The first phase of TransCom, which involved about a dozen modeling groups from around the world, examined the atmospheric concentration response to surface emissions of fossil fuel CO2 and the activity of terrestrial ecosystems. These experiments were designed to address two salient features of atmospheric CO2:
(1) the annual mean north-south (meridional) gradient arising from fossil fuel emissions; and
(2) the seasonal cycle arising from the seasonal exchange of CO2 between the atmosphere and terrestrial ecosystems, with a net zero flux at each grid point but with strong uptake during the growing season balanced by release by decomposition.
Results of this initial intercomparison were reported by Rayner and Law (1995) and by Law et al (1996). With a few exceptions, there was good agreement among the models with regard to the annual mean meridional distribution of the "fossil fuel" tracer at the surface. A few models simulated extremely strong interhemispheric gradients of fossil-fuel at the surface, and these models also simulated very low concentrations aloft over the emissions region. This suggests that the high surface mixing ratios simulated by these models resulted from vertical trapping of tracer in the emissions region, rather than from weak interhemispheric transport.
There was qualitative agreement for the seasonal variations of CO2 in the biosphere experiment, though this agreement diminishes over continental regions which lack observational constraints. The annual mean meridional response of the models to seasonal biotic forcing can be classified into two groups. Models which represent turbulent mixing in the planetary boundary layer simulate a pole-to-pole gradient in surface CO2 that is roughly half as strong as that obtained in the fossil fuel experiment. The other models simulate a very weak meridional structure in these runs.