Spatio-temporal variability of atmospheric CO2 and its main causes: A case study in Xi'an city, China

Cities account for most anthropogenic carbon dioxide (CO2) emissions, and thus play an important role in carbon mitigation. However, the variability of atmospheric CO2 caused by the complex emission sources and topography within city remains a challenge in evaluating urban CO2 emissions. Specifically, the mechanism of spatial and temporal patterns is still unclear at city scale. In this work, we explored the spatial and seasonal patterns of atmospheric CO2 at the city scale in Xi'an and attempted to illustrate the mechanisms of these patterns based on the analysis of CO2 sources and sinks. Using the observed radiocarbon (14C) in atmospheric CO2 as a tracer, we determined the CO2 emissions from fossil fuel (CO2ff) and the biosphere (CO2bio) in different seasons and city areas. The results showed that CO2ff in urban areas is significantly (p < 0.05) higher than that in suburban areas, and explained 82% and 74% of the spatial differences in winter and summer, respectively. Based on the analysis of δ13C of CO2 in the local sources, the depleted δ13C of urban site indicated greater influences from traffic or natural gas in the urban area than that in the suburban area. These findings suggest that the spatial differences of atmospheric CO2 are dominated by different emission intensities of fossil fuels between urban and suburban areas. In contrast, the contributions of CO2ff to the seasonal variations of atmospheric CO2 were different between urban and suburban areas, that is CO2ff contributed as much as CO2bio to the seasonal variations of atmospheric CO2 in suburban area, but the contribution increased with the increasing of human activities from suburban to urban area, and accounted for 57% of seasonal variations of atmospheric CO2 in the urban area. In addition, we also found that under the influence of the special topography of Xi'an city, the frequent fluctuations of atmospheric CO2 concentration in winter largely depend on the different wind conditions.