The rate of atmospheric CO₂ has exceeded the atmospheric system’s natural ability to efficiently absorb it, thereby gradually affecting global climatic warming. The use of model simulations has become essential to help us improve our understanding of the carbon cycle. We simulated CO₂ concentration using the atmospheric chemical transport model GEOS-Chem driven by Biosphere Model SiB3 (Simple Biosphere version 3) and optimized Carbon Tracker 2016 (CT2016) terrestrial biospheric carbon fluxes for 2008-2010. Knowing the importance of improving carbon flux inversion accuracy in terrestrial ecosystems through analysis of the spatial distribution and seasonal variation in simulated CO₂ concentrations can deepen our understanding of global carbon source distribution and explore the impact of carbon flux uncertainty on simulation results. SiB3 and the optimized CT2016 terrestrial ecosystem carbon flux show obvious seasonal changes, but are opposite in European regions and there is great uncertainty about the global total and spatial distribution. Simulation results show that the carbon flux of terrestrial ecosystems are found to be key in the distribution of CO₂ concentration near the ground in areas of less human activity, particularly in the southern hemisphere and Europe. Seasonal differences in the simulation re sults rely on the seasonal variations in the terrestrial biospheric carbon fluxes. Simulation results are compared with the data of nine observation sites to select the suitable terrestrial ecosystem carbon flux for improving the accuracy of the GEOS-Chem simulated CO₂ concentration. Experimental results indicated that two kinds of simulation results can significantly simulate the peak and valley of seasonal change of CO₂ concentration, although the simulated CO₂ concentration using CT2016 is close to the observation data with a high simulation accuracy at most stations.