Abstract: Air pollution concentration levels are typically dependent on both emissions from pollution sources and meteorological conditions. To accurately assess the impact of emission changes on pollutant mass concentrations, it is essential to reduce the interference from meteorological factors. In this study, the Kolmogorov-Zurbenko(KZ)filtering method is applied to analyze the temporal and seasonal characteristics of the time components of PM_2.5, PM₁₀ and O₃ mass concentrations in Suqian from March 1, 2020 to March 1, 2024. For each pollutant, an optimal regression model is developed to adjust for meteorological factors, enabling quantitative evaluation of the respective contributions of pollution source emissions and meteorological conditions to the observed concentration trends. The results indicate that the long-term component of PM_2.5, driven primarily by pollution source emissions, remained relatively stable. In contrast, the long-term component of PM₁₀ showed a fluctuating downward trend, while for O₃ it showed a fluctuating upward trend. Compared with the period from March 1, 2020 to March 1, 2021, the contributions of pollution source emissions and meteorological conditions to the PM_2.5 trend during March 1, 2023 to March 1, 2024 are 0.46 and 1.27 μg·m^-3, respectively, for PM₁₀, the corresponding contributions are -3.36 and 3.58 μg·m^-3, while for O₃ they are 5.25 and -2.70 μg·m^-3. These findings suggest that the relative impacts of pollution source emissions and meteorological conditions differ substantially across pollutants, underscoring the need for pollutant-specific and targeted air quality management strategies.