Key factors affecting atmospheric visibility during the 19th Asian Games in Hangzhou

Identifying the key factors that influence atmospheric visibility is critical for effective air pollution prevention and control. However, the complex nonlinear relationships between visibility and its influencing factors pose significant challenges. Most current studies rely on empirical models for statistical analysis, which can introduce significant inaccuracies.

This study integrates the traditional IMPROVE analysis method with machine learning models to investigate the driving factors of atmospheric visibility and the causes of low visibility events during the 19th Asian Games in Hangzhou. The results indicate that SO₄²⁻, NO₃⁻, organic aerosols (OA), and black carbon (BC) are the primary light extinction contributors among PM_2.5 chemical components. Secondary formation processes were the dominant factor that accounted for 57 % of extinction coefficient, relative humidity and vehicles together contributed 30 % of extinction coefficient. During periods of low visibility (<10 km), the contributions of secondary nitrates and relative humidity content to extinction coefficient increased significantly, suggesting that hygroscopic growth of secondary nitrates was the primary driver of such events.

The results of the study show that machine learning models closely match the IMPROVE approach during periods of low visibility. In addition, machine learning models outperform the IMPROVE method in their ability to capture the complex non-linear relationships between forcing factors and visibility. To effectively manage secondary nitrate pollution, it is critical to reduce emissions of nitrogen oxides (NOx) and ammonia (NH₃) while improving overall air quality. This comprehensive analytical approach provides scientific evidence for improving urban atmospheric visibility.