The ozone mass concentration is affected by meteorological elements and emissions from air pollution sources. In order to quantitatively evaluate the effect of air pollution control measures, it is necessary to separate the contributions from air pollution sources. The Kolmogorov-Zurbenko filter was used to decompose the time series of daily ozone mass concentration during 2014-2019 as well as time series of meteorological factors in the same period in Guangzhou into long-term, short-term and seasonal components, and the variance contribution of each component to the total variance of the original ozone mass concentration data was calculated. Then multiple linear stepwise regression method was used to establish the relationship between ozone mass concentration data and meteorological variables for each time scale, the contributions of meteorological factors and pollutant emissions to ozone mass concentration were separated to obtain the contribution of ozone mass concentration influenced by meteorological conditions only. The results are as follows: (1) The long-term series of ozone mass concentration generally fluctuated and increased, the seasonal component showed a high value in late spring and early summer, a secondary peak in late summer and early autumn, and a trough in winter. (2) By analyzing the variance contribution rates of each component to the total variance of ozone mass concentration, the short-term component contributed the most, followed by the seasonal component, and the long-term component contributed the least, which indicated that the fluctuation of ozone mass concentration in Guangzhou was mainly caused by the short-term and seasonal changes of meteorological conditions and precursor emissions, and the long-term changes of emissions and climate conditions were not the main reasons for the fluctuation of ozone mass concentration. (3) In terms of explanatory variance, meteorological variables had the highest explanatory ability to the long-term component of ozone mass concentration, followed by the seasonal component. (4) The long-term component of ozone mass concentration which eliminated the influence of meteorological conditions by stepwise regression showed a fluctuating decreasing trend. Combined with the aggravation of ozone pollution from 2014 to 2019 in Guangzhou, the meteorological conditions were unfavorable for ozone diffusion in recent years.