren, I.G. Rigor, N. Untersteiner, V.F. Radionov, N.N. Briazgan, Y.I. Aleksandrov and R. Colony

University of Washington, Seattle, USA

Arctic and Antarctic Research Institute, St Petersburg, RUSSIA

International ACSYS Project Office, Norsk Polarinstitutt, Oslo, NORWAY

Snow depth and density were measured at Soviet drifting stations on multiyear Arctic sea ice, beginning in 1937. Measurements were made daily at fixed stakes at the meteorological ground, and once- or thrice-monthly at 10-m intervals on a line extending outward 500 or 1000 m from the station. Analyses are performed here for the 37 years 1954-1990, during which time at least one station was always reporting.

Snow depth at the stakes was sometimes higher than on the lines, and sometimes lower, due to influences of the station, but no systematic trend of snow depth was detected as a function of distance from the station along the 1000-m lines. In forming the seasonal cycles of snow depth for each year at each station, priority was given to snow lines if available; otherwise the fixed stakes were used, with an offset applied if necessary to account for the station’s influence. Reporting of snow depth ceased each summer when the snow melted completely. Therefore, for stations that were continuously occupied during summer, we insert a snow depth of zero through the summer until the record resumes with the first autumn snowfall.

The average snow depth is near zero in August, and increases to a basin-average maximum of 34 cm (11.5 g/cm²) in May. The deepest snow is just north of Greenland and Ellesmere Island, peaking in early June at 46 cm, when the snow is already melting north of Siberia and Alaska.

The snow accumulates rapidly in September and October, moderately in November, very slowly in December and January, then moderately again from February to May. This pattern is seen on the basin average, but is most pronounced in the Greenland-Ellesmere sector, which shows almost no net accumulation from November to March. The Chukchi region shows a steadier accumulation throughout the autumn, winter, and spring.

Usually only one or two stations were in operation in any particular year, so there is insufficient information to obtain the geographical variation of interannual anomalies. Therefore, to represent the geographical and seasonal variation of snow depth, a two-dimensional quadratic function is fitted to all data for a particular month, irrespective of year. Higher-order fits are not justified for the available data density. We then express the geographical variation as contour maps of multi-year average snow depth (and snow-water-equivalent) for each month.

So as not to confuse geographical variation with interannual variation, interannual anomalies for each month of each year are obtained relative to the long-term mean snow depth not for the basin average, but rather for the geographical location of the station operating in that particular year. Weak negative trends of snow depth are found for all months. The largest trend is for May, the month of maximum snow depth, a decrease of 8 cm over 37 years. This trend, however, is not significantly different from zero, because of the large interannual variability: the standard deviation for 37 months of May is 6 cm.

Ip16u

Comparison of Heat Transfer Process Between the Chukchi and Okhotsk Ice Seas

G Naito, Y Sasaki and Y Muraji

Dept of Geoscience, National Defence Academy, Yokosuka, japan
Earth Science and Technology Organisation, Tokyo, Japan
Energysharing Co Ltd, Japan

Vertical heat fluxes were observed in two ice seas; one is the Chukchi Sea of the Arctic Ocean and the other the Sea of Okhotsk located lowest in latitude of all ice seas. The heat budget between the marginal sea and the atmospheric boundary layer in higher latitudes is estimated from comparison of turbulent and radiative fluxes for two seas.

Snow cover on sea ice plays a very important role in the heat transfer in ice sea, because dry snow layer on sea ice has rather lower thermal conductivity and much higher albedo than snow-free ice surface.

Net radiative flux is small and flows upward or downward for temporal sea ice conditions. Both sensible and latent heat fluxes are upward and much larger than the radiative flux. Resulting in the fact that the large heat transfer is always upward in the Chukchi Sea, however, both sensible and latent heat fluxes are small and the total heat exchange is almost zero in the Sea of Okhotsk.

IP16v

Floe Buoys: Time Series of Atmosphere-Ice-Ocean Interaction From Sea-Ice Core Analysis

Hajo Eicken

Alfred-Wegener-Institute for Polar and Marine Research, Bremerhaven, Germany

The exchange between ocean and atmosphere in the polar regions depends significantly on the transfer of energy through the ice cover. This, in turn, is controlled to a large degree by the properties and the microstructure of the ice. Apart from modifying atmosphere-ocean interaction, the growing sea-ice cover may also represent an archive of a number of processes relevant in this context. This presentation will discuss the importance of ice properties and microstructure in determining fluxes of energy and mass through the ice cover. Particular attention will be paid to the question, whether parameters such as ice texture or the stable-isotope composition may provide a quantitative record of atmosphere-ice-ocean interaction, such that sea-ice cores could provide time series of relevant parameters, very much similar to data buoys.

For a set of ice, brine and water samples obtained along a transect through the Weddell Sea, the dependence of texture, salinity and O-18 concentrations of sea ice on the mode and rate of ice growth were studied. Apart from identifying the contribution of snow or meteoric ice to the total ice thickness, a stagnant boundary-layer fractionation model has been utilized to explain growth-rate dependent fractionation of O-18 for sea ice. The fractionation model has been validated through comparison of averaged delta O-18 profiles with simulation results from a one-dimensional ice-growth model. Based on this intercomparison it is shown that growth rate and, under given boundary conditions at the top surface, the oceanic heat flux may be derived from ice-core data.

The results are discussed with respect to the ice-growth regime in the Antarctic. Furthermore, comparisons will be made with core data from the Arctic Ocean, pointing out the similarities and differences between the ice-growth regimes and their linkage to ice properties and microstructure.

IP16w

Thermal Transport and Convection in Sea Ice

H.J. Trodahl, K. Collins and M. McGuinness

Department of Physics, Victoria University of Wellington, Wellington, New Zealand

Department of Mathematics, Victoria University of Wellington, Wellington, New Zealand

The exchange of energy between the ocean and the atmosphere is controlled, over a large fraction of the Earth’s polar regions, by the transport of heat through sea ice. Despite the importance of this process, the thermal conductivity assumed in climate and sea ice models is based on a prediction, rather than on a measurement. Several years ago we set up a thermal diffusivity experiment designed to determine this parameter. Initial measurements indicated a value approximately 25% higher than expected, a result which has been reported earlier. More recent (and more accurate) measurements have established that the excess heat flow occurs only at temperature gradients larger than about 10ƒ C per metre, which suggests that convection in brine channels, or migration of brine pockets, may be responsible.

ip16x

decadal simulations of the ice cover and thermohaline structure in the arctic-gin sea basin

S. Piacsek, R. Allard, P. Posey and P. Jayakumar

Naval Research Laboratory, Stennis Space Centre,
Mississippi, USA

Centre for Marine Science, University of Southern Mississippi, Stennis Space Centre, Mississippi, USA

The coupled ice-ocean system in the Arctic-GIN Sea Basin has been simulated for the period 1978-1996, using atmospheric forcings from the NMC reanalysis effort (1978-1995) and the US Navy’s global forecast products (1986-1996). The results are analysed from four principle aspects: (1) relation between ice extent/volume and ocean theromhaline structure; (2) ocean circulation with and without ice cover; (3) comparison of ice cover with microwave imagery and (4) comparison of results obtained with the different forcings.

In general, the range of interannual variations in ice cover were found to be much greater in the summer than in the winter, with maximum summer ice volumes occurring in the 1986-1988 period and strong minima in 1990, 1991 and 1993. the ice thickness variations at the Pole agreed generally well with the submarine observed drafts. We have found an inverse correlation between annual wind strengths and ice cover in the 1986-1990 period. The GIN Sea has acquired a cold bias in the years 1986-1990 and the corresponding watermass census exhibited anomalously large volumes of Greenland Sea and Norwegian Sea Deep Waters (GSDW and NSDW), as well as Lower Arctic Intermediate Water (LAIW). The simulations produced as excess of ice in the Barents and Greenland Sea in winter and a deficiency in the Greenland Sea in the summer, when compared to SSM/I remotely sensed data.

IP16z

Seasonal change of the surface heat flux over the Southern Ocean from ECMWF atmospheric data and ERBE radiation data in 1985-1989

I. Okada and T. Yamanouchi

Centre for Environmental Remote Sensing, Chiba University, Chiba, JAPAN

National Institute of Polar Research, Tokyo, JAPAN

The heat budget of the earth has net income at low latitudes and net outgo at high latitudes, and heat transport by the atmosphere and the ocean exists to maintain the budget. The Southern Ocean is one of the places where a large amount of heat is released to the atmosphere from the ocean. And sea ice, which exhibits a large amplitude for seasonal change there, may affect surface heat exchange considerably.

However, the effects are not well known, since it is impossible to have routine in-situ observation, and to use satellite observation data due to the ice surface. There have been few data taken in the Southern Ocean. But efforts to estimate surface heat flux using climatology have been made. And many numerical experiments show the relationship between sea ice variation and surface heat flux; however few experiments consider temporal and spatial change (eg. seasonal change).

The quality of the atmospheric objective analysis data which are supplied from the weather forecast center such as ECMWF (European Centre for Medium-Range Weather Forecast) has been improved and produces more reliable interpolated values than that in the past. ERBE (Earth Radiation Budget Experiment) operated to obtain global shortwave and longwave radiative flux at the top of the atmosphere with high accuracy. The present study aims to show the seasonal change of the atmospheric heat budget, especially the surface heat exchange, calculated from these data in the seasonal sea ice area over the Southern Ocean, and to relate the surface heat exchange with the seasonal change of the sea ice.

An atmospheric belt around the Antarctic continent, bounded by 60 degree S, 70 degree S, the top of the atmosphere, and the surface, was analysed in the present study from satellite observed data over 60% of the region is covered by sea ice at the maximum. The atmospheric heat budget was analysed from January 1985 to December 1989. The tendency of the atmospheric heat storage and the convergence of the meridional atmospheric heat transport from ECMWF/WMO atmospheric data, and net radiative flux at the top of the atmosphere from ERBE radiation data were calculated as zonal and monthly averages. The surface heat exchange was obtained from the heat budget equation as the residual of these known terms.

The resulting surface heat flux was about -100 W/m² (the ocean gained heat) at the minimum value in December, and about 100 W/m² (the ocean lost heat) at the maximum in May. The curve shows rapid increase toward winter and slow decrease toward summer. The surface heat flux decreased 30 W/m² from May to July, when the average sea ice concentration in the same area increased from 33% to 60%. This contrasts remarkably with the band from 50 degrees S to 60 degrees S where sea ice is less prevalent. The surface heat flux in this area shows a flat maximum from May to July.

The heat flux for freezing and melting sea ice was simply estimated from the averaged area of sea ice and compared with the surface heat flux from the atmospheric heat budget: it was assumed that the thickness of the sea ice was constant. The surface heat flux can explain the seasonal change of sea ice of 2 m thickness, and both the variations are in phase. It is deduced from this similarity of both estimated fluxes that the atmospheric heat budget includes sea ice information, although it is a widely averaged value. And the similarity shows that it may be possible to analyse the inter-annual variation of the surface heat flux from the atmospheric heat budget.

IP16aa

Radiative Transport in First-year sea ice

H.J. Trodahl, R.G. Buckley and E.M. Haines

Victoria University of Wellington, Wellington, New Zealand

New Zealand Institute for Industrial Research and Development, Lower Hutt, NEW ZEALAND

Radiative transport in sea ice is diffusive but with an optical depth insufficient to model the transport with the diffusion equation. In that situation it is not an easy matter to develop an experimentally-based characterisation of the process. Most attempts to do so have relied on measurements of the transmission and albedo for sunlight falling on the ice, and the results do not provide clear direct guidance for the depth dependence of the scattering coefficients. As a consequence work in the field has focussed on models of physical structure of sea ice, with data used only as a comparison with predictions of radiative transport in those models.

We have developed an experiment intended to provide direct experimental evidence of the depth of dependence of the scattering coefficients. Monochromatic light is introduced at a point on the surface and the profiles of light emerging from the top and the bottom surfaces are measured. Direct evidence is found for a strongly scattering surface layer, and anisotropic bulk layer, and an absorbing algal layer. Measurements carried out on first-year sea ice in McMurdo Sound over several years reveal dependencies on season and on ice structure. Some of the results have been discussed in earlier publications; this presentation will bring together data from the 1986-92 seasons and extensive Monte Carlo modelling performed as an aid to their interpretation.

ip16bb

ICEBERG INDUCED UPWELLING AS AN ORIGIN FOR a CHIMNEY

L.G. Pisarevskaya

The State Research Center of Russian Federation - Arctic and Antarctic Research Institute, St.-Petersburg, Russia

Estimations of the water upwelling induced by air bubbles release while glacier ice is melting are presented. At first melting rates are evaluated for the bottoms of icebergs drifting in ambient water with temperature from -1 to 3ƒ C and a relative speed from 1 to 30 cm per second. Then for several values of air bubble content by volume - from 1 to 15% - the decay of an ice path along the iceberg bottom of 1 m width is evaluated. Finally the values of induced volume discharge of water delivered to the sea surface, the width at the sea surface of the upwelling plume, and the vertical velocities in the plume are presented depending on the iceberg draft and air volume discharge. For a middle-sized iceberg the volume of upwelled water can reach about 1000 m³ s^-1 . These estimations are applied to field measurements of the vertical distribution of temperature and salinity 50 m from an iceberg drifting among scattered ice floes to explain the observed warm water upwelling at 100 m. Since the water brought to the surface is more dense, once the air is released, than the ambient water, it has a tendency to plunge back. Due to irregularities of the process ( for example, existence of layers with high and low air bubble content in glacier ice) there will be formed a layered system. Thus mixing of Warm Deep Water with the overlying Winter Water will finally result in open ocean convection with double diffusion processes in the layered system acting to average temperature and salinity towards a chimney structure.

IP16cc

Sea Ice Thermodynamics

Donald K. Perovich

USACRREL, Hanover, NH, USA

An understanding of sea ice thermodynamics is critical in determining the heat and mass balance of the ice pack and in assessing potential impacts to the ice cover due to climate change. A key aspect of sea ice thermodynamics is the interaction of solar radiation with the ice. This is particularly important in the Arctic where solar radiation drives the summer decay cycle. This decay is influenced by the ice-albedo feedback mechanism, as illustrated by the profound changes the ice cover undergoes during summer melt. At the onset of melt the ice is snow-covered with very little open water present and the areally-averaged albedo is near 0.8. As the melt season progresses, the snow melts and the ice surface becomes a variegated combination of bare ice, drained hummocks, and melt ponds, with a decrease in albedo to roughly 0.5. The amount of open water increases due to melting and divergence of the ice pack, which also results in an increase in absorbed solar radiation. These large changes in the surface albedo provide the potential for positive feedback between ice albedo and ice melting. Solar radiation absorbed in the ice contributes to surface melting and internal melting. Radiation deposited in the upper ocean results in lateral ablation on the edges of floes and bottom ablation on the underside of floes. Leads play a major role in the melt cycle absorbing over 90% of the incident solar energy. How this energy is partitioned between lateral melting, bottom ablation, and storage in the water column significantly affects the overall melt cycle. Lateral melting increases the amount of open water, which increases the heat input to the system constituting a positive feedback mechanism. Melt ponds have a lower albedo than bare ice, so their areal coverage and duration are important considerations when assessing ice-albedo feedback.

IP16dd

Numerical modeling of seasonal ice cover evolution in the Sea of Okhotsk

Tatsuro Watanabe and Motoyoshi Ikeda

Japan Sea National Fisheries Research Institute, Fisheries Agency, Niigata, Japan
Graduate School of Environmental Earth Science, Hokkaido University, Sapporo, Japan

The Sea of Okhotsk is one of the lowest-latitude ice covered seas in the world. Sea ice first appears at the northwestern corner of the Sea and spreads southward as winter progresses. The ice cover is observed from December to April, although its duration and area are different from year to year. Atmospheric conditions (air temperature and wind) are considered to be major mechanisms to produce the inter-annual variabilities in the ice cover.

In order to explore the mechanisms of the ice cover evolution, a coupled ice-ocean model is developed. The ice model component was taken from the dynamic and thermo-dynamic model developed by Hibler (JPO, 1979), while the ocean component is identical to the GCM developed by Bryan and Cox (1984 version). The coupling between these models was implemented with vertical transfer of horizontal momemtum and heat content. The model sea is so far isolated from the exterior regions: Soya and Kuril Islands straits are closed. The model is initialized with the fall condition of the Levitus climatology, and forced with observed atmospheric data from the beginning of November for 7 months. The model is run for different years separately from other years. Sensitivity of the ice cover to atmospheric conditions is extensively examined.

The seasonal ice extent and the year-to-year variability are well reproduced in the model. In addition to the ice cover, water formation is also studied: ice is pushed southward by wind and melts in the marginal ice zone in early winter. Melting ice supplies fresh water to the mixed layer, being responsible for the shallow, fresh and cold mixed-layer. Coastal downwelling is observed at the northwestern corner with large density water. It is suggested a large amount of ice forms there.

IP16ee

Modeling the surface turbulent fluxes from a freezing lead

A. Alam and J.A. Curry

University of Colorado-Boulder, USA

Leads provide a significant source of heat and moisture to the arctic winter atmosphere and may have a major impact on the polar climate. We have developed a new model to compute turbulent surface heat and momentum fluxes over leads in the Arctic sea ice. The momentum roughness length uses a sea state parameterization which is fully consistent with the surface turbulent flux parameterization. The surface roughness length for heat is determined from an application of surface renewal theory to the air-sea interface. The flux parameterization accounts for the fetch limitation of the airflow over a lead. The model has been evaluated using the Arctic Ice Dynamics Joint Experiment 1974 surface flux data and compared with a bulk flux algorithm which has been commonly used to evaluate surface heat fluxes from leads. The large heat release from the lead surface results in the formation of frazil ice crystals at the surface. These frazil crystals are advected downwind by both the wind waves and a wind-driven surface current to collect at the downwind edge of the lead. The piled-up ice advances to the upwind edge with time as ice formation and downwind advection continues. The present model takes account of the freezing at the lead surface and the resultant effective open water and surface turbulent fluxes.

We have computed time-dependent integral heat fluxes as a function of lead width/fetch for various atmospheric states to determine the magnitude of heat flux in a mesoscale model grid in which a lead is present.

IP16ff

Turbulent mixing in Arctic leads

T.P. Stanton

Naval Postgraduate School, Monterey, USA

The ONR-sponsored Leads Experiment provided an opportunity to measure time series of microstructure properties in the oceanic boundary layer under freezing leads during field programs in the Arctic Ocean north of Alaska in 1991 and 1992. Measurements were made for several days at downcurrent sides of newly formed leads by deploying huts and instrumentation within helicopter range of a central camp. At each lead the automated Loose Tethered Microstructure Profiler (LMP) continuously profiled the water column from the surface to 75m depth, spanning the ~35m deep mixed layer and upper pycnocline. The LMP was equipped with a microscale shear probe, fast fp07 thermistor, and a microconductivity cell, providing cm resolution estimates of salinity and temperature structure while resolving the thermal and turbulent kinetic energy gradient spectrum to produce estimates of thermal dissipation rates, c and kinetic energy dissipation rates e . The c estimates have been successfully used with very high resolution thermal gradients to estimate heat fluxes and thermal diffusivities within both the mixed layer and pycnocline.

Results from three lead deployments are presented, with the first having higher insolation spring conditions, and the other two representative of late winter freezing conditions. A 24 hour observation of a 1 km wide lead made in late April 1991, with low ice-ocean currents showed a diurnal weakly stratified mixing layer reaching as deep as 15m, as a result of near-surface solar heating and ice melting. At the peak of the solar heating a fresh melt layer with 0.1 psu salinity jump extended down to 15m, with low, surface boundary-driven turbulent dissipation levels dropping from 2 x 10^-6 near the surface to 2 x 10^-8 W kg^-1 below the weak stratification. The fresh layer largely mixed and advected away through the following 12 hours. These conditions will be contrasted with the effects on the mixed layer by steady surface freezing, and the resulting large negative buoyancy flux, which occurred at leads 3 and 4 observed in early April 1992. Further comparisons will be made between lead 3, with rapid ice motion and strong downward buoyancy fluxes, and very low current conditions and moderate buoyancy fluxes observed at lead 4.

IP16gg

The origin and evolution of sea-ice anomalies in the Beaufort Sea

Bruno Tremblay and Lawrence A. Mysak

McGill University, Department of Atmospheric and Oceanic Sciences and Center for Climate and Global Change Research, Montreal, Quebec, Canada

The origin and space-time evolution of Beaufort Sea ice anomalies are studied using data and a recently developed sea-ice model, based on granular material rheology. In particular, the influence of river run-off, atmospheric temperature and wind anomalies in creating anomalous sea ice condition in the Beaufort Sea is studied. The sea-ice model is then used to track the position of an ice anomaly as it is transported by the Beaufort Gyre and the Transpolar Drift Stream out of the Arctic Basin. The results indicate that wind anomalies are the dominant factor responsible for interannual variability in the Beaufort Sea ice cover, whereas temperature and river run-off play a larger role for longer time scale fluctuations. The results also show that sea-ice anomalies originating in the western Beaufort Sea can survive a few seasonal cycles and still account for a significant amount of ice export into the Greenland Sea via Fram Strait. This is a non-negligible contribution to the fresh water budget in this area when compared to fluctuations due to variable wind speed and direction.

Ip16hh

An Overview of the sea ice mechanics (SIMI) program

Jacqueline. A. Richter-Menge

U.S. Army Cold Regions Research and Engineering Laboratory, Hanover New Hampshire USA

The processes governing sea ice deformation and failure and the forces associated with them are important to understand for a wide range of problems. These include loads on structure sand vessels located in ice infested water, the bearing capacity of an ice sheet, and the development of models that simulate the motion and the mass balance of the arctic ice pack. Compilations of laboratory test results, field measurements and parameterised estimates of the compressive stress required to fall the ice suggest that the failure stress decreases significantly as the length scale increases. Laboratory tests on 10-cm ice specimens indicate that the compressive strength of the ice is between 1-10 MPa while sea ice dynamic models assume a compressive strength of about 0.01 Mpa to achieve reasonable agreement between model results of ice motion and drifting buoy measurements. The U.S. Navy’s Office of Naval Research recently sponsored a 5-year program to investigate this apparent scale dependency, considering the mechanical behaviour of sea ice over a range in scale from 10^-2 to 10⁵ m. In addition to modelling efforts and laboratory studies, the Sea Ice Mechanics Initiative (SIMI) supported a significant field effort during the fall-winter-spring of 1993-94. Field investigations included carefully controlled, in situ mechanical property and fracture tests on specimens between 1-100 m in size. Acoustic signals generated by the pack ice were monitored to determine their usefulness in mapping future events. A stress and deformation array was established over a 20-km region of pack ice to evaluate the rheological characteristics of the ice at this scale. Early results from these various investigations highlight the importance of carefully defining the dominant models of failure in the problem of interest before comparing results at different scales.

IP16ii

Effect of a Highly Reflecting Coastline on Radiative Transfer in a Cloudy Polar Maritime Atmosphere

Igor Podgorny and Dan Lubin

Center for Clouds, Chemistry and Climate, Scripps Institution of Oceanography, University of California, San Diego, USA

A backward three-dimensional Monte Carlo radiative transfer model has been developed in order to investigate effect of a highly reflecting coastline on biospherically active insolation over Antarctic waters under cloudy skies. Specifically, we calculated both downwelling spectral irradiance in the ultraviolet and visible as well as downwelling irradiance integrated over UV-B, UV-A and PAR wavelength bands as a function of a distance from the coast line for a broad range of illumination and cloud conditions. A particular set of numerical experiments has been performed for typical springtime ozone concentrations. Finally, we used a realistic surface albedo distribution as an input for the model and computed a two-dimensional surface downwelling irradiance distribution in the vicinity of Palmer Station (65S,64W) with 200mx200m grid spacing. Along with the aforementioned applications to the Antarctic coastal regions, the model can be easily employed for calculating downwelling irradiances over marginal ice zones as well as on the boundary between ice floes and leads. The results of this study therefore have relevance to both marine ecology under episodes of springtime ozone depletion and to the surface heat budget of a polar ocean.

IP16jj

The effects of variable forcing on ice/ocean and ice/atmosphere exchange

Marika Holland, Judith Curry and Julie Schramm

Program in Atmospheric and Oceanic Sciences, University of Colorado, Boulder, USA

In general circulation model experiments, the Arctic region shows an amplified response to global warming scenarios. This suggests that the effects of anthropogenic CO₂ forcing on the climate system may first be detected in these regions. However, the natural variability which occurs in the Arctic climate system and the effect of variable forcing on the simulated ice cover is unclear. These issues are addressed with the use of a single column ice/ocean coupled model. The sea ice component contains an ice thickness distribution in which the sea ice is divided into several classes depending on thickness, area, and age. It includes a spectral treatment of the disposition of shortwave radiation through the sea ice and the effects of leads and melt ponds. A mechanical redistribution function is included which parameterizes the effects of ridging and divergence. The ocean mixed layer component of the model is an integral model with variable mixed layer temperature, salinity, and depth. The ice and ocean are coupled through the exchange of heat and fresh water. The effects of daily varying dynamic and thermodynamic forcing on the ice/ocean and ice/atmosphere exchange are examined using data from AIDJEX. The interannual variability of the ice cover is simulated using data from the Russian drifting ice stations which were present from 1950 until 1991.

IP16kk

The Sensitivity of an Ice Thickness Distribution to Parameterizations of Radiation and Surface Processes

Julie L. Schramm, Marika M. Holland and Judith A. Curry

Program in Atmospheric and Oceanic Sciences, University of Colorado-Boulder, USA

A single column ice thickness distribution model has been coupled to a two stream radiative transfer model. The domain corresponds to a grid cell at 80ƒN with a horizontal extent of 20-500 km, depending on the aggregate scale of the ice thickness distribution. A coupled one-dimensional ocean mixed layer model is also included, as is a level and ridged ice thickness distribution, a melt pond parameterization, and a complex surface albedo parameterization. Ice thickness and area, meltwater ponds, ice salinity and age, snow cover and surface albedo evolve independently for each ice thickness. Surface radiation fluxes are calculated interactively in the model, allowing the shortwave fluxes to interact fully with the surface state. Sensitivity of the surface fluxes and ice thickness distribution to details of the radiation, albedo, and meltpond parameterizations are explored. In particular, the sensitivity of the surface shortwave fluxes to cloud overlap, ice crystal precipitation, and the classification of surface types in calculating surface albedo are examined. The influence of these parameterizations on the ice albedo feedback are also described.

IP16ll

Marginal Ice Zone Meteorology

Peter Guest

Department of Meteorology, Naval Postgraduate School, Monterey, California, USA

Marginal ice zones (MIZs) are regions with intense air-sea interactions generated by the contrasts between a well-insulated surface and an ocean directly exposed to the air. During cold months, the surface heat flux over an ice floe is typically small, O(10 Wm^-2), while over an adjacent open ocean surface it is O(100 Wm^-2) upward, due to turbulent heat loss. In summer, the large differences in albedos cause the opposite contrast: much greater downward surface heat flux over the open water due to absorption of solar radiation. The surface flux gradients have a powerful effect on the dynamics and thermodynamics of the atmospheric boundary layer over MIZs. For example, in the Greenland Sea MIZ during spring the 10-m air temperature increases by 10 ƒ C, the wind speed doubles, and the depth of the atmospheric boundary layer change from 150 m to 1350 m going across a 150 km MIZ region, on the average. Surface roughness, as parameterized by the neutral drag coefficient, C_DN, has more spatial variability in MIZs than any other marine surface. Values range from 0.7 ¥ 10^-3 over grease ice to 5.5 ¥ 10^-3 over multi-year floes that have been subjected to intense rafting due to swell action. Estimates of surface momentum fluxes in MIZs must take into account these surface roughness changes as well as effects due to wind speed and atmospheric stability changes. The location of MIZs and the associated low-level baroclinicity can affect synoptic and mesoscale cyclone tracks, frequencies and intensities. In many respects, the meteorology of marginal ice zones is similar to the meteorology of coastlines. Both regions have large horizontal variations in surface forcing. However, there is an important difference: the actual geometry of the MIZ is itself affected by the atmospheric and oceanographic forcing, thus opening the door to a wide variety of possible feedback mechanisms. MIZ and nearby regions are expected to be particularly sensitive to global climate change because a small change in the location of an MIZ can strongly influence atmospheric and oceanic climate locally and on larger scales due to changes in thermalhaline circulations and other teleconnections.

IP16mm

THE ROLE OF OCEANIC HEAT FLUX IN THE HEAT AND MASS BALANCE OF ICE IN THE WEDDELL SEA

Miles McPhee

McPhee Research Company, Naches, USA

Heat entering the sea-ice/mixed layer system from below plays a critical, almost paradoxical role in maintaining the relatively compact but thin ice cover of the central and eastern Weddell Sea in winter. The paradox arises from the marginal stability of the water column. In some regions, ice growth in the absence of oceanic heat flux would inject enough salt to overturn the water column, leading to deep convections and associated heat flux sufficient to preclude further ice formation. Thus at least in a simplified view, relatively large heat flux from below maintains the ice cover, preventing much larger ocean-to-atmosphere heat flux from deep convection. During the years of the Weddell Polynya in the late 1970’s, it appears that deep convection kept winter ice concentration low over a relatively large region well within the limits of the seasonal ice pack. The talk will review current knowledge of Weddell Sea heat flux, including (i) several numerical modeling studies covering a wide range of presumed or inferred heat flux; (ii) entrainment inferred from mixed layer temperature, salinity, and other chemical properties; (iii) direct measurements of turbulent fluxes during the ANZFLUX experiment; and (iv) heat flux from drifting buoy measurements of mixed layer characteristics. Although average values of observed oceanic heat flux do not in general differ drastically from modeled or inferred estimates, it is shown that the process by which ocean heat is exchanged with the surface is quite complex, depending both on extreme turbulent mixing during storms, and or large excursions in the depth of the pycnocline (thermocline). Numerical modeling shows that both effects were about equally important during the ANZFLUX experiment, and will be used to conjecture about the impact of changes in forcing on the state of the ice cover.

ip16nn

satellite remote sensing of polynyas in the laptev sea

Lawson Brigham

Department of Oceanography, Naval Postgraduate School, California, USA

The Laptev Sea in the center of the Russian maritime Arctic is covered by sea ice for more than half the year (7.5 months in 1990 and 9 months in 1983). Russian studies in the 1960s revealed a major Laptev Sea springtime polynya and a winter flaw lead along the fast ice located in the southeastern region of the sea. Passive microwave and visible data for 1995 from the Defence Meteorological Satellite Programme (DMSP) are compared for identification of major sea ice features throughout the Laptev Sea. Sea ice concentration maps for 1979-95, derived from the satellite passive microwave record (Scanning Multichannel Microwave Radiometer(SMMR) and Special Sensor Microwave Imager (SSM/I) sensors). are analysed for the interannual variability of flaw leads and polynyas observed each May as significant sea ice features. The mid-May record reveals three distinct features: for 13 of 17 years a predominant Laptev Sea Polynya along the fast ice edge; annual recurring polynyas surrounding the islands of Severnaya Zemlya; and detectable, but less frequent polynyas along the coast of the Taymyr Peninsula. The areal extent of the Laptev Sea Polynya is shown to be as high as 18.7 percent of the sea’s total area (662,000 km²). Minimal coastal/land effects were observed in the passive microwave record. The passive microwave data also consistently showed the presence of new and young ice within the polynyas, a key observation for the estimation of sea ice production in this polar region. These studies further confirm that the SMMR and SSM/I sensors can effectively monitor regional polynyas (and other sea ice features) in the Russian Arctic coastal seas and can provide significant cluse for environmental change within the entire Arctic Basin.

IP16oo

A model for the influence of wind and oceanic currents on the size of a steady state latent heat polynya

A.J. Willmott, M.A. Morales Maqueda and M.S. Darby

Department of Mathematics, Keele University, UK
Department of Mathematics, University of Exeter, UK

We present a model for determining the size and shape of a steady state latent heat coastal polynya in terms of (1) the frazil ice production rate (F), (2) the wind stress (t), (3) the surface ocean velocity field (u), (4) the offshore consolidated thin ice transport (T), (5) the coastline shape, and (6) the intersection of the polynya edge with the coastline, all of which must be prescribed. Frazil ice trajectories are determined via the free-drift ice momentum balance. Analytical solutions for the polynya shape are derived for a straight coastline in the special case when u, T and t are uniform in the alongshore direction and the rotation of the Earth is neglected. When the latter constraint is relaxed, an expression for the asymptotic uniform polynya width is obtained. An expression for the alongshore adjustment lenght scale of the polynya associated with alongshore variations in the coastline shape, u, T and t is derived, with rotation of the Earth included. The model is applied to simulate the wind-driven polynya that sometimes forms off the northern Greenland coast during winter and early spring between the Henrik Kroyer Islands and the Ob Bank.

IP16pp

Sensitivity of a global sea ice model to the treatment of ice thermodynamics and dynamics

M.A. Morales Maqueda and T. Fichefet

Department of Mathematics, Keele University, UK
Institut d'Astronomie et de Geophysique G. Lemaitre, Universite Catholique de Louvain, BELGIUM

The sensitivity of a global thermodynamic-dynamic sea ice model to degradations of its physics is studied. The model takes into consideration the snow cover on top of ice, the storage of sensible and latent heat in the system, the influence of the subgrid-scale snow-ice thickness distributions on ice thermodynamics, the formation of snow ice, and the existence of leads in the ice cover. Regarding ice dynamics, ice is treated as a viscous-plastic fluid.

A single set of parameters is used to simulate both the Arctic and Antarctic ice regimes. The model simulates reasonably well the seasonal cycle of both ice covers. The sensitivity study focuses on: (1) vertical processes (thermal inertia, heat conduction, and snow), (2) lateral processes (leads), and (3) dynamics (ice motion and shear strength). Each experiment consisted in removing a single parameterization from the control run. Neglecting thermal inertia, has a significant effect on the model's response in the Arctic. Snow and snow ice formation largely affect the modeled ice cover in the Antarctic. The thermodynamic effect of the subgrid-scale thickness distribution, the existence of leads, and the ice motion play a crucial role in the modeled behavior of both ice packs. Ice shear strength has a non-negligible effect in both hemispheres. We conclude that all these processes should be represented in global climate models.

IP16qq

East Antarctic Sea Ice Drift: Variability and Trend

Petra Heil, Ian Allison and Victoria I. Lytle

Antarctic CRC and IASOS, Hobart, AUSTRALIA,
Australian Antarctic Div. Hobart, AUSTRALIA
Antarctic CRC Hobart, AUSTRALIA

In the polar regions sea ice is an important component of the climate system since it modifies processes taking place on the interface between ocean and atmosphere. The transfer of momentum from the atmosphere to the ocean is an important example of such a process. Compared to the Arctic, the Antarctic sea ice region is not well studied with many investigations being concentrated in the Weddell Sea, a large, nearly decoupled gyre branching off the Antarctic Circumpolar Current (ACC). In contrast, the sea ice in the East Antarctic region is a narrow band of highly mobile ice. Here we present sea ice drift data from the East Antarctic Sea Ice Zone between 40ƒ and 160ƒ E, which was collected by a total of 33 drifting buoys in selected years from 1985 onwards as part of the Australian Antarctic Research programme. Satellite-located buoys deployed on ice floes are used to measure the sea ice drift, which is frequently as high as 0.5 m/s.

The East Antarctic sea ice Zone includes Prydz Bay, a cyclonic gyre; otherwise the dominant oceanic features are the coastal current near the continental shelf flowing to the West and further north the eastward flowing ACC. A general sea ice drift pattern for the East Antarctic Zone is derived from the buoy drift, and we show the existence of persistent recirculation cells between the coastal current and the ACC. Using mean sea ice velocities and meander coefficients we identify different drift regimes in the East Antarctic region. The spatial distribution of the daily ice velocity shows that little sea ice enters the East Antarctic Zone from the eastward Ross Sea region compared to the westward export of sea ice from Prydz Bay and adjacent regions. This suggests that the East Antarctic Sea Ice Zone is a net source of sea ice within the Antarctic region. In addition the variability and the trend in the sea ice drift over the whole time series are shown.

IP16tt

Relationship between the upper ocean and sea ice during the Antarctic melting season

Kay I. Ohshima

Institute of Low Temperature Science, Hokkaido University, Sapporo, Japan

During the Antarctic ice-melting season, high resolution sea ice data were collected with a video monitoring system aboard the icebreaker Shirase, along with temperature and salinity monitoring of the upper ocean. On the basis of these data, relationships between sea ice concentration, temperature, and salinity are investigated. In the region away from ice-free ocean (ice interior region), ice concentration is negatively correlated with temperature and positively correlated with salinity for the spatially averaged data, which suggests that the local balances of energy and salt nearly hold in a bulk area. At the ice margin, ice concentration is negatively correlated with both temperature and salinity, suggesting that the local balances are overwhelmed by the effects of ice advection. XBT profiles at the ice margin also show that a considerable amount of sea ice was advected into ice-free ocean and subsequently melts there. It is pointed out that a polynya works as an "ice melting factory" in summer; it absorbs solar radiation during the period of opening, and then melts the ice advected there. From a heat budget analysis and ocean structure in the melting season, we propose a simple ice/upper ocean coupled model in which sea ice melts from the bottom and lateral faces with the heat source supplied to open water area by solar radiation. The relationships between ice concentration, temperature, and salinity derived from the model are consistent with those observed relations in the ice interior region. The negative feedback effect on the upper ocean temperature explains why the temperature-concentration plot shows similar relation for any region. Dependence of the relationships between ice concentration, temperature, and salinity on the spatial scale are also discussed.

IP16uu

Senstivity of a global sea ice model to variations in cloud fraction

T. Fichefet and M.A. Morales Maqueda

Institut d'Astronomie et de Geophysique G. Lemaitre, Universite Catholique de Louvain, Belgium
Department of Mathematics, Keele University, UK

Since downwelling radiative fluxes are significantly affected by cloud fraction, sea ice is potentially sensitive to climatic changes in cloud conditions. Numerical simulations with models of the Arctic sea ice cover carried out in the past suggest that this is case. However, these experiments have very often yielded contradictory results. Some models simulate an increase in ice thickness with decreasing cloudiness, leading to the conclusion that changes in longwave fluxes due to variations in cloud amount dominate the shortwave fluxes. Other models exhibit the opposite behaviour. It has been speculated that these discrepancies could be due to differences between the models as regards the treatment of sea ice thermodynamics and dynamics, surface radiative fluxes or cloud amount itself. Here, we present a series of numerical experiments with a thermodynamic dynamic sea ice model in which the sensitivity of the Arctic and Antarctic sea ice covers to variations in cloud fraction is investigated. The model is forced with climatological atmospheric fields. It is shown that the response of the simulated ice cover strongly depends on the way in which turbulent surface fluxes are dealt with in the model. If sensible and latent heat fluxes are parameterized using bulk aerodynamic formulae, their response to changes in the simulated surface temperature field tends to balance any modification in the radiative fluxes due to alterations in the cloud fraction. The model is actually near insensitive to alterations in cloudiness. On the other hand, if the turbulent surface heat fluxes are prescribed rather than predicted in the simulation, the model sensitivity increases drastically. It is also demonstrated that the sign of the model response depends on whether the dependence of the downwelling radiative fluxes on cloud fraction is assumed to be linear or non-linear. These results could help explaining the discrepancies encountered in previous sensitivity experiments by other authors.

IP16vv

ON THE VARIABILITY OF LAND-FAST SEA-ICE FACIES IN THE VICINITY OF ICE SHELVES

J.-L. Tison, and M. Dini

Laboratoire de Glaciologie, Departement des Sciences de la Terre et de l'Environnement, UniversitÈ Libre de Bruxelles, Belgium
Laboratorio di Geochimica Isotopica, Dipartimento di Scienze della Terra, Universitý di Trieste, Italy

Physical and chemical properties of land-fast sea ice forming in the vicinity of ice shelves often differ from those of drifting pack ice. This work presents further arguments showing the variability of land-fast sea ice facies and its potential use in demonstrating ice shelf-ocean interactions. Two sets of land-fast first year sea ice cores were retrieved in January 1994 and October 1995 in front of Hells Gate Ice Shelf (Terra Nova Bay, Ross Sea) and analyzed for their texture, ice fabrics, salinity and d ¹⁸O. Results are compared to simple laboratory experiments where small individual freshwater platy ice crystals are poured to saturation into a sea water reservoir at -1.9ƒC and subsequently frozen applying realistic constant freezing rates.

It is shown that during most of the winter, a granular orbicular facies with sub-vertical c-axes forms at the bottom of the initial sea ice cover in front of Hells Gate Ice Shelf. At the end of the winter, platelet ice with dispersed fabrics is more commonly formed. Both facies result from frazil ice production in the Deep Ice Shelf Water outflow associated with Deep Thermohaline Convection, the differences in textures and fabrics probably depending on the geometry of the ice shelf and on the thermodynamics of the genesis and consolidation processes. Bottom accretion still occurs during the summer as a distinct banded rectangular facies with vertical c-axes, only described, until now, in marine ice outcrops at the surface of Hells Gate Ice Shelf. This process appears to be related to an active circulation mode-3, in which occasional intrusions of Modified Circumpolar Deep Water can be forced below the ice shelf by tidal oscillations. Consequent melting at the base of the ice shelf produces a Shallow Ice Shelf Water, where suitable conditions are shown to exist for the production of new frazil ice crystals.

IP16xx

Morphology and structure of the sea ice in the Southern Ocean sector between
60ƒE and 150ƒE

Ian Allison, Anthony P. Worby, Robert A. Massom,
Victoria I. Lytle and Petra Heil

Antarctic CRC and Australian Antarctic Div. Hobart, AUSTRALIA
Antarctic CRC, University of Tasmania, Hobart, AUSTRALIA
Antarctic CRC and Institute of Antarctic and Southern Ocean Studies, Hobart, AUSTRALIA

The seasonal sea ice cover in the eastern Indian and western Pacific Oceans is less extensive in all seasons than for other parts of the Antarctic. This region is also one of generally high ice drift rates (with daily averages of more than 0.2 m s^-1) both in the westward flowing coastal current, and in the eastward flowing Antarctic Circumpolar Current . Both the dynamic nature of the pack, and the largely divergent motion over much of the area, play important roles in determining ice characteristics. In this paper we summarise ship-board observations made over the last ten years between 60ƒE and 150ƒE, and between August and April, to describe the morphology and structure of the pack, and seasonal variations.

Characteristics considered here include ice concentration, floe size, and the thickness distribution of both the ice, and of the snow cover. In late September, at the time of maximum ice extent and maximum ice volume, the ice thickness distribution has a strong modal peak at about 0.6 m, and the snow cover thickness distribution has a peak at around 0.15 m. Un-ridged ice greater than 1 m-thick forms a relatively small areal fraction of the total pack, but ridges, typically with sail heights of not more than a metre, contain as much as 50% of the total ice mass. Ice of 0.4 to 0.6 m thickness forms the basic building blocks from which thicker ice is formed by rafting and ridging. Seasonal changes in the ice thickness distribution are primarily related to the thermodynamic growth rate of new ice and, in late spring and summer, to the melt of thin ice categories.

The textural structure, bulk salinity, and stable isotope ratio of the ice provide evidence of the formation processes and show the importance of ice growth in divergent leads, and of deformational thickening. Flooding and freezing of surface snow resulting from over burden load is also a significant ice formation process, and 13% of the total ice mass is formed by this method. Floe sizes continually change due to aggregation, and to swell induced break-up, even 300 km or more from the ice edge.

IP16yy

Sensitivity of an Arctic sea ice model to variations in the North Atlantic temperature and salinity flux

D. Stark, S. Piacsek and P. Posey

Naval Research Laboratory, Stennis Space Center, USA

The North Atlantic oceanic heat flux is known to significantly impact the amount of ice in the marginal ice zone of the Barents-Greenland seas. Understanding the sensitivity of present ice models to variations in the temperature and salinity forcing from the North Atlantic is helpful in estimating the Arctic impact of lower latitude global change.

The coupled ice-ocean system in the Arctic-Gin Sea Basin has been simulated using the Navy's Polar Ice Prediction System (PIPS). This code is based on the dynamic-thermodynamic Hibler ice model coupled to an ocean model including a Mellor-Yamada mixed layer model. Atmospheric forcing is provided by NMC reanalysis effort and the US Navy's global forecast products. Comparisons are made between a control run with standard climatological fluxes and simulations where the temperature and salinity flux are varied independently in magnitude and temporal frequency ranging from artificial process models to values derived from observations. The results are analyzed in terms of seasonally and spatially varying fields of ice thickness, compactness, velocity, and surface temperature.

IP16zz

The Relationship Between Internal Stress and Ice Thickness Within Boundary Regions

Jinro Ukita

NASDA-EORC, Tokyo, Japan

Through numerical studies it has been speculated that the large scale spatial pattern of ice thickness depends on the way in which internal stress acts in the momentum balance. Despite its importance to ice dynamics, very little is known as to how the momentum balance is modified through internal ice stress, which is physically the representation of floe-floe interactions. Recently the author has developed an analytical framework that is suitable to study this stress effect. Essentially it is a decomposition scheme for the divergence of a stress tensor. It provides us with not only a canonical representation under the plastic formulation, but also the physical interpretation of momentum fluxes arising from ice interactions. Within this context, we shall discuss the relationship between internal stress and ice thickness for coastal regions, with a particular emphasis on its geometrical aspects. Topics include (i) the roles of normal and shear stress, (ii) the separation of flow and material property effects, and (iii) the implication of plastic formulation. Also included is a discussion of the response of ice thickness to changes in atmospheric conditions.

IP16aaa

A comparison of observed sea ice parameters with modelled results in the region 60ƒ-150ƒE

Anthony P. Worby and Xingren Wu

Antarctic CRC and Australian Antarctic Division, Hobart, AUSTRALIA

The seasonal cycle of the ice thickness distribution for the East Antarctic pack ice between 60ƒand 150ƒE, derived from ten years of ship-based observations, is compared with the results of a dynamic-thermodynamic sea ice model coupled to a general atmospheric circulation model. The observed values are of the undeformed ice thickness but have been corrected using a simple model to include the effects of observed surface ridging. There are seven months with sufficient observational data to draw statistically meaningful conclusions. These are March-April and August-December.

The comparison shows the observed ice thickness values to be lower in all months than the modelled values. This may be due to inaccuracies in the prescribed ocean heat flux or snow conductivity in the model, and subsequent model runs will include sensitivity studies of these parameters.

Of more concern is the difference in the observed and modelled seasonal cycles of ice thickness. The observations show a decrease in the thickness between August and October, followed by a sharp increase in November and December. This is due to increasing radiation and air temperatures which, up until October, still supports ice formation in leads but without achieving the same thickness as observed in winter. The subsequent increase in ice thickness during November and December is due to the melting of thin ice. In contrast, the modelled result shows a much flatter distribution over the same period, but with a peak thickness in October.

The single mean thickness calculated by the model is most likely the cause of this disparity, as it will not reflect the significant changes in ice concentration, caused by the gradual removal of the thinnest ice types, as spring progresses. This is verified in the ice concentration values, which in the observational data decrease from 93% to 43% for the period August-December, but in the model decrease from only 84% to 60%.

The model also partitions ice growth in leads to the base and sides of floes depending on the mean ice thickness and concentration. This partioning is compared with the observed granular and columnar ice fractions determined from ice cores, which are representative of ice growth in leads and at the base of floes respectively. Additionally, the formation of snow ice by flooding and refreezing the base of the snow cover can be partitioned for comparison with observations.