Abstract :
[en] Despite the fact that the Southern Ocean (S.O.) is a high nutrients-low chlorophyll area
(HNLC), it acts as a significant sink for atmospheric CO2. We addressed the temporal and
spatial variations of the frontal system of the Indian sector of the S.O. using remote sensing
measurements of sea surface temperature (SST) and we present the first synoptic
partitioning of the main physical-biogeochemical provinces. In the Crozet basin, if the
signature of the fronts is well marked in the mesoscale distribution of the partial pressure of
CO2 (pCO2), this latter hardly reflects the chlorophyll a (Chl a) distribution during the summer
post-bloom period. Scaling in situ pCO2 measurements using remote sensing measurements
of SST and Chl a, we assessed spring and summer air-sea flux of CO2 per physicalbiogeochemical
province. Spring and summer air-sea CO2 fluxes in the Indian sector of the
S.O. ranges from -0.048 PgC, to -0.057 PgC and -0.04 PgC in the North Subantarctic, South
Subantarctic and Polar Frontal zones, respectively. A further collaborative effort was carried
out applying a similar approach to the western Pacific sector of the S.O.. Integrating CO2
fluxes over the year shows that this area acts as a sink for atmospheric CO2 of 0.08 PgC yr-1.
Both studies provide lower estimates than the Takahashi et al. (Takahashi et al., 2002;
Takahashi, 2003) climatology but corroborate (Metzl et al., 1999; Takahashi et al., 2002) the
conclusions of inverse models, indicating that this climatology overestimates the CO2 sink in
the S.O. (Gurney et al., 2004; Jacobson et al., 2005).
We present a three years survey of pCO2 in Subantarctic coastal waters surrounding the
Kerguelen Archipelago, with a particular attention on the role of Macrocystis giant kelp beds.
Primary production of Macrocystis lasts from early spring to late autumn and is tightly linked
to solar irradiance. Maximum net kelp community production can be as high as 15 gC m-2 d-1
at the solar irradiance climax. Such production strongly affects pCO2 within kelp bed. Coastal
waters of the archipelago experience earlier and more intense phytoplanktonic blooms than
offshore waters, which markedly affect pCO2. However, over the year near-shore waters of
the archipelago act as a source of CO2 of 0.32TgC yr-1.
The role of sea ice cover in the budgets of exchanges of CO2 between the S.O. and the
atmosphere has been neglected, since it was assumed as an impermeable and inert cover
that prohibited air-sea fluxes of gases. We report the first direct measurements of pCO2
within first year pack ice, multi-year pack ice and land fast sea ice, and corresponding CO2
fluxes at the air-sea ice interface. Internal spring and summer sea ice specific processes
(dilution with ice crystals, dissolution of carbonate minerals and primary production), drive
drastic decreases of pCO2 and lead to marked undersaturation of CO2 with respect to the
atmosphere. Despite its thinness, the Antarctic sea ice cover thus appears to sustain a
significant uptake of atmospheric CO2. We scaled measurements CO2 fluxes over bare sea
ice using remote sensing measurement of sea ice surface temperature: in spring and
summer, the Antarctic sea ice cover acts as a sink of atmospheric CO2 ranging from 0.015
PgC to 0.024 PgC which represents 6% to 9% of the annual uptake of the S.O. south of 50°S
(0.27 PgC yr-1). However, we surmise that the present evaluation of the sea ice CO2 sink is
an underestimate, since it does not account for the uptake of CO2 by biologically active sea
ice surface communities. Eddy-covariance measurements of air-sea ice CO2 fluxes over
slush - a mixture of melting snow, ice and flooding seawater covering sea ice - which hosts
abundant surface communities showed 4-fold higher fluxes than over bare sea ice. On the
whole, sea ice represents an additional significant CO2 sink that should be taken into account
when budgeting exchanges of CO2 fluxes between the S.O. and the atmosphere.
