We address an ongoing debate regarding the geographic distribution of interannual variability in ocean - atmosphere carbon exchange. We find that, for 1983-1998, both novel high-resolution atmospheric inversion calculations and global ocean biogeochemical models place the primary source of global CO₂ air-sea flux variability in the Pacific Ocean. In ocean biogeochemical models, this variability is clearly associated with the El Niño / Southern Oscillation cycle. Both inversion and models indicate that the Southern Ocean is the second-largest source of air-sea CO₂ flux variability, and that variability is small throughout the Atlantic, including the North Atlantic, in contrast to previous studies.

A discrepancy exists between current estimates of the Southern Ocean air-sea flux of CO₂.  The most recent estimate using a combination of direct and climatologically-derived pCO₂ measurements [Takahashi et al., 2002] (herein referred to as T02) suggests a Southern Ocean CO₂ sink that is nearly two times greater that that suggested from general circulation models, atmospheric inverse models [Gurney et al., 2002] and oceanic inverse models [Gloor et al., 2003]. Here we employ an independent method to estimate the Southern ocean air-sea flux of CO₂.  Our method exploits all available surface measurements for Dissolved Inorganic Carbon (DIC) and total alkalinity (ALK) from 1986 to 1996. We show that surface age-normalized DIC can be predicted to within ~8mmol/kg and ~10mmol/kg for ALK using standard hydrographic properties, independent of season.  The predictive equations are used in conjunction with World Ocean Atlas (2001) climatologies to estimate an annual cycle of DIC and ALK, while the pCO₂ distribution is calculated using standard carbonate chemistry.  For consistency we use the same gas transfer relationship and wind product from Takahashi et al, [2002] however, we include the effects of sea-ice. We estimate a Southern Ocean CO₂ sink (>40°S) of -0.19±0.26 Pg C for 1995. Our estimates are smaller than those estimated by Takahashi et al, [2002], but consistent with atmospheric / oceanic inverse methods, general circulation models and provides further evidence that the Southern Ocean CO₂ sink in relation to its oceanic surface area, is moderate on a global scale.

Four CARIOCA Lagrangian buoys drifted in the North-East Atlantic Ocean between 38° and 45°N between February and August 2001. Daily cycles of pCO₂, SST and DIC are observed even in winter. Biological rates of carbon consumption, gross and net primary production,are determined in situ from the amplitude of the diel cycles and the time evolution of surface dissolved inorganic carbon. Over the 6 months period, February-August, the ocean in the studied area is a sink for atmospheric CO_2.The mean absorbed flux is equal to 3.8 mmoles/ m²/ day.

Since 1993, regular seasonal water sampling has been conducted along a ship-track between Island and Newfoundland in the open ocean of the North Atlantic subpolar gyre in the frame of the long-term SURATLANT program. In this study, we analyse the interannual variation of the carbon dioxide system, including seawater fugacity (fCO₂) and air-sea CO₂ fluxes for the period 1993-2004. During 1993-1997, the data present a clear seasonality in this region marked by a strong CO₂ sink in summer and near-equilibrium in winter. For recent years, 2001-2004, we observed a dramatic change of the source/sink seasonality. An extreme case was observed in 2003 when oceanic fCO₂ was above equilibrium during all seasons. This strong anomaly was driven by ocean warming.

The decadal variability of air-sea CO₂ fluxes is poorly known in the southern hemisphere. To evaluate the changes or stability of these fluxes over several years, we compare seasonal observations obtained in 1991 and 2000 the Southern Indian Ocean. For summer and winter, we observed a significant increase of ocean fugacity (fCO₂) in subtropical waters (20°-35°S), about the same rate as in the atmosphere. In polar waters south of 40°S where meso-scale biological activity is high in summer, the rising of oceanic fCO₂ is only well detected when comparing austral winter data. The decadal evolution of fCO₂ observed in the cold waters certainly results from anthropogenic CO₂ emissions, but is also probably modulated by variations of primary production.

Inverse modeling methods are now commonly used for estimating surface fluxes of carbon dioxide, using atmospheric mass fraction measurements combined with a numerical atmospheric transport model. Michalak et al. [2004] recently developed a geostatistical approach to flux estimation that takes advantage of the spatial and/or temporal correlation in fluxes and does not require prior flux estimates. In this work, a geostatistical implementation of a fixed-lag Kalman smoother is developed and applied to the recovery of gridscale carbon dioxide fluxes for 1997 – 2001 using data from the NOAA-CMDL Cooperative Air Sampling Network.

We report the interannual variations of winter CO₂ partial pressure in surface waters (pCO₂^sea) and overlying air (pCO₂^air) and air-sea CO₂ flux in the extensive area (3-34°N) from subtropical to equatorial along 137°E during the period of 1983-2003. The pCO₂^sea varied largely in the equatorial region of 3-6°N, depending on the variations of the oceanographic conditions related to the El Niño-Southern Oscillation (ENSO) events. The pCO₂^sea variations in the subtropical gyre north of 23°N were small due to highly counteracting effects between anti-correlated sea surface temperature (SST) and dissolved inorganic carbon (DIC) anomalies through the entrainment process, irrespective of large variations of SST. By contrast, it was found that there occurred a low negative correlation between SST and DIC in the region restricted around 15-18°N in the North Equatorial Current, which resulted in a large amplitude of variations of pCO₂^sea and hence CO₂ influx. The interannual variations of CO₂ flux depended predominantly on those of the difference between pCO₂^sea and pCO₂^air (ΔpCO₂) south of 18°N but on those of wind speed in the northern region. 

Physical processes in the Southern Ocean are known to profoundly impact the global carbon cycle, but this region is one of the most difficult to simulate consistently in ocean general circulation models (OGCMs). Here we show that Southern Hemisphere winds, by altering the volume of light, actively-ventilated ocean water as well as the relative contribution to this volume from Ekman transport, exert strong control over both the magnitude and distribution of anthropogenic carbon uptake in an OGCM. These results are provocative in suggesting that climate warming, by increasing the magnitude of the wind stress at high southern latitudes, may act as a negative feedback on the global carbon cycle.

Accurate estimates of the spatial distribution of pre-industrial and anthropogenic air-sea carbon fluxes are crucial to understanding the processes driving ocean carbon uptake. We present regional anthropogenic and pre-industrial air-sea fluxes estimated separately from their reconstructed concentrations and Ocean General Circulation Models (OGCM). The ocean interior carbon transports required to explain these fluxes are calculated and their implications for the global carbon cycle are discussed. 

The Orbiting Carbon Observatory (OCO) mission will deliver space-based observations of atmospheric CO₂ with the potential to resolve many of the uncertainties in the spatial and temporal variability of carbon sources and sinks.  Our assessments of the measurement requirements for space-based remote sensing of atmospheric CO₂ conclude that the data must support retrievals of the column-averaged CO₂ dry air mole fraction, X_CO2, with precisions of 3 to 4 ppm to resolve the annually averaged gradients between the Northern and Southern hemispheres, but higher precision (1 to 2 ppm) will be needed to resolve East-West gradients and questions like the location and spatial extent of the Northern Hemisphere terrestrial carbon sink.  These conclusions are derived from the results of observational system simulation experiments (OSSEs) and synthesis inversion models [Rayner and O’Brien, 2001; O’Brien and Rayner, 2002; Rayner et al., 2002]. The X_CO2 precision requirements also considered the OCO mission design, the amplitude of X_CO2 spatial and temporal gradients, and the relationship between X_CO2 data precision and regional scale surface CO₂ flux uncertainties inferred from X_CO2 data.