Spatiotemporal analysis of PM2.5 and its cross-multifractality with black carbon and dust aerosols in Indian cities: crop residue burning impact
摘要
The study investigates the spatiotemporal variability of PM2.5 and its cross-multifractality with black carbon (BC) and dust aerosols (DUST) across nine Indian cities from 2017 to 2023, focusing on two distinct periods: Period 1 (October-February), influenced by crop residue burning and Period 2 (March-June), dominated by dust storms. Ground-based PM2.5 data and satellite-derived BC and DUST concentrations (MERRA-2) are analysed using Multifractal Detrended Cross-Correlation Analysis (MFXDFA) to capture the nonlinear, scale-dependent interactions among pollutants. Results show distinct spatial heterogeneity, with the highest PM2.5 levels in Delhi, decreasing with distance from the crop-burning belt. Period-wise analysis reveals a marked seasonal shift in pollutant concentration, with PM2.5 and BC concentrations peak during the crop burning season, while DUST concentrations increase during the dust storm period. Period 2 exhibits up to a 56.5% reduction in PM2.5 and a modest decline in BC, alongside an increase of up to 38.9% in DUST concentrations, reflecting a seasonal shift in dominant emission sources. Cross-multifractality between PM2.5 and BC is more pronounced in Period 1, capturing the short-term, episodic nature of biomass burning events, while PM2.5-DUST interactions in Period 2 display greater multifractal complexity, likely due to dust storms. Hurst exponent analysis indicates persistent temporal dynamics in both PM2.5-BC and PM2.5-DUST, with stronger persistence in Period 2, suggesting more continuous dust-related influences compared to the episodic nature of crop burning. The multifractality strength assessed through the width of the multifractal spectra (Δβ), is greater in Period 2, supporting the presence of more complex pollutant dynamics during dust-dominated months. A significant correlation between Δβ and the mean time interval between extreme events highlights their contribution to the observed temporal variability, particularly in cities downwind of crop burning zones. The findings highlight the role of emission source characteristics and distance from the emission source on the evolving temporal dynamics of PM2.5 in urban areas. The study highlights cross-multifractality analysis as a powerful tool for uncovering nonlinear pollutant interactions and capturing city-specific pollution behaviour. The approach is extendable to other pollutants and regions to advance understanding of complex and non-linear temporal interactions.