Effective sea ice area based on a thickness threshold

As Arctic sea ice declines in response to climate change, a shift from thick multiyear ice to a thinner ice cover is occurring. With this transition, ice thicknesses approach a threshold below which ice no longer insulates the atmosphere from oceanic surface fluxes. While this is well known, there are no estimates of the magnitude of this threshold, nor of the proportion of sea ice area that is below this threshold as ice thins. We determine this threshold by simulating the atmospheric response to varying thicknesses, ranging from 0.0 to 2.0 m and determine that threshold to be 0.40–0.50 m. The resulting “effective” ice area is 4–14% lower than reported total ice area, as 0.39–0.97 × 10⁶ km² of the total ice area falls below the threshold throughout the twentieth century, including during notable ice minima. The atmosphere above large non-insulating ice-covered regions is susceptible to more than 2 °C of warming despite ice presence. Observed mean Arctic Ocean ice thickness is projected to fall below this threshold as early as the mid-2020s. Studies on ocean–atmosphere interactions in relation to sea ice area should focus on this insulating sea ice area, where ice is at least 0.40–0.50 m thick, and treat ice regions below 0.40–0.50 m thickness with caution.

     

Coupled ice‐ocean variability in the Greenland Sea

Abstract

A major surface feature of the Greenland Sea during winter is the frequent eastward extension of sea ice south of 75°N and an associated embayment to the north. These features are nominally connected with the East Greenland Current, and both the promontory and the embayment are readily apparent on climatic ice charts. However, there are significant changes in these features on time‐scales as short as a few days. Using a combination of satellite microwave images (SSM/I) of ice cover, meteorological data and in situ velocity, temperature and salinity records, we relate the ice distribution and its changes to the developing structure and circulation of the upper ocean during winter 1988–1989. Our measurements illustrate the preconditioning that leads to convective overturn, which in turn brings warmer water to the surface and results in the rapid disappearance of ice. In particular, the surface was cooled to the freezing point by early December and the salinity then increased through ice formation (about 0.016 m d^‐1) and brine rejection. Once the vertical density gradient was sufficiently eroded, a period of high heat flux (>300 W m^‐2) in late January provided enough buoyancy loss to convectively mix the upper water column to at least 200 m. We estimate vertical velocities at about 3 cm s^‐1 downward during the initial sinking. The deepening of the thermocline raised surface temperatures by over 1°C resulting in nearly 1.5 × 10⁵ km² of ice‐melt within two days. Average rates of ice retreat are about 11 km d^‐1 southwestward, generally consistent with a wind‐driven flow. Comparison of hydrographic surveys from before and after the overturning indicate the fresh water was advected out of the area, possibly to the south and east of our moorings.     

Upward flushing of sea water through first year ice

Abstract

Observations in both the ice and slush layers suggest that sea water intrudes into the snow layer following a snow storm. Ice temperature values recorded at 1 cm below the snow‐ice interface show that the upward flux of sea water is of short duration. This is followed by a period of intense brine drainage characterized by the migration of a salty brine layer, with salinities up to 42 psu. These results suggest that a snow storm induces a complete (upward) flushing of the brine channel network and major modifications of snow and ice characteristics.

Melt rates and downward brine fluxes were calculated using salinity measured in a 40 cm deep box placed on the ice‐water interface, which isolated a 50 × 50 cm area of sea ice from ocean mixing processes. In this semi‐isolated environment, observed salinity changes allowed us to determine melt water fluxes and brine drainage or flushing even though ice thickness measurements did not show any significant change. Melt rates up to 21 cm/month and equivalent growth rates up to 32 cm/month were measured.     

Consistent past half-century trends in the atmosphere, the sea ice and the ocean at high southern latitudes

Simulations performed with the climate model LOVECLIM, aided with a simple data assimilation technique that forces a close matching of simulated and observed surface temperature variations, are able to reasonably reproduce the observed changes in the lower atmosphere, sea ice and ocean during the second half of the twentieth century. Although the simulated ice area slightly increases over the period 1980–2000, in agreement with observations, it decreases by 0.5 × 10⁶ km² between early 1960s and early 1980s. No direct and reliable sea ice observations are available to firmly confirm this simulated decrease, but it is consistent with the data used to constrain model evolution as well as with additional independent data in both the atmosphere and the ocean. The simulated reduction of the ice area between the early 1960s and early 1980s is similar to the one simulated over that period as a response to the increase in greenhouse gas concentrations in the atmosphere while the increase in ice area over the last decades of the twentieth century is likely due to changes in atmospheric circulation. However, the exact contribution of external forcing and internal variability in the recent changes cannot be precisely estimated from our results. Our simulations also reproduce the observed oceanic subsurface warming north of the continental shelf of the Ross Sea and the salinity decrease on the Ross Sea continental shelf. Parts of those changes are likely related to the response of the system to the external forcing. Modifications in the wind pattern, influencing the ice production/melting rates, also play a role in the simulated surface salinity decrease.     

Boundary-layer modification in wintertime cold-air outbreaks from the Arctic sea ice

During the field experiment ARKTIS 1993 ten cases of boundary-layer modification in wintertime cold-air outbreaks from the Arctic sea ice in the Spitsbergen region were observed by aircraft over a distance ranging from about 50 km over the ice to about 300 km over the water. The modification depends decisively on the initial conditions over the ice, the boundary conditions at the bottom and top of the boundary layer and on the conditions of the large-scale flow. The modification of the bulk boundary-layer characteristics in relation to these conditions is presented.Besides the air-sea temperature contrast, the most important role for the boundary-layer modification is played by the stability on top of the boundary layer and by the divergence of the large-scale flow. According to the high variability of these conditions the observed boundary-layer modifications were very variable ranging from 100 to 300 m thick boundary layers with air temperatures between -32 and -22 °C over the ice to thicknesses between 900 and 2200 m and air temperatures between -15 and -5 °C after 300 km fetch over the open water. In most cases the large-scale flow was anticyclonic and divergent over the ice and changed to cyclonic and convergent over the water and an ice-sea breeze was superimposed on it.The sensible and latent heat fluxes are the dominant terms in the surface energy budget over the open water and ranged between 200 and 700 W m^-2 whereas the net longwave radiation is the dominating term over the ice with the heat fluxes only about 10 W m^-2.     

The effect of snow on Antarctic sea ice simulations in a coupled atmosphere-sea ice model

 The effect of a snow cover on sea ice accretion and ablation is estimated based on the ‘zero-layer’ version sea ice model of Semtner, and is examined using a coupled atmosphere-sea ice model including feedbacks and ice dynamics effects. When snow is disregarded in the coupled model the averaged Antarctic sea ice becomes thicker. When only half of the snowfall predicted by the atmospheric model is allowed to land on the ice surface sea ice gets thicker in most of the Weddell and Ross Seas but thinner in East Antarctic in winter, with the average slightly thicker. When twice as much snowfall as predicted by the atmospheric model is assumed to land on the ice surface sea ice also gets much thicker due to the large increase of snow-ice formation. These results indicate the importance of the correct simulation of the snow cover over sea ice and snow-ice formation in the Antarctic. Our results also illustrate the complex feedback effects of the snow cover in global climate models. In this study we have also tested the use of a mean value of 0.16 Wm^-1 K^-1 instead of 0.31 for the thermal conductivity of snow in the coupled model, based on the most recent observations in the eastern Antarctic and Bellingshausen and Amundsen Seas, and have found that the sea ice distribution changes greatly, with the ice becoming much thinner by about 0.2 m in the Antarctic and about 0.4 m in the Arctic on average. This implies that the magnitude of the thermal conductivity of snow is of considerable importance for the simulation of the sea ice distribution. An appropriate value of the thermal conductivity of snow is as crucial as the depth of the snow layer and the snowfall rate in a sea ice model. The coupled climate models require accurate values of the effective thermal conductivity of snow from observations for validating the simulated sea ice distribution under the present climate conditions. Received: 20 November 1997/Accepted: 27 July 1998     

Using δ<Superscript>18</Superscript>O as a Tracer of the Formation of Water Masses in the Laptev Sea. Part 1. Quantification of Ice Formation and Melting

Data on salinity and δ¹⁸O from the NASA open-source database are used to estimate the Laptev Sea water mass transformation during ice formation and melting. The indicator of these processes is salinity variation. The estimates for the Laptev Sea show that the amount of meltwater can reach 40% for the sea water with salinity below 7 psu. In this case, sea water salinity reduction due to the meltwater inflow alone can be equal to 0.2-0.7 psu. In the sea water with salinity above 7 psu, ice formation prevails over ice melting. This process is the most strongly pronounced in the range of sea water salinity from 15 to 25 psu. In this salinity range, the average water removal for the ice formation makes up 9% (the maximum is 24%), and the average salinity growth is 0.5 psu (the maximum is 1.7 psu). The most transformed sea water masses during ice formation are located in the bottom layer of the shallow southern and southeastern parts of the Laptev Sea, where the sea depth is not more than 50 m.     

Sea ice and wind: Effects on primary productivity in the Barents Sea

Abstract

The Barents Sea is divided into a northern and a southern part by the Polar Front (at about 75–76° N) where Atlantic waters descend under Arctic waters. Near to and north of the Polar Front, the spring bloom of phytoplankton is triggered by the stability induced in the upper 20 m by the melting of ice. The pycnocline is too strong to be eroded by wind. Primary productivity after the bloom is therefore small and largely regenerative. Underneath the pycnocline there is a 3–5 m thick layer characterized by dense, slow‐growing algal populations. New productivity north of the Polar Front is no more than 40 g C m^?2 a^?1.

In permanently open waters south of the Polar Front, the spring bloom starts in early May. Rhythmic wind‐induced mixing related to the atmospheric low‐pressure belt reaches an average 40–60 m depth in the growth season, and secondary phytoplankton maxima may arise. As a result, new annual productivity is more than doubled, i.e. 90 g C m^?2 a^?1, relative to the same system without wind. Although productivity is highest south of the Polar Front, it is more concentrated north of it, in the sense that high new production is mainly related to a 20–50 km wide belt that sweeps the area following the ice edge northwards while the ice melts through the summer.     

Unprecedented low twentieth century winter sea ice extent in the Western Nordic Seas since A.D. 1200

We reconstructed decadal to centennial variability of maximum sea ice extent in the Western Nordic Seas for A.D. 1200–1997 using a combination of a regional tree-ring chronology from the timberline area in Fennoscandia and δ¹⁸O from the Lomonosovfonna ice core in Svalbard. The reconstruction successfully explained 59% of the variance in sea ice extent based on the calibration period 1864–1997. The significance of the reconstruction statistics (reduction of error, coefficient of efficiency) is computed for the first time against a realistic noise background. The twentieth century sustained the lowest sea ice extent values since A.D. 1200: low sea ice extent also occurred before (mid-seventeenth and mid-eighteenth centuries, early fifteenth and late thirteenth centuries), but these periods were in no case as persistent as in the twentieth century. Largest sea ice extent values occurred from the seventeenth to the nineteenth centuries, during the Little Ice Age (LIA), with relatively smaller sea ice-covered area during the sixteenth century. Moderate sea ice extent occurred during thirteenth–fifteenth centuries. Reconstructed sea ice extent variability is dominated by decadal oscillations, frequently associated with decadal components of the North Atlantic Oscillation/Arctic Oscillation (NAO/AO), and multi-decadal lower frequency oscillations operating at ~50–120 year. Sea ice extent and NAO showed a non-stationary relationship during the observational period. The present low sea ice extent is unique over the last 800 years, and results from a decline started in late-nineteenth century after the LIA.     

Influence of coupling on atmosphere, sea ice and ocean regional models in the Ross Sea sector, Antarctica

Air–sea ice–ocean interactions in the Ross Sea sector form dense waters that feed the global thermohaline circulation. In this paper, we develop the new limited-area ocean–sea ice–atmosphere coupled model TANGO to simulate the Ross Sea sector. TANGO is built up by coupling the atmospheric limited-area model MAR to a regional configuration of the ocean–sea ice model NEMO. A method is then developed to identify the mechanisms by which local coupling affects the simulations. TANGO is shown to simulate realistic sea ice properties and atmospheric surface temperatures. These skills are mostly related to the skills of the stand alone atmospheric and oceanic models used to build TANGO. Nonetheless, air temperatures over ocean and winter sea ice thickness are found to be slightly improved in coupled simulations as compared to standard stand alone ones. Local atmosphere ocean feedbacks over the open ocean are found to significantly influence ocean temperature and salinity. In a stand alone ocean configuration, the dry and cold air produces an ocean cooling through sensible and latent heat loss. In a coupled configuration, the atmosphere is in turn moistened and warmed by the ocean; sensible and latent heat loss is therefore reduced as compared to the stand alone simulations. The atmosphere is found to be less sensitive to local feedbacks than the ocean. Effects of local feedbacks are increased in the coastal area because of the presence of sea ice. It is suggested that slow heat conduction within sea ice could amplify the feedbacks. These local feedbacks result in less sea ice production in polynyas in coupled mode, with a subsequent reduction in deep water formation.     

The formation and maintenance of the North Water Polynya

Abstract

This paper investigates the formation and maintenance of the North Water Polynya, Baffin Bay in winter using a multi‐category sea‐ice model coupled with the Princeton ocean model. Monthly climatological atmospheric data from the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalysis provides the forcing. An objectively‐analysed climatology provides the initial ocean temperature and salinity. Wind stress drives the ice in a cyclonic gyre around northern Baffin Bay. Localized regions of thin ice form where wind drives ice away from coastlines or fast ice. The regions of thin ice are characterized by enhanced ice growth, exceeding 1.2 m mo^?1. In the regions of thin ice, surface ocean heat flux is also enhanced and is between 30–60 W m^?2. Surface heat flux is, in part, attributable to convective mixing and entrainment driven by ice growth. The surface heat flux reflects advection of the warm West Greenland Current. Heat and salt balances show that horizontal advective exchange counterbalances surface fluxes of heat and salt.     

Future surface mass balance of the Antarctic ice sheet and its influence on sea level change, simulated by a regional atmospheric climate model

A regional atmospheric climate model with multi-layer snow module (RACMO2) is forced at the lateral boundaries by global climate model (GCM) data to assess the future climate and surface mass balance (SMB) of the Antarctic ice sheet (AIS). Two different GCMs (ECHAM5 until 2100 and HadCM3 until 2200) and two different emission scenarios (A1B and E1) are used as forcing to capture a realistic range in future climate states. Simulated ice sheet averaged 2 m air temperature (T_2m) increases (1.8–3.0 K in 2100 and 2.4–5.3 K in 2200), simultaneously and with the same magnitude as GCM simulated T_2m. The SMB and its components increase in magnitude, as they are directly influenced by the temperature increase. Changes in atmospheric circulation around Antarctica play a minor role in future SMB changes. During the next two centuries, the projected increase in liquid water flux from rainfall and snowmelt, together 60–200 Gt year^?1, will mostly refreeze in the snow pack, so runoff remains small (10–40 Gt year^?1). Sublimation increases by 25–50 %, but remains an order of magnitude smaller than snowfall. The increase in snowfall mainly determines future changes in SMB on the AIS: 6–16 % in 2100 and 8–25 % in 2200. Without any ice dynamical response, this would result in an eustatic sea level drop of 20–43 mm in 2100 and 73–163 mm in 2200, compared to the twentieth century. Averaged over the AIS, a strong relation between $\Updelta$ SMB and $\Updelta\hbox{T}_{2{\rm m}}$ of 98 ± 5 Gt w.e. year^?1 K^?1 is found.     

Seasonal climate information preserved in West Antarctic ice core water isotopes: relationships to temperature, large-scale circulation, and sea ice

As part of the United States’ contribution to the International Trans-Antarctic Scientific Expedition (ITASE), a network of precisely dated and highly resolved ice cores was retrieved from West Antarctica. The ITASE dataset provides a unique record of spatial and temporal variations of stable water isotopes (δ¹⁸O and δD) across West Antarctica. We demonstrate that, after accounting for water vapor diffusion, seasonal information can be successfully extracted from the ITASE cores. We use meteorological reanalysis, weather station, and sea ice data to assess the role of temperature, sea ice, and the state of the large-scale atmospheric circulation in controlling seasonal average water isotope variations in West Antarctica. The strongest relationships for all variables are found in the cores on and west of the West Antarctic Ice Sheet Divide and during austral fall. During this season positive isotope anomalies in the westernmost ITASE cores are strongly related to a positive pressure anomaly over West Antarctica, low sea ice concentrations in the Ross and Amundsen Seas, and above normal temperatures. Analyses suggest that this seasonally distinct climate signal is due to the pronounced meridional oriented circulation and its linkage to enhanced sea ice variations in the adjacent Southern Ocean during fall, both of which also influence local to regional temperatures.     

Seasonal variations of oceanic conditions near the southeastern coast of Sakhalin Island

Seasonal variations of hydrological conditions in the area adjoining the southeastern coast of Sakhalin Island are described based on the analysis of monthly mean temperature and salinity obtained over standard oceanic sections Makarov-Cape Georgii and Cape Svobodny-the sea and from nine oceanic surveys. The Poronai River runoff that promotes the formation of a warm surface layer with low salinity largely influences the water area of Terpeniya Bay in the northern part of the area studied. In spring, these waters primarily spread southward along the coast; in summer, they flow southeastward, forming a weak vortex structure at 144° E. In the fall, major changes occur below 20 m, where waters of the cold intermediate layer are replaced by warmer waters (4–6°C) of low salinity connected with the Amur River runoff. The destruction of the CIL core near the shelf edge at depths of about 100 m resulting from the fall intensification of the East Sakhalin Current is pronounced in the southern, abyssal part of the region. The coastal area is covered by waters with salinity below 32‰ connected with the Amur River runoff. The volume of low-salinity waters coming through the Cape Svobodny-the sea section into the southern part of the Sea of Okhotsk is estimated at 3000 km³ taking into account instrumental measurements of flow rates.     

Application of a coupled ice‐ocean model to the Labrador Sea

Abstract

The steady, coupled ice‐ocean circulation model of Willmott and Mysak (1989) for a meridional channel is applied to the Labrador Sea for the winter season. The model consists of a thermodynamic reduced‐gravity ocean combined with a variable thickness ice cover that is in thermal equilibrium. Upon specifying the forcing fields of surface air temperature, wind stress and water temperature along the open southern boundary, the winter climatological ice‐edge position, ice thickness, ocean circulation and temperature fields are determined in the channel domain. The sensitivity of the results to the various model parameters is examined. In particular, the optimum heat exchange coefficients for the interfaces of air‐water, ice‐water and air‐ice are found.

The model ice‐edge position compares favourably with the 50% winter climatological ice concentration isoline obtained from an analysis of 32 years (1953–84) of sea‐ice concentration data. The simulations of the ocean temperature and ice thickness are also quite realistic according to the observed records available. The model is also applied to two specific winters (1981 and 1983) during which anomalous sea‐ice and weather conditions prevailed in the Labrador Sea.     

Sea ice in the Barents Sea: seasonal to interannual variability and climate feedbacks in a global coupled model

Sea ice variability in the Barents Sea and its impact on climate are analyzed using a 465-year control integration of a global coupled atmosphere–ocean–sea ice model. Sensitivity simulations are performed to investigate the response to an isolated sea ice anomaly in the Barents Sea. The interannual variability of sea ice volume in the Barents Sea is mainly determined by variations in sea ice import into Barents Sea from the Central Arctic. This import is primarily driven by the local wind field. Horizontal oceanic heat transport into the Barents Sea is of minor importance for interannual sea ice variations but is important on longer time scales. Events with strong positive sea ice anomalies in the Barents Sea are due to accumulation of sea ice by enhanced sea ice imports and related NAO-like pressure conditions in the years before the event. Sea ice volume and concentration stay above normal in the Barents Sea for about 2 years after an event. This strongly increases the albedo and reduces the ocean heat release to the atmosphere. Consequently, air temperature is much colder than usual in the Barents Sea and surrounding areas. Precipitation is decreased and sea level pressure in the Barents Sea is anomalously high. The large-scale atmospheric response is limited with the main impact being a reduced pressure over Scandinavia in the year after a large ice volume occurs in the Barents Sea. Furthermore, high sea ice volume in the Barents Sea leads to increased sea ice melting and hence reduced surface salinity. Generally, the climate response is smallest in summer and largest in winter and spring.     

A continuous simulation of global ice volume over the past 1 million years with 3-D ice-sheet models

Sea-level records show large glacial-interglacial changes over the past million years, which on these time scales are related to changes of ice volume on land. During the Pleistocene, sea-level changes induced by ice volume are largely caused by the waxing and waning of the large ice sheets in the Northern Hemisphere. However, the individual contributions of ice in the Northern and Southern Hemisphere are poorly constrained. In this study, for the first time a fully coupled system of four 3-D ice-sheet models is used, simulating glaciations on Eurasia, North America, Greenland and Antarctica. The ice-sheet models use a combination of the shallow ice and shelf approximations to determine sheet, shelf and sliding velocities. The framework consists of an inverse forward modelling approach to derive a self-consistent record of temperature and ice volume from deep-sea benthic δ¹⁸O data over the past 1 million years, a proxy for ice volume and temperature. It is shown that for both eustatic sea level and sea water δ¹⁸O changes, the Eurasian and North American ice sheets are responsible for the largest part of the variability. The combined contribution of the Antarctic and Greenland ice sheets is about 10 % for sea level and about 20 % for sea water δ¹⁸O during glacial maxima. However, changes in interglacials are mainly caused by melt of the Greenland and Antarctic ice sheets, with an average time lag of 4 kyr between melt and temperature. Furthermore, we have tested the separate response to changes in temperature and sea level for each ice sheet, indicating that ice volume can be significantly influenced by changes in eustatic sea level alone. Hence, showing the importance of a simultaneous simulation of all four ice sheets. This paper describes the first complete simulation of global ice-volume variations over the late Pleistocene with the possibility to model changes above and below present-day ice volume, constrained by observations of benthic δ¹⁸O proxy data.     

Physical properties and isotopic characteristics of landfast sea ice around the North Water (NOW) Polynya region

Abstract

An ice core sampling program was conducted during the North Water (NOW) Polynya Project 1998 Experiment in northern Baffin Bay during April‐May 1998. The physical properties of snow and sea ice as well as the microstructure and stable isotopic composition of first‐year landfast sea ice near the polynya were investigated. The thickness of sea ice at the sampling sites ranged between 147 and 194 cm with thinner snow cover during the period between mid‐April and late May. The ice was characterized as typical first‐year landfast sea ice, being composed of a thin granular ice layer at the top and an underlying columnar ice layer towards the bottom of the ice. The samples obtained at a site closer to the ice edge of the polynya contained a thin granular ice layer originating from frazil ice near the ice bottom. Formation of frazil ice was considered to be caused by turbulent processes induced by winds, waves and currents forced from the polynya and also mixing with water masses produced at the polynya.     

Sensitivity of open-water ice growth and ice concentration evolution in a coupled atmosphere-ocean-sea ice model

A coupled atmosphere-ocean-sea ice model is applied to investigate to what degree the area-thickness distribution of new ice formed in open water affects the ice and ocean properties. Two sensitivity experiments are performed which modify the horizontal-to-vertical aspect ratio of open-water ice growth. The resulting changes in the Arctic sea-ice concentration strongly affect the surface albedo, the ocean heat release to the atmosphere, and the sea-ice production. The changes are further amplified through a positive feedback mechanism among the Arctic sea ice, the Atlantic Meridional Overturning Circulation (AMOC), and the surface air temperature in the Arctic, as the Fram Strait sea ice import influences the freshwater budget in the North Atlantic Ocean. Anomalies in sea-ice transport lead to changes in sea surface properties of the North Atlantic and the strength of AMOC. For the Southern Ocean, the most pronounced change is a warming along the Antarctic Circumpolar Current (ACC), owing to the interhemispheric bipolar seasaw linked to AMOC weakening. Another insight of this study lies on the improvement of our climate model. The ocean component FESOM is a newly developed ocean-sea ice model with an unstructured mesh and multi-resolution. We find that the subpolar sea-ice boundary in the Northern Hemisphere can be improved by tuning the process of open-water ice growth, which strongly influences the sea ice concentration in the marginal ice zone, the North Atlantic circulation, salinity and Arctic sea ice volume. Since the distribution of new ice on open water relies on many uncertain parameters and the knowledge of the detailed processes is currently too crude, it is a challenge to implement the processes realistically into models. Based on our sensitivity experiments, we conclude a pronounced uncertainty related to open-water sea ice growth which could significantly affect the climate system sensitivity.     

The role of sea ice in climatic variability: Theories and evidence 1

Abstract

The climatic role of sea ice is assessed in a survey of the recent literature. Theoretical or model‐based results are compared with existing evidence of ice‐atmosphere interactions over scales ranging from the local and regional to the hemispheric and global.

The evidence shows that sea‐ice fluctuations are meteorologically important locally, primarily through associations with air temperature. On the regional and hemispheric scales, atmospheric and sea‐ice fluctuations are correlated according to both observational evidence and model experiments. While the causal links have not been evaluated quantitatively, there is evidence that the stronger signal occurs in the response of the ice to the atmosphere. On the longer time‐scales, model experiments and qualitative arguments suggest that sea ice may play a major role in the climatic change. However, the results of large‐scale coupled model simulations contain deficiencies and must be viewed with caution pending more realistic treatments of sea‐ice dynamics, leads, ice thickness variations, and the areally‐integraled effects of the small‐scale features of sea ice.