To evaluate the impact of air pollution prevention and control policies on the spatiotemporal evolution of Ozone (O₃) and fine particulate matter (PM_2.5), this study analyzed the spatiotemporal stratified heterogeneity of O₃ and PM_2.5 pollution across 331 Chinese cities from 2014 to 2024 using a Bayesian spatiotemporal hierarchical model. The GeoDetector model was further employed to quantify the determinant power and interactive effects of meteorological, socioeconomic, and other pollutant factors on summer O₃ and winter PM_2.5 pollution during the implementation of three key policies: the Air Pollution Prevention and Control Action Plan (2014~2017), the Blue Sky Defense Action Plan (2018~2020), and a circular on further Promoting the Nationwide Battle to Prevent and Control Pollution (2021~2024). The results indicate that O₃ concentration exhibited a trend of initial rapid growth followed by a slower rate, with hotspots mainly distributed in North China and the Yangtze River Delta (YRD). Conversely, PM_2.5 concentration continuously decreased, and high-value areas shrank and concentrated in North China, the YRD, and the Chengdu-Chongqing region. A positive correlation between O₃ and PM_2.5 was frequently observed in the southern and northeastern humid regions, whereas a negative correlation prevailed in the northwestern semi-arid areas. Although the dominant factors for O₃ and PM_2.5 evolved across policy stages, the determinant power of the interactions among different factors were consistently stronger than that of individual factors. Specifically, from 2014 to 2017, O₃ was mainly affected by maximum temperature (Geo_q=0.24), industrial output (Geo_q=0.22), and PM₁₀ (Geo_q = 0.60); PM_2.5 was primarily affected by maximum temperature (Geo_q = 0.41), population density (Geo_q= 0.40), and PM₁₀ (Geo_q = 0.88). From 2018 to 2020, the dominant factors for O₃ shifted to precipitation, population density, and PM₁₀, while those for PM_2.5 shifted to average temperature, population density, and PM₁₀. From 2021 to 2024, the dominant factors for both shifted toward average temperature, population density, and PM₁₀, albeit with significant differences in their determinant power. This research reveals the dynamic evolution of the spatiotemporal patterns of O₃ and PM_2.5 pollution driven by policies implementation, emphasizing the dominant role of multi-factor interactions in the composite pollution formation process. The findings provide a crucial scientific basis for formulating regionalized and phased synergistic control strategies for O₃ and PM_2.5.