Sea ice is a dynamic biogeochemical reactor and a double interface actively interacting with both the atmosphere and the ocean. However, proper understanding of its annual impact on exchanges, and therefore potentially on the climate, notably suffer from the paucity of autumnal and winter data sets. Here we present the results of physical and biogeochemical investigations on winter Antarctic pack ice in the Weddell Sea (R. V. Polarstern AWECS cruise, June–August 2013) which are compared with those from two similar studies conducted in the area in 1986 and 1992. The winter 2013 was characterized by a warm sea ice cover due to the combined effects of deep snow and frequent warm cyclones events penetrating southward from the open Southern Ocean. These conditions were favorable to high ice permeability and cyclic events of brine movements within the sea ice cover (brine tubes), favoring relatively high chlorophyll‐a (Chl‐a) concentrations. We discuss the timing of this algal activity showing that arguments can be presented in favor of continued activity during the winter due to the specific physical conditions. Large‐scale sea ice model simulations also suggest a context of increasingly deep snow, warm ice, and large brine fractions across the three observational years, despite the fact that the model is forced with a snowfall climatology. This lends support to the claim that more severe Antarctic sea ice conditions, characterized by a longer ice season, thicker, and more concentrated ice are sufficient to increase the snow depth and, somehow counterintuitively, to warm the ice.

This study reports concentrations of iron (Fe) and organic matter in young Antarctic pack ice and during its initial growth stages in situ. Although the importance of sea ice as an Fe reservoir for oceanic waters of the Southern Ocean has been clearly established, the processes leading to the enrichment of Fe in sea ice have yet to be investigated and quantified. We conducted two in situ sea-ice growth experiments during a winter cruise in the Weddell Sea. Our aim was to improve the understanding of the processes responsible for the accumulation of dissolved Fe (DFe) and particulate Fe (PFe) in sea ice, and of particulate organic carbon and nitrogen, dissolved organic carbon, extracellular polymeric substances, inorganic macro-nutrients (silicic acid, nitrate and nitrite, phosphate and ammonium), chlorophyll a and bacteria. Enrichment indices, calculated for natural young ice and ice newly formed in situ, indicate that during Antarctic winter all of the measured forms of particulate matter were enriched in sea ice compared to underlying seawater, and that enrichment started from the initial stages of sea-ice formation. Some dissolved material (DFe and ammonium) was also enriched in the ice but at lower enrichment indices than the particulate phase, suggesting that size is a key factor for the incorporation of impurities in sea ice. Low chlorophyll a concentrations and the fit of the macro-nutrients (with the exception of ammonium) with their theoretical dilution lines indicated low biological activity in the ice. From these and additional results we conclude that physical processes are the dominant mechanisms leading to the enrichment of DFe, PFe, organic matter and bacteria in young sea ice, and that PFe and DFe are decoupled during sea-ice formation. Our study thus provides unique quantitative insight into the initial incorporation of impurities, in particular DFe and PFe, into Antarctic sea ice.

In the Southern Ocean, polynyas exhibit enhanced rates of primary productivity and represent large seasonal sinks for atmospheric CO2. Three contrasting east Antarctic polynyas were visited in late December to early January 2017: the Dalton, Mertz, and Ninnis polynyas. In the Mertz and Ninnis polynyas, phytoplankton biomass (average of 322 and 354 mg chlorophyll a (Chl a)/m2, respectively) and net community production (5.3 and 4.6 mol C/m2, respectively) were approximately 3 times those measured in the Dalton polynya (average of 122 mg Chl a/m2 and 1.8 mol C/m2). Phytoplankton communities also differed between the polynyas. Diatoms were thriving in the Mertz and Ninnis polynyas but not in the Dalton polynya, where Phaeocystis antarctica dominated. These strong regional differences were explored using physiological, biological, and physical parameters. The most likely drivers of the observed higher productivity in the Mertz and Ninnis were the relatively shallow inflow of iron‐rich modified Circumpolar Deep Water onto the shelf as well as a very large sea ice meltwater contribution. The productivity contrast between the three polynyas could not be explained by (1) the input of glacial meltwater, (2) the presence of Ice Shelf Water, or (3) stratification of the mixed layer. Our results show that physical drivers regulate the productivity of polynyas, suggesting that the response of biological productivity and carbon export to future change will vary among polynyas.