We develop and extend a theoretical framework to analyze the impacts of changes in temperature and alkalinity on the ocean-atmosphere carbon partitioning. When investigating the impact of temperature, we assume that there is no change in the global ocean alkalinity. This idealized situation is probably most relevant on intermediate timescales of hundreds to thousands of years. Our results show that atmospheric pCO2 depends approximately exponentially on the average ocean temperature, since the chemical equilibria involved have an exponential (Arrhenius-type) dependence. The dependence of pCO2 on alkalinity is more complicated, and our theory suggests several regimes. The current ocean-atmosphere system appears to have an exponential dependence of pCO2 on global mean ocean alkalinity, but at slightly higher alkalinities, the dependence becomes a power law. We perform experiments with a numerical physical-biogeochemical model to test the validity of our analytical theory in a more complex, ocean-like system: in general, the numerical results support the analytical inferences.< br />
©2011 American Geophysical Union.