Modeling Carbon-Climate Interactions

How fast greenhouse warming will proceed depends on a large part on how fast carbon dioxide is increasing in the atmosphere. The abundance and growth rate of atmospheric CO2 is determined not only by the rate of anthropogenic emissions but also by the rates of biogeochemical processes that sequester carbon in the land and ocean. The biogeochemical processes in turn responds to and alters climate and circulation changes. A goal of carbon cycle modeling is to understand the interactions among the physical and biogeochemical processes that exchange carbon among the mobile reservoirs on seasonal to millennial time scales, and to project the future co-evolution of CO2 and climate. The task requires the synthesis of atmospheric, oceanic and terrestrial observations of carbon inventories and carbon fluxes, and the coupling of the physical and biogeochemical systems. CO2 is radiatively active but chemically inert in the atmosphere. The geographic and intraseasonal variations are observed to be small (<10%) and are radiatively insignificant compared to the global annual mean concentrations (~350 ppmv). In this way CO2 departures from the global mean (referred to as “tracer CO2”) are unimportant for the radiative and biogeochemical processes, and can be treated as an inert tracer in the atmosphere. This has permitted the traditional separation of climate models and carbon cycle models. In climate models, the global annual mean CO2 concentration is specified. In carbon cycle models, the only crucial atmospheric and oceanic processes to model are those related to large-scale advection and turbulent mixing; the challenge lies in the prescription of source/sink functions for the tracer. In this paper, we review the status of carbon cycle modeling, and discuss the outlooks for global environment modeling, both in the future and of the future.