Decoding temporal patterns and trends of PM10 pollution over Delhi: a multi-year analysis (2015-2022)

Abstract

The current study delved into an extensive analysis of multi-year observations on PM₁₀ to have trends at various time scales in Delhi, India. High-resolution ground observations from all 37 monitoring stations from 2015 to 2022 were used. This study used non-parametric generalized additive model (GAM) based smooth-trend and Theil-Sen slope estimator techniques to analyze temporal trends and variations. The long-term PM₁₀ concentration, both in its ambient and de-seasonalized forms, exhibited a statistically significant decreasing trend. An average decrease of - 7.57 [95% confidence interval (CI) - 16.51, 0.18] µg m^-3 year^-1 for ambient PM₁₀ and - 8.45 [95% CI - 11.96, - 5.58] µg m^-3 year^-1 for de-seasonalized PM₁₀ mass concentration was observed. Breaking it down into seasons, we observed significant declines in PM₁₀ concentrations during monsoon (- 10.71 µg m^-3 year^-1, p < 0.1) and post-monsoon (- 7.49 µg m^-3 year^-1, p < 0.001). On the other hand, summer and winter displayed statistically insignificant declining trends of - 5.32 µg m^-3 year^-1 and - 6.06 µg m^-3 year^-1, respectively. Remarkably, all months except March displayed declining PM₁₀ concentrations, suggesting a gradual reduction in particle pollution across the city. Further analysis of PM₁₀ across various wind sectors revealed a consistent decreasing trend in all wind directions. The most substantial decrease was observed from the northwest (- 10.24 µg m^-3 year^-1), while the minimum reduction occurred from the east (- 5.67 µg m^-3 year^-1). Throughout the 8-year study period, the daily average PM₁₀ concentration remained at 228 ± 124 µg m^-3, ranging from 33 to 819 µg m^-3. Seasonal variations were apparent, with concentrations during winter, summer, monsoon, and post-monsoon seasons averaging 279 ± 133, 224 ± 117, 135 ± 95, and 323 ± 142 µg m^-3, respectively. November had the highest and August had the lowest concentration. Weekend PM₁₀ concentration is slightly lower than weekdays. These findings emphasize the need for more stringent government action plans.

Keywords: De-seasonalization; GAM smooth-trends and Theil-Sen trends; Particulate matter; Seasonal variations.