A comparison between the numerical simulation results of ozone fields with different experimental data makes it possible to estimate the quality of models for their further use in reliable forecasts of ozone layer evolution. We analyze time series of satellite (SBUV) measurements of the total ozone column (TOC) and the ozone partial columns in two atmospheric layers (0-25 and 25-60 km) and compare them with the results of numerical simulation in the chemistry transport model (CTM) for the low and middle atmosphere and the chemistry climate model EMAC. The daily and monthly average ozone values, short-term periods of ozone depletion, and long-term trends of ozone columns are considered; all data sets relate to St. Petersburg and the period between 2000 and 2014. The statistical parameters (means, standard deviations, variations, medians, asymmetry parameter, etc.) of the ozone time series are quite similar for all datasets. However, the EMAC model systematically underestimates the ozone columns in all layers considered. The corresponding differences between satellite measurements and EMAC numerical simulations are (5 +/- 5)% and (7 +/- 7)% and (1 +/- 4)% for the ozone column in the 0-25 and 25-60 km layers, respectively. The correspondent differences between SBUV measurements and CTM results amount to (0 +/- 7)%, (1 +/- 9)%, and (-2 +/- 8)%. Both models describe the sudden episodes of the ozone minimum well, but the EMAC accuracy is much higher than that of the CTM, which often underestimates the ozone minima. Assessments of the long-term linear trends show that they are close to zero for all datasets for the period under study.