Analysis of Driving Factors of Ozone in the Urban Area of Meizhou Based on Machine Learning

Key Points

• Ozone levels show slight variation in Meizhou from 2019 to 2023, with peaks in September and April.
• Weather factors contribute 63% to ozone generation, while pollutants contribute 37%, particularly influenced by photochemical conditions.
• Assessment using machine learning techniques like random forest and SHAP modeling illustrates seasonal differences in driving factors for ozone concentration.
• Clusters indicate that high ozone levels in Meizhou mainly arise from long-distance northern transport and short-distance southern influences.

Abstract