Inter-annual variability of the atmospheric carbon dioxide concentrations as simulated with global terrestrial biosphere models and an atmospheric transport model

Seasonal and inter-annual variations of atmospheric CO2 for the period from 1961 to 1997 have been simulated using a global tracer transport model driven by a new version of the Biome BioGeochemical Cycle model (Biome-BGC). Biome-BGC was forced by daily temperature and precipitation from the NCEP reanalysis dataset, and the calculated monthly-averaged CO2 fluxes were used as input to the global transport model. Results from an inter-comparison with the Carnegie-Ames-Stanford Approach model (CASA) and the Simulation model of Carbon CYCLE in Land Ecosystems (Sim-CYCLE) model are also reported. The phase of the seasonal cycle in the Northern Hemisphere was reproduced generally well by Biome-BGC, although the amplitude was smaller compared to the observations and to the other biosphere models. The CO2 time series simulated by Biome-BGC were compared to the global CO2 concentration anomalies from the observations at Mauna Loa and the South Pole. The modeled concentration anomalies matched the phase of the inter-annual variations in the atmospheric CO2 observations; however, the modeled amplitude was lower than the observed value in several cases. The result suggests that a significant part of the inter-annual variability in the global carbon cycle can be accounted for by the terrestrial biosphere models. Simulations performed with another climate-based model, Sim-CYCLE, produced a larger amplitude of inter-annual variability in atmospheric CO2, making the amplitude closer to the observed range, but with a more visible phase mismatch in a number of time periods. This may indicate the need to increase the Biome-BGC model sensitivity to seasonal and inter-annual changes in temperature and precipitation