The two-dimensional wavelet transform is a highly efficient band-pass filter, which can be used to track features in satellite images from sequential paths. Wavelet analysis of NASA scatterometer and Special Sensor Microwave/Imager data has been used to obtain daily sea-ice drift information for the Arctic region. Comparison with ice motion derived from ocean buoys shows good quantitive agreement. Furthermore, the scatterometer results definitely complement passive-microwave radiometer results when there are cloud or surface effects. This outcome allows three sets of sea-ice-drift daily results from scatterometer, radiometer and buoy data to be merged as a composite map by data-fusion techniques. Based on the composite maps, the ice-flow streamlines are highly correlated with surface air-pressure contours. In order to quantify the wind effects on ice motion, empirical orthogonal functions are used in the principal-component analysis to isolate generalized patterns inherent in 6 months (fall/winter) of daily sea-ice motion data. It is found that 30% of sea-ice motion is highly correlated with 50% of the pressure field in modes 1 and 2. For the higher modes, sea-ice motion is also affected by ocean current, bathymetry and coastal boundary and therefore is not highly correlated with the wind field.

Results of East Antarctic (20−160° E) sea-ice motion derived from drifting buoys (collected in eight selected years between 1985 and 1997) are compared with sea-ice motion derived from sequential images from the 37 and 85.5 GHz channels of the Scanning Multichannel Microwave Radiometer (SMMR, 1979−87) and the Special Sensor Microwave/Imager (SSM/I, 1987−97). For the latitudes of interest, buoys yield velocity fields at 2 hourly resolution/while SMMR and SSM/I provide daily data. The spatial coverage of the drifting buoys is relatively sparse compared to the complete coverage of the southern sea-ice zone by SMMR and SSM/I, but the accuracy of the buoy position is better than ±280 m, while the resolution for the 85.5 GHz channel is only about 12.5 km, and about 25 km for the 37 GHz channel. The buoy measurements represent point motion, equal to the velocity of an individual ice floe, while the satellite data derived via the maximum cross-correlation method represent the motion of surface patterns within a gridcell. Comparison of the mean distributions of daily velocity derived from the datasets shows that while the broad-scale velocity patterns agree, the magnitude of the satellite-derived sea-ice velocities is significantly lower (typically by 40% or less) than the velocities derived from buoy measurements. This is in contrast with previous studies for the Arctic Ocean and the Weddell Sea, Antarctica. To elucidate differences in the velocity magnitude, individual trajectories from a number of drift regimes off East Antarctica are directly compared. Discrepancies between the two datasets are discussed with regard to the difference in spatial resolution of the data, the data-collection and −processing methods and the observed region.

Observed and modeled sea-ice motions, combined via an optimal-interpolation assimilation method, are used to study two synoptic events in the Arctic. The first is a convergence event along the north Alaska coast in the Beaufort Sea during November 1992. Assimilation indicates stronger convergence than the stand-alone model, in agreement with Advanced Very High Resolution Radiometer-derived ice motions and Special Sensor Microwave/Imager-derived ice concentrations. The second event pertains to ice formation and advection in Fram Strait and the Barents and Greenland, Iceland and Norwegian Seas. Assimilation indicates export of thick, less saline ice out of the central Arctic into the East Greenland Sea. However, the model indicates little flow through Fram Strait, instead showing strong flow of thin, more saline first-year ice from the Barents Sea westward into the Greenland Sea. These results indicate that assimilation is a useful tool for investigating synoptic events in the Arctic and may be useful for both climate studies and operational analyses

A sea-ice dynamics experiment was performed in the Baltic Sea in March 1997. A new type of drifter was constructed based on the global positioning system and mobile-phone data transmission. The drifters worked well and the accuracy was reasonable (std dev. 40 m). Five drifters were used to map ice kinematics in a 20 km size array in the coastal drift-ice zone. The study included periods of onshore motion with ridge formation, and alongshore motion with narrow shear lines. The motion of ice largely occurred in short pulses between which the field stood nearly still. The level of ice speed was much less than that of free drift, while the deformation field was more uniaxial due to the presence of the solid boundary. Drift-ice strength was estimated as 4 × 104 N m−2. Large deformation rates of up to 2% h−1 were observed.

While sea-ice drift tends to be a relatively coherent field moving largely in response to spatially averaged wind, strain-rate observations show relative ice motion to be much more of a chaotic field with deformation occurring along narrow intersection strikes between effectively rigid plates. This relative motion also has considerable energy at high frequencies which are not coherent with the wind forcing. Plastic rheologies for a strongly interactive medium are inherently able to predict such fractures. To investigate this phenomenon, this paper presents a basic anisotropic theory for multiple-oriented weaknesses. When coupled with short-term evolution equations for ice strength, this model is found to yield strong zones of failure over the Arctic basin, with relatively rigid motion in between. These failure lines typically occur along geostrophic wind contours. These zones are also found to be created with appropriate "Coulombic" isotropic rheologies. When combined with a previously developed inertial embedding procedure for the oceanic boundary layer, the resulting deformation field tends to fluctuate with considerable power near the inertial period even though the wind field is constant. The mechanisms necessary for obtaining such oriented features are discussed.

A sea-ice thickness profile obtained in September 1996 from the Greenland Sea and the Eurasian Basin, extending as far as the North Pole, has shown an unusually open ice cover with low mean drafts, large amounts of open water and little deep pressure ridging. Comparisons with data obtained from the same region in October 1976 show that mean ice draft has declined by 43% and that the decline can largely be ascribed to a loss of the thickest ice.

The morphology of the Arctic sea-ice cover undergoes large changes over an annual cycle. These changes have a significant impact on the heat budget of the ice cover, primarily by affecting the distribution of the solar radiation absorbed in the ice-ocean system. In spring, the ice is snow-covered and ridges are the prominent features. The pack consists of large angular floes, with a small amount of open water contained primarily in linear leads. By the end of summer the ice cover has undergone a major transformation. The snow cover is gone, many of the ridges have been reduced to hummocks and the ice surface is mottled with melt ponds. One surface characteristic that changes little during the summer is the appearance of the bare ice, which remains white despite significant melting. The large floes have broken into a mosaic of smaller, rounded floes surrounded by a lace of open water. Interestingly, this break-up occurs during summer when the dynamic forcing and the internal ice stress are small During the Surface Heat Budget of the Arctic Ocean (SHEBA) field experiment we had an opportunity to observe the break-up process both on a small scale from the ice surface, and on a larger scale via aerial photographs. Floe break-up resulted in large part from thermal deterioration of the ice. The large floes of spring are riddled with cracks and leads that formed and froze during fall, winter and spring. These features melt open during summer, weakening the ice so that modest dynamic forcing can break apart the large floes into many fragments. Associated with this break-up is an increase in the number of floes, a decrease in the size of floes, an increase in floe perimeter and an increase in the area of open water.

Upward-looking sonar (ULS) data are presented from a prototype instrument deployed at 63° 18’ S, 107°49’ E in 1994. These data show the seasonal evolution of the ice-draft distribution from May when predominantly thin ice is present, through October when substantially thicker ice has been formed by deformation. The mean ice draft reaches a maximum in August at 1.21 m, the same month in which ship-based observations from the same region show a peak in ice thickness. The observed distribution from ULS data is only for drafts > 0.3 m due to data losses caused by the low acoustic reflectivity of actively forming ice. The spring distributions show very little development of drafts > 3.0 m, and it is hypothesized that this is due to the cyclical nature of deformation in the East Antarctic pack-ice zone, and that periods of sustained pressure required to form very thick ice are uncommon in this region

In order to consider the effect of different ice conditions on the momentum exchange between atmosphere and ocean, a parameterization for variable drag coefficients is introduced. A square dependence on ice concentration is combined with a linear dependence on deformation energy to account for the influence of floe edges as well as for roughness elements in regions with a more compact ice cover. The approach is not regionally specified and is applied over the whole model region. Empirical parameters are optimized via comparison with observed buoy-drift data.

Sea-ice thickness and roughness data collected on three cruises in the Ross Sea, Antarctica, showed interseasonal, regional and interannual variability. Variability was reduced to season, or age of ice floe, when sea-ice roughness values from around Antarctica were compared. There were statistically significant correlations between mean snow elevation and mean ice thickness; snow surface roughness and mean ice thickness; and snow surface roughness and ice bottom roughness, which appeared to be independent of season, geographical location and deformation history of ice floes. Our field data indicate that ice thickness can be predicted from snow elevation measurements with higher accuracy in summer. The feasibility of using snow surface roughness to infer ice thickness and ice bottom roughness is promising, and can provide us with a means to study the thickness and underside of Antarctic sea ice at good spatial and temporal resolution.

In the framework of the Surface Heat Budget of the Arctic (SHEBA) study, indirect, non-invasive ice mass-balance measurements were carried out at a drifting station in the northern Chukchi Sea between May and August 1998. Ice thickness was derived from electromagnetic induction (EM) measurements of apparent conductivity along 13 profiles (60−900 m long). As shown through sensitivity studies with a one-dimensional model, the apparent conductivity data from individual points can be inverted to yield estimates of ice thickness and ablation with an accuracy of approximately 0.05 m (for 2 m thick level ice). Ablation rates were 8−18 mm d−1, with total ablation amounting to roughly 0.9−1.2 m. Measurements of thickness and melt rates along different profiles in undeformed multi-year ice corresponded closely, indicating that the sampling statistics are adequate. The roughness of undeformed ice has been found to increase during the summer due to deepening of melt ponds and enhanced bottom melt. Ice under melt ponds was disproportionately thinner, most likely a result of thicker snow cover reducing winter accretion.

Winter ice dynamics plays an important role in the energy budget of the air-ice-ocean system, through the formation of leads and ridges. In summer, thermodynamic processes cause a transition in the ice pack from a mechanical continuum to an ensemble of floes that move in a state of free drift, with little floe-floe interaction. Results from the recent Surface Heat Budget of the Arctic Ocean (SHEBA) experiment have demonstrated that even under summer conditions, ice dynamics can still cause dramatic changes in the characteristics of the ice-ocean matrix that affect the energy budget. To illustrate this, we present observations taken before and after a period of sustained, moderate winds in late July 1998, which was preceded by an extended period of low winds. These conditions resulted in significant differential motion of ice floes in the vicinity of SHEBA. The measurements include the mass balance of the ice cover, the distribution of ice and open water, and salinity and temperature profiles in leads. The data show that after the storm there was a significant change in the amount and distribution of open-water areas, that there was an increase in the rate of bottom ablation, and that a stratified layer of warm fresh water that had formed at the top of leads during melt had become mixed.

Vertical profiles of temperature from the air through the snow and ice and into the upper ocean were measured over an annual cycle, from October 1997 to October 1998, as part of a study of the Surface Heat Budget of the Arctic Ocean (SHEBA). These observations were made at nine locations, including young ice, ponded ice, undeformed ice, a hummock, a consolidated ridge and a new blocky ridge. All of the sites had similar environmental forcing, with air temperatures at the different sites typically within 1°C. In general, the seasonal evolution of ice temperature followed a pattern of (1) a cold front propagating down through the ice in the fall, (2) cold ice temperatures and ice growth in late fall, winter and early spring, and (3) warming to the freezing point in the summer. Within this general pattern, there was considerable spatial variability in the temperature profiles, particularly during winter. For example, snow/ice interface temperatures varied by as much as 30°C between sites. The coldest ice temperatures were observed in a consolidated ridge with a thin snow cover, while the warmest were in ponded ice. The warm pond temperatures were a result of two factors: the initial cooling in the fall was retarded by freezing of pond water, and the depressed surface of the pond was quickly covered by a deep layer of snow (0.6 m). In an 8 m thick unconsolidated ridge, the cold front did not penetrate to the ice bottom during winter, and a portion of the interior remained below freezing during the summer. The spatial variability in snow depth and ice conditions can result in situations where there is significant horizontal transport of heat.

The temperature of the snow-ice interface was measured every 2.4 h throughout winter 1997/98 at 30 locations near the Surface Heat Budget of the Arctic Ocean (SHEBA) camp in the Beaufort Sea. Measurements were obtained from young ice, ridges, refrozen melt ponds and ice hummocks. Average snow depths at these locations were 567 cm, while mean interface temperatures ranged from −8° to −25°C, with minimums varying from −12° to −39°C. Interface temperatures were linearly related to snow depth, with increasing scatter at greater depths. The conductive heat flux during the winter, Fc, was estimated for each location using air and interface temperatures, snow depths and measured snow thermal conductivities. Fc was integrated to determine total heat loss for the winter at each site. Losses varied by a factor of four, with variations over short distances (10 m) as large as the variations between ice floes. Spot measurements along traverse lines confirm that large variations in interface temperature are common, and imply that small-scale spatial variability in the conductive flux is widespread. A comparison of the dependence of Fc on snow depth and ice thickness based on our observations with the dependence predicted by a one-dimensional theoretical model suggests that spatial heterogeneity may be an important issue to consider when estimating the heat flux over large aggregate areas. We suggest that the small-scale variability in the conductive flux arises because the combined snow and ice geometry can produce significant horizontal conduction of heat.

Leads act as windows through which solar radiation enters the upper ocean. Studies of the heat budget and ice melt require accurate parameterization of the albedo. Results from summertime measurements of albedo over leads show that under cloudy conditions the average albedo is 0.066 with a standard deviation of 0.007. The clear-sky case is more complex because the albedo is strongly dependent on wind speed at large solar zenith angles. This dependence has not been included in prior modeling efforts. We suggest a simple modification to an existing parameterization of the albedo" which includes the effect of wind speed.

The RADARSAT geophysical processor system (RGPS) uses sequential synthetic aperture radar images of Arctic sea ice taken every 3 days to track a large set of Lagrangian points over the winter and spring seasons. The points are the vertices of cells, which are initially square and 10 km on a side, and the changes in the area of these cells due to opening and closing of the ice are used to estimate the fractional area of a set of first-year ice categories. The thickness of each category is estimated by the RGPS from an empirical relationship between ice thickness and the freezing degree-days since the formation of the ice. With a parameterization of the albedo based on the ice thickness, the albedo may be estimated from the first-year ice distribution. We compute the albedo for the first spring processed by the RGPS, the early spring of 1997. The data include most of the Beaufort and Chukchi Seas. We find that the mean albedo is 0.79 with a standard deviation of 0.04, with lower albedo values near the edge of the perennial ice zone. The biggest source of error is likely the assumed rate of snow accumulation on new ice.

The heat budget over the ice-covered area of the southern Sea of Okhotsk is estimated from in situ meteorological and ice observation for four years, 1996−99. The data are from about 1 week in early February in each of four years. Ice-thickness distributions required for calculating the heat budget are quantitatively obtained from video analysis. A one-dimensional thermodynamical model is used to calculate the heat flux. The total heat flux is obtained by summing up the area-weighted heat flux of each ice-thickness category. In addition, to determine the characteristics of the heat budget in this region, we also calculated the heat budget in the northern Sea of Okhotsk using Special Sensor Microwave/Imager ice-extent data and European Centre for Medium-range Weather Forecasts meteorological data, and compared the results. Our investigations show the following characteristics in the southern Sea of Okhotsk: (1) Due to relatively thin ice thickness, the average turbulent heat flux is upward. (2) Thin ice and open water contribute significantly to the total turbulent heat flux. (3) Thermodynamic ice growth is limited to about 1 cm d−1 on average. (4) The heat budget is largely characterized by abundant solar radiation. The first, third and fourth results are characteristic of this region located at a relatively low latitude, while the second one is similar to that for polar regions.

Turbulent surface fluxes were studied using observations taken over sea ice in the Baltic Sea in March 1998. The fluxes of momentum and sensible heat were measured by a sonic anemometer and compared with fluxes derived from wind velocity and air-temperature profiles. The neutral 10 m drag coefficient showed no apparent dependence on wind speed (in the range 2–20 m s–1), resulting in a mean value of 1.0 × 10–3 for smooth snow-covered ice and 1.5 × 10−3 for deformed ice. The overall mean value was 1.28 × 10–3. The roughness length for temperature revealed a greater apparent dependence on wind speed and was slightly larger than the aerodynamic roughness for low wind speeds, and vice versa for moderate and high winds. We give an empirical expression that predicts how the scalar roughness depends on the aerodynamic roughness (drag coefficient) and wind speed. Agreement of the gradient-method results with the eddy-flux results supports the validity of the Monin-Obukhov similarity theory. Fluxes modelled by a coupled air-ice-sea model compared well with the eddy-flux and gradient methods. Surface temperature estimates by the three methods also agreed well. Tests and sensitivity analysis emphasize the need for especially accurate sensor calibration and strict information about the sensor heights for the gradient method.

In this paper, we present the results of a one-dimensional, thermodynamic sea-ice model applied to the Baltic Air-Sea-Ice Study (BASIS) field data. In general, the model results are in good agreement with the measurements, which were made during mild weather conditions with distinct areal and temporal variations in the snow and ice thickness. The total amount of refrozen ice calculated from the surface melting water gives a first-order estimate of snow-ice formation during the BASIS experiment, and this agreed well with the total observed variation in ice thickness. The model slightly overestimated ice growth at the bottom. This may be due to the variation in sea-ice thermal properties, affected by a slush layer between the snow and ice, or to the lack of ice-ocean interaction in the model.

The spatial and temporal variability of surface, cloud and radiative properties of sea ice are examined using new satellite-derived products. Downwelling short- and longwave fluxes exhibit temporal correlation over about 180 days, but cloud optical depth and cloud fraction show almost no correlation over time. The spatial variance of surface properties is shown to increase much less rapidly than that of cloud properties. The effect of small-scale inhomogeneity in surface and cloud properties on the calculation of radiative fluxes at ice- and climate-model gridscales is also investigated. Annual mean differences between gridcell fluxes computed from average surface and cloud properties and averages of pixel-by-pixel fluxes are 9.46% for the downwelling shortwave flux and −7.04% for the longwave flux. Therefore, using mean surface and cloud properties to compute surface radiative fluxes in a gridcell results in an overestimate of the shortwave flux and an underestimate of the longwave flux. Model sensitivity studies show that such biases may result in substantial errors in modeled ice thickness. Clearly, the sub-gridscale inhomogeneity of surface and atmospheric properties must be considered when estimating aggregate-area fluxes in sea-ice and climate models.