Heat Polynya Research Articles

A salt flux model for salinity change through ice production in the Greenland Sea, and its relationship to winter convection

A salt flux model which incorporates ice formation, advection and melt, as well as time‐dependent brine drainage from frazil‐pancake ice, has been developed for the central Greenland Sea in winter. Information from ice drifting buoys suggests that this region can be thought of as similar to a latent heat polynya, with the wind blowing newly formed sea‐ice constantly in a southerly or southeasterly direction such that it melts in a different area from that of its formation. The result is a salt refinement process generating a net positive salt flux over the ice formation region, which may play an important role in stimulating deep open‐ocean convection. By combining buoy data, ECMWF meteorological data and passive microwave ice concentration data the model is used to estimate the sea ice‐derived salt flux over the 1996–1997 winter. Salt fluxes of over 40 kg m−2 were obtained in the center of the region, which corresponds closely an area where a maximum depth of convection is seen in winter cruises. The model gives realistic results when compared with in situ ice and oceanographic data gathered in the region, and suggests that ice formation and drift play an important role in the preconditioning of surface waters for deep water formation events.

Wintertime oceanography of the Adélie Depression

Abstract The formation of high-salinity shelf waters beneath coastal polynyas from enhanced sea-ice production and brine rejection during wintertime is critical to the production of Antarctic Bottom Waters. We report on the first wintertime measurements of high-salinity shelf water formation over the Adelie Depression off the East Antarctic coast between 143° and 146°E, during the Mertz Glacier Polynya Experiment in July–August 1999. The general circulation and evolution of water masses in the Adelie Depression during winter are described, and in particular we quantify the rates of brine rejection (sea-ice growth), ocean heat flux, and latent heat flux, in the Mertz polynya central to the formation of high-salinity shelf waters. We find shelf waters above the minimum sill depth with sufficient density to become Adelie Land Bottom Water and suggest the highest production rates occur near Commonwealth Bay. Sea-ice growth and heat transfer rates are calculated from the divergence of heat and freshwater fields around a closed volume beneath the Mertz polynya. The sea-ice growth estimates and associated Monte Carlo errors are found to range from 4.8±1.7, 4.1±1.5, 8.4±1.7 cm d−1, respectively, over a 3-week period. The average growth rate is 5.8 cm d−1. Near the Mertz polynya the sensible heat from ocean transport is between 15±6 and 43±11 W m−2, with an average of 30 W m−2, and the latent heat budget is between 125±45 and 254±52 W m−2, with an average of 174 W m−2. We conclude that during winter the Mertz polynya is primarily a latent heat polynya.

Modelling the impacts of persistent Antarctic polynyas with an atmosphere–sea-ice general circulation model

Abstract A coupled atmosphere–sea-ice model is used to assess the effect of Antarctic polynyas on the heat exchange between the surface and the atmosphere, the ice production rate within the polynyas, and the possible mechanisms of formation and maintenance of the polynyas. Results are presented for sensitivity studies with the model using prescribed minimum open-water fractions (25% and 30%) at a number of locations around the coast of East Antarctica and in the Amundsen, Ross and Weddell Seas. Although to some extent unphysical for the ocean, this technique allows the heat fluxes and freezing rates to be computed for the constrained conditions compared with an unconstrained control simulation. The model simulations suggest that Antarctic coastal polynyas, including the Mertz Glacier polynya, are primarily latent heat polynyas (where the heat loss from open water is balanced by the latent heat released with ice production) rather than sensible heat ones (where the heat loss is balanced by heat from the deep ocean). When a polynya exists, the turbulent latent heat flux from the ocean surface to the atmosphere can increase several times in magnitude and the average latent heat flux (over open-water and sea-ice fractions) can be sustained at over 60 W m −2 . The sensible heat flux changes sign and increases by about 2 orders of magnitude. This loss of turbulent heat (sensible plus latent) is offset only slightly by increased absorbed short-wave radiation in winter, but this offset increases greatly in spring when more short-wave radiation is available. In the polynyas, the large heat loss to the atmosphere is mostly balanced by the heat released due to ice formation, and this is accompanied by a high rate of salt rejection. The average ice production simulated in the polynyas is more than 11 m/year, with 5 m formed during 3 months of winter. These results provide quantitative support for the sustained high ice-freezing rates in polynyas estimated in other studies, which have used satellite and in situ observational data.

One-Dimensional Models for the Closure of a Coastal Latent Heat Polynya

Abstract One-dimensional models for the closure of a coastal latent heat polynya due to the onshore drift of frazil ice and consolidated ice are presented in this paper. The models predict the width of a polynya during an opening and closing cycle and include the cases in which the collection thickness of consolidated new ice at the polynya edge (during opening) or at the coast (during closing) is (a) a prescribed constant or (b) parameterized in terms of the depth of frazil ice arriving at the polynya edge (during opening) or the coast (during closing) and the relative velocity of the frazil ice with respect to the “convergence boundary” (i.e., the consolidated ice during opening or the stationary coastal wall during closing). Four dimensionless parameters specify the opening and closing cycle of a polynya, and throughout most of the four-dimensional parameter space the closing timescale is shorter than the opening timescale. In this case, it is found that the heat released from the polynya to the atmosp...

Numerical modelling of the Storfjorden (Svalbard) polynya development due to wind stress: role of the sea ice rheology and damping forces

Remote sensing of the ice cover in Storfjorden (Svalbard) revealed the persistence and evolution of latent heat polynyas during the winter of 1997/98. Latent heat polynyas open mechanically under wind stress or ocean currents that transport the ice cover away. In the present work we used mathematical modelling to simulate the Storfjorden polynya size and geometry caused by wind stress, measured at the meteorological station on the island of Hopen in winter 1997/98. The dependence of the polynya outlines on the wind velocity is presented. Two approaches were used: quasi-static and dynamic. Quasi-static simulations are based on a time-independent, linear ice stress-strain relationship valid for the low strain rates only. Time dependence of the ice cover fracture is joined with stress-strain nonlinearity caused by ice delayed-elastic recovery and viscosity. Results are compared to satellite observations from the synthetic aperture radar (SAR) of ERS-2. The simulation results show that a northern wind opens a larger polynya (ca. 30%) than does a north-eastern wind with the same speed. The results also indicate that the bathymetry and geometry of the fjord might have a stronger influence on the polynya opening and development than the location of individual islands and reefs.

Storfjorden (Svalbard): Modeling of the Polynya Development and the Sea Ice Ridging Process

Remote sensing observations of Storfjorden ice cover described the persistence and evolution of latent heat polynyas during winter 1997/1998. The induced important ice production was quantitatively estimated by simple modeling of the sea ice dynamics and growth. In the present work we used mathematical modeling to qualitatively simulate the polynya open and developing. First, we estimate the values of internal stresses appeared in Storfjorden ice sheet due to the wind stress and used them in the model calibration algorithm. Then, we model the polynya open and developing in simple quasi-static approach. The sea ice ridging process is modeled last and the maximum of ice ridges height is estimated. The results of the simulations are compared to satellite observations from ERS-2. They show that the equilibrium stages of the model developing coincide in common with in-situ data of satellite observations. From another side, the estimations show that the ice ridge keel can achieve height up to 8 m along the eastern shore of Spitsbergen.

Ice production and brine formation in Storfjorden, Svalbard

Brines, which appear as salty and cold bottom water layers on Arctic shelves, form by salt rejection during sea ice formation. High ice production occurs in latent heat polynyas such as the one that appears in Storfjorden under northerly winds. Using ERS‐2 synthetic aperture radar imagery, we observed the ice cover in Storfjorden over the winter 1997/1998, revealing an area of over 6000 km2 of the open water and thin ice that are characteristic of a polynya. Changes in the polynya extent correlated well with a simple wind‐driven polynya size algorithm. Furthermore, using a frazil ice formation algorithm and the hypothesis of ice accumulation at the lee side of the polynya, we distinguished between open water, thin ice, and fast/pack ice. The average polynya width during the 6 winter months was about 30 km, with a maximum of 130 km in March; thus Storfjorden was composed of 5/6 fast and pack ice and 1/6 polynya. Open water occupied about half of the polynya; the other half was composed of brash ice and thin ice involved in ridging and rafting. A total ice volume of 40 km3 was produced in Storfjorden between November 15, 1997, and May 15, 1998. About 57% of it was formed in the open water area; 9%, in the thin ice area; and 34%, in the fast/pack ice area. This ice production released more than 1 Gt (gigaton) of salt, able to increase the salinity of the Storfjorden waters by 1.4 practical salinity units (psu). Assuming that brine‐enriched waters are 0.3 psu more saline than the parent waters, the above salt release is able to ventilate Storfjorden 4–5 times during the course of the productive season.

Polynyas in a high-resolution dynamic–thermodynamic sea ice model and their parameterization using flux models

This paper presents an analysis of the solutions for a steady state latent heat polynya generatedby an applied wind stress acting over a semi-enclosed channel using: (a) a dynamic–thermodynamicsea ice model, and (b) a steady state flux model. We examine what processes in the seaice model are responsible for the maintenance of the polynya and how sensitive the results areto the choice of rheological parameters. We find that when the ice is driven onshore by anapplied wind stress, a consolidated ice pack forms downwind of a zone of strong convergencein the ice velocities. The build-up of internal stresses within the consolidated ice pack becomesa crucial factor in the formation of this zone and results in a distinct polynya edge. Furthermore, within the ice pack the across-channel ice velocity varies with the across-channel distance. It isdemonstrated that provided this velocity is well represented, the steady state polynya flux modelsolutions are in close agreement with those of the sea ice model. Experiments with the sea icemodel also show that the polynya shape and area are insensitive to (a) the sea ice rheology;(b) the imposition of either free-slip or no-slip boundary conditions. These findings are usedin the development of a simplified model of the consolidated ice pack dynamics, the output ofwhich is then compared with the sea ice model results. Finally, we discuss the relevance of thisstudy for the modelling of the North Water Polynya in northern Baffin Bay.

Development of an Antarctic Regional Climate System Model. Part I: Sea Ice and Large-Scale Circulation

Abstract A coupled atmosphere–ice regional model previously used for simulations in the Arctic has been implemented in the Antarctic. Three 14-month simulations were performed for 1988–89, with different oceanic specifications. The year 1988 was interesting as there were several sensible heat polynya events in the Cosmonaut Sea region, the investigation of which is the goal of future finer-resolution simulations. Overall, the regional climate model produces reasonable simulations of the Antarctic at 100-km resolution. Root-mean-square errors range from 4 hPa in surface pressure, 4 K in near-surface air temperature, and 3 m s−1 in near-surface winds, to 1 K in air temperature and 2 m s−1 in winds at the 500-hPa level. Tests of the coupled system response to oceanic heat flux suggest that the sea-ice simulation is more sensitive than the atmospheric circulation, but it could be expected that the atmosphere would respond to these changes in sea ice over longer time periods than those of interest here. This s...

The Storfjorden polynya: ERS-2 SAR observations and overview

Persistent polynyas have been observed over several winters in Storfjorden, situated between Spitsbergen and Barentsøya/Edgeøya in the south of the Svalbard archipelago. Polynyas are in general active regions with respect to ocean-atmosphere heat exchange, presenting strong convection phenomena and as such being involved in important water mass formation and having an impact on the marine ecosystem. Hydrographic observations have revealed very dense (cold and saline) brine-enriched bottom waters leaving the continental shelf as gravity driven plumes into the deep sea west of Spitsbergen. Satellite observations, using ERS-2 SAR imagery, reveal the evolution of the Storfjorden polynya during winter 1997/98. After forming a complete ice cover until mid-January, Storfjorden responds dynamically to northerly winds by opening a large latent heat polynya. It occupies at its largest extent a region of up to 6000 km2 of open water, thin ice and brash ice. Comparable in size to other large Arctic polynyas, the Storfjorden polynya might have the same or even greater importance in the thermohaline circulation and bottom water mass formation. Ice production is estimated at 30 km3 in Storfjorden, rejecting around 700 Mt (Megatons) of salt that can raise the salinity in Storfjorden by 0.9-1.0 PSU. First studies and the winter 1997/98 evolution of this polynya are presented in this paper.

Shallow, brine‐driven free convection in polar oceans: Nonhydrostatic numerical process studies

A three‐dimensional nonhydrostatic convection model, which accounts for small‐scale ice‐ocean interactions, is used to study convection in shallow sea (coastal) ice formation regions which contribute significantly to water mass formation in both the Arctic and Antarctic Ocean. For certain conditions the results presented in this paper are also transferable to shallow open ocean convection. The model is applied to an initial well‐mixed ocean at rest with a temperature close to the freezing point. The ocean, initially free of ice, is exposed to cold and dry polar air. We consider situations in which the mean wind stress is negligible but wind fluctuations result in (small) sensible and latent heat fluxes corresponding to a wind speed of 2 m s−1. Cellular convection patterns develop in the ocean, finally occupying a mean aspect ratio of 2. Convection is driven by salt release during frazil ice formation due to supercooling. Newly forming sea ice is collected along convergent (downwelling) regions at the surface, thus showing also cellular structures. Because the area of insulating sea ice remains small, new ice can be formed continuously, and the surface buoyancy forcing remains large. This collection of ice in small fractions of the sea surface results in a latent heat polynya type, which is very effective in terms of dense water mass formation. A comparison of the three‐dimensional model and a two‐dimensional (slice) model shows that key results can be reproduced with the slice model. In summary, the results of the process studies indicate that cellular features in the sea ice cover, which may be detectable by remote sensing techniques, are closely related to active brine‐driven convection.

On the influence of coastline orientation on the steady state width of a latent heat polynya

This paper investigates under what conditions the steady state shape of a coastal latent heat polynya is influenced by (1) the direction of flow of free‐drifting frazil ice and the consolidated offshore ice sheet and (2) coastline orientation. Analytical solutions are obtained for coastal polynyas when the coastline is straight and when a coastline has a single step. From the solutions it is shown that the asymptotic polynya width is given by the generalized Pease (1987) width, L cos θ, where L is the “classical Pease width” and θ is the angle between the consolidated ice sheet velocity and the normal to the coastline. An expression is also obtained for the alongshore length scale LC, over which the polynya adjusts toward its asymptotic width, and it is shown that LC depends upon the directions of travel of the frazil ice and the consolidated ice sheet. When the frazil ice and the consolidated ice sheet both drift offshore, normal to the coast, LC = 0, in agreement with the classical Pease theory. Variations in the coastline which occur on a length scale much smaller than LC are not reproduced in the polynya shape. Two numerical simulations are presented for Terra Nova Bay, Antarctica, and the St. Lawrence Island Polynya in the Bering Sea. Both examples employ realistic coastline geometry and information about the direction of drift of the consolidated ice sheet. In each example a realistic polynya shape is obtained. The examples highlight the versatility of the model because in the former case the ice sheet drifts almost parallel to the coast, while in the latter it drifts offshore.

The 1992 summer circulation in the Northeast Water Polynya from acoustic Doppler current profiler measurements

A July–August 1992 snapshot of the circulation in the Northeast Water Polynya located off the coast of Greenland is presented using data collected with an acoustic Doppler current profiler (ADCP) during the cruise of the USCGC Polar Sea. The data reveal anticyclonic circulation with velocity and transport cores aligned with the local topographic trough. Transport through the trough is between 0.5 and 0.9 Sv. The total ADCP measured velocities are generally less than 10 cm s−1, but some regions have velocities larger than 20 cm s−1. The ratio of the average velocity of the upper 100 m to the average velocity below 100 m shows surface layer flows faster by a factor of between 1.2 and 2. This ratio in the ice‐free northwest portion of the polynya is smaller (≈1.5) than found over the adjacent continental shelf (≈2.0), indicating smaller vertical shear. The data show northward flow along the coast, supporting the idea that ice carried by this flow is blocked from entering the polynya by a persistent ice wedge to the south that extends from the coast, east over the topographic trough, to a shallow central bank. By blocking ice advecting from the south, the ice wedge enhances the polynya, supporting the idea that the Northeast Water Polynya is a latent heat polynya.

A Nonlinear Steady-State Model of the North Water Polynya, Baffin Bay

Abstract A nonlinear, steady-state model of the North Water (NOW), the Arctic's largest polynya, is presented. The model follows in the spirit of the recently developed latent and sensible heat polynya model of Mysak and Huang, but extends it in several important ways: finite amplitude displacements of the upper-layer thickness are allowed; the channel walls diverge to the south; the sensible-heat flux from the lower layer is physically well defined in terms of a vertical entrainment velocity; and the free-drifting frazil ice to the north of the NOW ice edge is allowed to move to the right of the northerly winds. An important result found here is that with the exception of late spring, the asymptotic southern ice edge position of the NOW can be simulated in terms of a latent heat model alone. In this case, the observed equatorward curvature of the ice edge in the region adjacent to the west Greenland coast can be produced by a combination of a channel that widens in the equatorward direction, together wit...

Recurring polynyas over the Cosmonaut Sea and the Maud Rise

Two polynyas over the deep ocean were observed in the Antarctic region during the winter of 1980: one near 43°E, 66°S (Cosmonaut polynya) and another near 2°E, 64°S (Maud Rise polynya). The time history of these two polynyas was examined on an alternate day basis using ice concentration maps from the Nimbus 7 scanning multichannel microwave radiometer (SMMR). A quantitative analysis of a study area around it shows that the totally enclosed Cosmonaut polynya attained a maximum size on July 25, 1980, with an open water area of as much as 137,700 km2. This polynya lasted for a few weeks, disappeared on August 16, 1980, and was not observed for the rest of the winter. Similar polynyas in the same region have occurred for several years, including 1973, 1975, 1979, 1982, and 1986. The Maud Rise polynya, on the other hand, was observed as a reduction in ice concentration to about 37% within the SMMR resolution of about 900 km2. However, the open water area in the region amounted to 92,800 km2 on July 20, and the polynya recurred several times during the same winter period. It is proposed that both polynyas are products of deep‐reaching convection which introduces warmer deep water into the surface layer. In this way, they are viewed as sensible heat polynyas in that they are maintained by oceanic heat. The oceanographic settings of these two polynyas are similar. The hydrographic data at both sites indicate the existence of localized doming of the pycnocline. This brings warmer, saltier deep water closer to the sea surface, which has been demonstrated to be an effective preconditioner for deep‐reaching convection. It is probable that the polynyas are terminated by “invasion” of sea ice from the sides, which attenuates convection. The capability of a polynya to survive an entire winter period may be related to its size; the larger polynyas are better protected from convergence of the surrounding sea ice. For example, a simple model shows that a polynya with a diameter of about 100 km or greater is much more likely to survive a winter season than a smaller one.