We report first direct measurements of the partial pressure of CO₂ (pCO₂) within Antarctic pack sea ice brines and related CO₂ fluxes across the air‐ice interface. From late winter to summer, brines encased in the ice change from a CO₂ large oversaturation, relative to the atmosphere, to a marked undersaturation while the underlying oceanic waters remains slightly oversaturated. The decrease from winter to summer of pCO₂ in the brines is driven by dilution with melting ice, dissolution of carbonate crystals, and net primary production. As the ice warms, its permeability increases, allowing CO₂ transfer at the air‐sea ice interface. The sea ice changes from a transient source to a sink for atmospheric CO₂. We upscale these observations to the whole Antarctic sea ice cover using the NEMO‐LIM3 large‐scale sea ice‐ocean and provide first estimates of spring and summer CO₂ uptake from the atmosphere by Antarctic sea ice. Over the spring‐summer period, the Antarctic sea ice cover is a net sink of atmospheric CO₂ of 0.029 Pg C, about 58% of the estimated annual uptake from the Southern Ocean. Sea ice then contributes significantly to the sink of CO₂ of the Southern Ocean.

We present an Arctic seasonal survey of carbon dioxide partial pressure (pCO₂) dynamics within sea ice brine and related air‐ice CO₂ fluxes. The survey was carried out from early spring to the beginning of summer in the Arctic coastal waters of the Amundsen Gulf. High concentrations of pCO₂ (up to 1834 μatm) were observed in the sea ice in early April as a consequence of concentration of solutes in brines, CaCO₃ precipitation and microbial respiration. CaCO₃ precipitation was detected through anomalies in total alkalinity (TA) and dissolved inorganic carbon (DIC). This precipitation seems to have occurred in highly saline brine in the upper part of the ice cover and in bulk ice. As summer draws near, the ice temperature increases and brine pCO₂ shifts from a large supersaturation (1834 μatm) to a marked undersaturation (down to almost 0 μatm). This decrease was ascribed to brine dilution by ice meltwater, dissolution of CaCO₃ and photosynthesis during the sympagic algal bloom. The magnitude of the CO₂fluxes was controlled by ice temperature (through its control on brine volume and brine channels connectivity) and the concentration gradient between brine and the atmosphere. However, the state of the ice‐interface clearly affects air‐ice CO₂ fluxes.

The impacts of the seasonal evolution of sea‐ice physical properties on ice‐ocean biogeochemical exchanges were investigated in landfast ice at Barrow (Alaska) from January through June 2009. Three stages of brine dynamics across the annual cycle have been identified based on brine salinity, brine volume fraction, and porous medium Rayleigh number (Ra). These are sea‐ice bottom‐layer convection, full‐depth convection, and brine stratification. We further discuss the impact of brine dynamics on biogeochemical compounds in sea ice: stable isotopes of water (δD, δ¹⁸O), nutrients (NO₃⁻, PO₄³⁻, NH₄⁺), microalgae (chlorophyll‐a), and inert gas (argon). In general, full‐depth convection events favor exchanges between sea ice and seawater, while brine stratification limits these exchanges. However, argon responds differently to brine dynamics than the other biogeochemical compounds analyzed in this study. This contrast is attributed to the impact of bubble nucleation on inert gas transport compared to the other biogeochemical compounds. We present a scenario for argon bubble formation and evolution in sea ice and suggest that a brine volume fraction approaching 7.5–10% is required for inert gas bubbles to escape from sea ice to the atmosphere.

Although the presence of a gas phase in sea ice creates the potential for gas exchange with the atmosphere, the distribution of gas bubbles and transport of gases within the sea ice are still poorly understood. Currently no straightforward technique exists to measure the vertical distribution of air volume fraction in sea ice. Here, we present a new fast and non-destructive X-ray computed tomography technique to quantify the air volume fraction and produce separate images of air volume inclusions in sea ice. The technique was performed on relatively thin (4–22 cm) sea ice collected from an experimental ice tank. While most of the internal layers showed air volume fractions < 2 %, the ice–air interface (top 2 cm) systematically showed values up to 5 %. We suggest that the air volume fraction is a function of both the bulk ice gas saturation factor and the brine volume fraction. We differentiate micro bubbles (Ø < 1 mm), large bubbles (1 mm < Ø < 5 mm) and macro bubbles (Ø > 5 mm). While micro bubbles were the most abundant type of gas bubbles, most of the air porosity observed resulted from the presence of large and macro bubbles. The ice texture (granular and columnar) as well as the permeability state of ice are important factors controlling the air volume fraction. The technique developed is suited for studies related to gas transport and bubble migration.