Understanding long-term changes in large-scale sea ice drift in the Southern Ocean is of considerable interest given its contribution to ice extent, to ice production in open waters, with associated dense water formation and heat flux to the atmosphere, and thus to the climate system. In this paper, we examine the trends and variability of this ice drift in a 34-year record derived from satellite observations. Uncertainties in drift (~3 to 4 km day -1 ) were assessed with higher resolution observations. In a linear model, drift speeds were ~1.4% of the geostrophic wind from reanalyzed sea-level pressure, nearly 50% higher than that of the Arctic. This result suggests an ice cover in the Southern Ocean that is thinner, weaker, and less compact. Geostrophic winds explained all but ~40% of the variance in ice drift. Three spatially distinct drift patterns were shown to be controlled by the location and depth of atmospheric lows centered over the Amundsen, Riiser-Larsen, and Davis seas. Positively correlated changes in sea-level pressures at the three centers (up to 0.64) suggest correlated changes in the wind-driven drift patterns. Seasonal trends in ice edge are linked to trends in meridional winds and also to on-ice/ off-ice trends in zonal winds, due to zonal asymmetry of the Antarctic ice cover. Sea ice area export at flux gates that parallel the 1000-m isobath were extended to cover the 34-year record. Interannual variability in ice export in the Ross and Weddell seas linked to the depth and location of the Amundsen Sea and Riiser-Larsen Sea lows to their east. Compared to shorter records, where there was a significant positive trend in Ross Sea ice area flux, the longer 34-year trends of outflow from both seas are now statistically insignificant.

We present a first examination of Arctic sea ice snow depth estimates from differencing satellite lidar (ICESat‐2) and radar (CryoSat‐2) freeboards. These estimates cover the period between 14 October 2018 and the end of April 2019. Snow depth is related to freeboard differences by the refractive index/bulk density of the snow layer—the only free parameter in the approach. Area‐averaged snow depth ranges from 9 cm (on first‐year ice: 5 cm, multiyear ice: 14 cm) in late October to 19 cm (first‐year ice: 17 cm, multiyear ice: 27 cm) in April; on average, this snow is thinner over FYI. Spatial patterns and gradients of snow depth estimates compare well with reconstructions using snowfall from ERA‐Interim and ERA5, although snowfall from ERA5 is systematically higher. For all months, the results suggest that ~50% of the total freeboard is comprised of snow. Retrievals are within a few centimeters of snow depth data acquired by Operation IceBridge in April 2019. Sources of uncertainties associated with this freeboard‐differencing approach are discussed. Further, sea ice thicknesses calculated using the retrieved snow depth and a modified climatology are contrasted. Comparatively, the snow depth and calculated ice thickness using a modified climatology are higher by ~5 cm and 0.33 m, although these differences are not uniform throughout the season. Snow accumulation was slower between October and December but increased between December and January, unlike the modified climatology, which exhibited a monotonic accumulation for all months. Future opportunities for assessment and improvement of these estimates are discussed.

Abstract. We examine the variability of sea ice freeboard, snow depth, and ice thickness in three years (2011, 2014, and 2016) of repeat surveys of an IceBridge (OIB) transect across the Weddell Sea. Averaged over this transect, ice thickness ranges from 2.40±1.07 (2011) to 2.60±1.15 m (2014) and snow depth from 35.8±11.5 (2016) to 43.6±10.2 cm (2014), suggesting a highly variable but broadly thicker ice cover compared to that inferred from drilling and ship-based measurements. Spatially, snow depth and ice thickness are higher in the more deformed ice of the western Weddell. The impact of undersampling the thin end of the snow depth distribution on the regional statistics, due to the resolution of the snow radar, is assessed. Radar freeboards (uncompensated for snow thickness) from CryoSat-2 (CS-2) sampled along the same transect are consistently higher (by up to 8 cm) than those computed using OIB data. This suggests radar scattering that originates above the snow–ice interface, possibly due to salinity in the basal layer of the snow column. Consequently, sea ice thicknesses computed using snow depth estimates solely from differencing OIB and CS-2 freeboards (without snow radar) are therefore generally higher; mean differences in sea ice thickness along a transect are up to ∼0.6 m higher (in 2014). This analysis is relevant to the use of differences between ICESat-2 and CS-2 freeboards to estimate snow depth for ice thickness calculations. Our analysis also suggests that, even with these expected biases, this is an improvement over the assumption that snow depth is equal to the total freeboard, with which the underestimation of thickness could be up to a meter. Importantly, better characterization of the source of these biases is critical for obtaining improved estimates and understanding the limits of retrievals of Weddell Sea ice thickness from satellite altimeters.

Surface height and total freeboard from the Ice, Cloud, and Land Elevation Satellite‐2 (ICESat‐2, IS‐2) sea ice data products (ATL07/ATL10) are assessed with near‐coincident retrievals from the Airborne Topographic Mapper (ATM) lidar in four dedicated underflights during the 2019 Operation IceBridge Arctic deployment. Over a mix of seasonal and older ice, we find remarkable correlations between the ATM and IS‐2 height profiles and roughness (in ninety‐nine 10‐km segments) that averages to >0.95 and > 0.97, respectively. Regression slopes near unity, between 0.93 and 0.99, indicate close agreement of the height estimates. Larger differences between the surface heights are seen in rougher areas where it is more difficult for the photon heights (used in IS‐2 surface finding) to capture the surface distributions at short length scales. Total freeboard in 10‐km segments, calculated using three different approaches, show variability of 0.02 to 0.04 m. Sources of residual variance, attributable to differences between the two instruments, are discussed.

Abstract. We offer a view of the Antarctic sea ice cover from lidar (ICESat-2) and radar (CryoSat-2) altimetry, with retrievals of freeboard, snow depth, and ice thickness that span an 8-month winter between 1 April and 16 November 2019. Snow depths are from freeboard differences. The multiyear ice observed in the West Weddell sector is the thickest, with a mean sector thickness > 2 m. The thinnest ice is found near polynyas (Ross Sea and Ronne Ice Shelf) where new ice areas are exported seaward and entrained in the surrounding ice cover. For all months, the results suggest that ∼ 65 %–70 % of the total freeboard is comprised of snow. The remarkable mechanical convergence in coastal Amundsen Sea, associated with onshore winds, was captured by ICESat-2 and CryoSat-2. We observe a corresponding correlated increase in freeboards, snow depth, and ice thickness. While the spatial patterns in the freeboard, snow depth, and thickness composites are as expected, the observed seasonality in these variables is rather weak. This most likely results from competing processes (snowfall, snow redistribution, snow and ice formation, ice deformation, and basal growth and melt) that contribute to uncorrelated changes in the total and radar freeboards. Evidence points to biases in CryoSat-2 estimates of ice freeboard of at least a few centimeters from high salinity snow (> 10) in the basal layer resulting in lower or higher snow depth and ice thickness retrievals, although the extent of these areas cannot be established in the current data set. Adjusting CryoSat-2 freeboards by 3–6 cm gives a circumpolar ice volume of 17 900–15 600 km3 in October, for an average thickness of ∼ 1.29–1.13 m. Validation of Antarctic sea ice parameters remains a challenge, as there are no seasonally and regionally diverse data sets that could be used to assess these large-scale satellite retrievals.