The WRF-Chem model（v4. 4）was employed to simulate ozone（O3）concentrations in the Haikou re‐ gion of Hainan Province during the autumn of 2016（September to November），complemented by the FLEX‐ PART-WRF model for backward trajectory analysis. The research aimed to quantify the contributions of physical and chemical processes to O3 variations，examine the characteristics of regional transport，and identify potential source areas. Process analysis was applied to evaluate the influence of various mechanisms on O3 concentration changes. Results indicated significant diurnal variation in the roles of chemical processes，vertical mixing and dry deposition. During the day，chemical reactions contributed an average of 11. 5 μg·m-3 to O3 levels，with a peak of 17. 8 μg·m-3，while a net negative contribution was observed at night. Vertical mixing and dry deposition processes were major removal mechanisms during daylight hours but weakened considerably at night. Although advection transport had a relatively minor impact，its influence varied under different air mass origins. Cluster analysis of FLEXPART-WRF-simulated backward trajectories demonstrated that O3 concentrations in Haikou were strongly affected by the direction of incoming air masses. Higher concentrations were associated with north‐ ern and northeastern air masses，while lower levels were linked to cleaner air masses from the southeast and south. Overall，chemical processes dominated the increase in O3 concentrations，whereas vertical mixing and dry deposition processes served as the principal removal mechanisms. The contribution of advection transport to changes in O3 concentration is relatively small. Potential source regions were further identified using the Potential Source Contribution Function（PSCF）and Concentration Weighted Trajectory（CWT）methods. Key source ar‐ eas were located in the northeastern coastal regions and the southeastern part of Guangdong Province，contribut‐ ing an average of 60~70 μg·m-3 per month. These findings provide critical insight into the evolution of atmo‐ spheric composition in tropical coastal cities and offer a scientific basis for the development of regional air quali‐ ty management strategies.