Abstract. The rate at which atmospheric carbon dioxide (CO2) would decrease in response to a decrease in anthropogenic emissions or cessation (net zero emissions) is of great scientific and societal interest. Such a decrease in atmospheric CO2 on the centennial scale would be due essentially entirely to transfer of carbon into the world ocean (WO) and the terrestrial biosphere (TB), which are sink compartments on this timescale. The rate of decrease in excess atmospheric CO2 and the apportionment of this decrease into the two sink compartments have been examined in two prior model intercomparison studies, subsequent either to a pulse emission of CO2 or to abrupt cessation of anthropogenic CO2 emissions. The present study examines and quantifies inter-model anticorrelation in those studies in the net rate and extent of uptake of CO2 into the two sink compartments. Specifically, in each study the time-dependent coefficients characterizing the net transfer rate into the two sink compartments (evaluated as the net transfer rate normalized to excess atmospheric CO2 above the pre-pulse amount for the pulse experiment or as the net transfer rate divided by excess atmospheric CO2 above the preindustrial amount for the abrupt cessation experiment) were found to exhibit strong anticorrelation across the participating models. That is, models for which the normalized rate of uptake into the WO was high exhibited low uptake rate into the TB and vice versa. This anticorrelation in net transfer rate results in anticorrelation in net uptake extent into the two compartments that is substantially greater than would be expected simply from competition for excess CO2 between the two sink compartments. This anticorrelation, which is manifested in diminished inter-model diversity, can lead to artificially enhanced confidence in current understanding of the consequences of potential future reductions of CO2 emissions and in the global warming potentials of non-CO2 greenhouse gases relative to that of CO2.
Read full abstract