Abstract: Using the measurement data from an 8-level vertical profile of CO₂/H₂O in a temperate deciduous broadleaved forest at the Maoershan Forest Ecosystem Research Station, Northeast China, this paper quantified the errors of CO₂ storage flux (F_s) calculated with three scalar  variables, i.e., CO₂ density (ρ_c), molar fraction (c_c), and molar mixing ratio relative to dry air (χ_c). The dry air storage in the control volume of flux measurement was not a constant, and thus, the fluctuation of the dry air storage could cause the CO₂molecules transporting out of or into the control volume, i.e., the variation of the dry air storage adjustment term (F_sd). During nighttime and day-night transition periods, the relative magnitude of F_sd to eddy flux  was larger, and ignoring the F_sd could introduce errors in calculating the net CO₂ exchange between the forest ecosystem and the atmosphere. Three error sources in the F_s calculation could be introduced from the atmospheric hydrothermal processes, i.e., 1) air temperature fluctuation, which could cause the largest error, with one order of magnitude larger than that caused by atmospheric pressure (P), 2) water vapor, its effect being larger than that of P in warm and moist summer but smaller in cold and dry winter, and 3) P, whose effect was generally smaller throughout the year. In estimating the effective CO₂ storage (F_{s_E}), the F_s value calculated with ρ_c, c_c, and χ_c was overestimated averagely by 8.5%, 0.6%, and 0.1%, respectively. It was suggested that in the calculation of F_s, adopting the χ_c conservative to atmospheric hydrothermal processes could be more appropriate to minimize the potential errors.