2025 Volume 60 Issue 5 Pages 115-120
Atmospheric simulation models play a significant role in supporting the decision by the local government to issue photochemical-oxidant warnings. We evaluated the performance of an air pollution forecast system, VENUS, for simulating atmospheric ozone in the summer of 2024. Simulated ozone concentrations using meteorological forecast data were generally comparable in accuracy to those using meteorological analysis data. The VENUS also predicted the occurrence of high ozone events, with daily maximum ozone concentrations (DM1h-O3) over 100 ppbv in individual regions, achieving a true alert ratio of 70% and a detection probability of 55%. In addition, by implementing a bias correction using quantile mapping, the VENUS predicted the occurrence of high ozone events with DM1h-O3 over 120 ppbv, with a true alert ratio and a detection probability of 45%–50%.
