Comparison of СO<sub>2</sub> Content in the Atmosphere of St. Petersburg According to Numerical Modelling and Observations

Due to the increase in CO2 content in the Earth’s atmosphere, which is highly dependent on anthropogenic emissions of CO2, quality of emission estimation should be improved. Advanced experiment-based methods of the CO2 anthropogenic emission estimation are built on solution of an inverse problem using highly-accurate measurements of CO2 content and numerical models of transport and chemistry in the atmosphere. The accuracy of such models greatly determines errors of the emission estimations. In a current study temporal variations of column-average CO2 content in an atmospheric layer from surface to the height of ~70–75 km (XCO2) in the Russian megapolis of St. Petersburg during Jan 2019–Mar 2020 simulated by WRF-Chem model and measured by IR Fourier-transform spectrometer Bruker EM27/SUN are compared. The research has demonstrated that the WRF-Chem model simulates well the observed temporal variation of XCO2 in the area of St. Petersburg (correlation coefficient of ~0.95). However, using CarbonTracker v2022-1 data as chemical boundary conditions, the model overestimates XCO2 relative to the observations significantly during almost the whole period of investigation – systematic difference and standard deviation of the difference are 4.2 and 1.9 ppm (1 and 0.5%). A correction of the chemical boundary conditions which is based on analysis of a relation between near-surface wind direction and XCO2 variation notably decreases the systematic difference between the modelled and observed data (almost by a factor of 2). The XCO2 variation by the observations and modelling with uncorrected chemical boundary conditions are in a better agreement during vegetation season. Probably this is related to the compensation of the systematic difference by inaccuracies in estimated biogenic contribution. Hence, the reason of the still existing mean difference between the modelled and observed data can be inaccuracies in setting chemical boundary conditions for upper troposphere and in estimating how biosphere influences CO2 content.