Improved simulation of regional CO₂ surface concentrations using GEOS-Chem and fluxes from VEGAS

Abstract. CO₂ measurements have been combined with simulated CO₂ distributions from a transport model in order to produce the optimal estimates of CO₂ surface fluxes in inverse modeling. However, one persistent problem in using model–observation comparisons for this goal relates to the issue of compatibility. Observations at a single station reflect all underlying processes of various scales. These processes usually cannot be fully resolved by model simulations at the grid points nearest the station due to lack of spatial or temporal resolution or missing processes in the model. In this study the stations in one region were grouped based on the amplitude and phase of the seasonal cycle at each station. The regionally averaged CO₂ at all stations in one region represents the regional CO₂ concentration of this region. The regional CO₂ concentrations from model simulations and observations were used to evaluate the regional model results. The difference of the regional CO₂ concentration between observation and modeled results reflects the uncertainty of the large-scale flux in the region where the grouped stations are. We compared the regional CO₂ concentrations between model results with biospheric fluxes from the Carnegie-Ames-Stanford Approach (CASA) and VEgetation-Global-Atmosphere-Soil (VEGAS) models, and used observations from GLOBALVIEW-CO₂ to evaluate the regional model results. The results show the largest difference of the regionally averaged values between simulations with fluxes from VEGAS and observations is less than 5 ppm for North American boreal, North American temperate, Eurasian boreal, Eurasian temperate and Europe, which is smaller than the largest difference between CASA simulations and observations (more than 5 ppm). There is still a large difference between two model results and observations for the regional CO₂ concentration in the North Atlantic, Indian Ocean, and South Pacific tropics. The regionally averaged CO₂ concentrations will be helpful for comparing CO₂ concentrations from modeled results and observations and evaluating regional surface fluxes from different methods.