Urban forests play a vital role in mitigating mercury pollution, primarily through mercury accumulation in leaves. However, the dynamics of this accumulation and its economic implications are not well understood. This study examines mercury sequestration in urban forest leaves across the growing season in Changchun, a major industrial city in northeast China. A grid-based sampling method was used, with 526 leaf samples, 104 atmospheric samples, and 104 soil samples collected each month for mercury content analysis. The results revealed significant temporal variations in leaf mercury concentration, peaking in June and declining thereafter, with higher concentrations observed in the inner urban rings. A random forest model, interpreted using SHAP (Shapley Additive Explanations), identified atmospheric pressure (with an average SHAP value of 23.06), Number_6-114 and Number_6-48, representing specific combinations of months and sampling points (with average SHAP values of 11.33 and 8.03, respectively), and Moisture Content(MC), Diameter at Breast Height (DBH), or Humidity (with average SHAP values ranging from 2 to 3) as key predictors of mercury concentration. The high mercury concentrations observed in June were likely linked to seasonal physiological changes in trees. Furthermore, economic valuation of the ecosystem services provided by urban forests showed significant variations across seasons and tree species. Spring and summer exhibited stronger mercury absorption capacities, with poplar and ash trees providing higher economic value. This study offers valuable insights for urban forest management and mercury pollution control, highlighting the importance of considering seasonal variations and the economic value of mercury accumulation when selecting tree species.