Galen McKinley
University of Wisconsin - Madison
| Using a global version of the MIT ocean general circulation model driven by analyzed winds and fluxes, global air-sea CO2 flux variability are considered. The mean air-sea CO2 flux is consistent with the study of Takahashi et al. (2002), and the interannual variability from 1980-1998 is ±0.5 PgC/yr. Consistent with other ocean modeling studies, we find the equatorial Pacific to be the primary driver of global interannual air-sea flux variability, with the higher latitudes playing a minor role. Variability in the equatorial Pacific is driven by large-scale coordinated physical changes associated with ENSO. In the high latitude North Atlantic, by contrast, flux variability is driven by small-scale convective variability. Pronounced intra-regional cancellation of air-sea flux anomalies, the slow air-sea exchange timescale of CO2, and a rapid biological response to convective nutrient supply result in small net regional interannual variability. Our finding of small high latitude flux variability contrasts with recent atmospheric inversion results (Bousquet et al. 2000), but agrees with the inversion of Rodenbeck et al. (2003). The former may be compromised by the use of fixed spatial patterns based on the mean air-sea flux. This may bias the inversion toward large high-latitude variability because inter-regional flux cancellation, as observed in our modeling study, cannot occur. Rodenbeck et al. (2003) do not impose the mean-flux pattern, reducing high-latitude variability and substantially improving model-inversion agreement. |