Title : Seasonal strength and structural stability in urban air pollution: A comparative analysis of NO2 and SO2 dynamics in two major US cities
Abstract:
Urban air pollution displays complex temporal characteristics influenced by pollutant composition, regulatory frameworks, and atmospheric conditions. While numerous studies have focused on forecasting pollutant concentration levels, limited attention has been given to pollutant-specific temporal dynamics within the same environment. This study provides a comparative analysis of monthly nitrogen dioxide (NO2) and sulfur dioxide (SO2) concentrations in Baltimore, MD, and New York, NY, from 2018 to 2024. A forecasting model was developed for 2024-2026 and validated using observations from January-September 2025. Results indicate distinct temporal behaviour across pollutants. NO2 shows a strong, consistent seasonal pattern in both cities, characterized by recurring winter peaks and summer troughs. In contrast, SO2 exhibits a moderate seasonal pattern accompanied by a significant degree of stochastic variability. Additionally, while some residual dependence is detected in SO2 levels in Baltimore, no structural regime changes are evident for both pollutants. These findings suggest that transportation-related pollutants such as NO2 exhibit stable seasonal structures, whereas industrial-related pollutants such as SO2 display weaker seasonal coherence and greater variability. By reframing air pollution forecasting to emphasize pollutant-specific temporal patterns rather than merely concentrating on levels, this research seeks to offer a novel perspective on analysing air pollution within a defined environment.
