Characterization of paddy-residue burning derived carbonaceous aerosols using dual carbon isotopes

Abstract

A large scale paddy-residue burning (PRB) happens every year in the northwest Indo-Gangetic Plain (IGP) during the post-monsoon season, and winds transport pollutants from the source region up to the northern Indian Ocean affecting air quality of the IGP and marine region. In this study, day-night pairs of fine aerosol samples (n = 69) were collected during October–November over Patiala (30.2°N, 76.3°E, 250 m amsl), a site located in the source region of PRB. Carbonaceous aerosols (CA) were characterised using chemical species and dual carbon isotopes (13C and 14C) to estimate bio vs non-bio contributions and understand their characteristics. Percentage of bio fraction (fbio, estimated using 14C) in CA varied from 74 % to 87 % (avg: 80 ± 3) during days and 71 % to 96 % (avg: 85 ± 7 %) during nights. Further, the fbio was found to be better correlated with aerosol mass spectrometer derived f60 compare to levoglucosan (LG) or nsssingle bondK+, suggesting f60 a useful proxy for PRB. The δ13C varied from -27.7 ‰ to -26.0 ‰ (avg: −27.0 ± 0.4 ‰) and - 28.7 ‰ to -26.4 ‰ (avg: -27.5 ± 0.7 ‰) during day and night, respectively. Measured δ13C of the samples was found to be more enriched than expected by 0.3 to 2.0 ‰, indicating the presence of aged CA also in Patiala even during PRB period. From fbio versus δ13C correlation, and from Miller-Trans plot, δ13C of PRB is found to be −28.9 ± 1.1 ‰, which also infers that Miller-Trans plot can be used to understand source isotopic signature in the absence of radiocarbon measurements in aerosols. Further, the characteristics ratios of organic carbon (OC) to elemental carbon (EC) (11.9 ± 4.1), LG to potassium (K+) (0.84 ± 0.15), OC/LG (19.7 ± 2.0) and K+/EC (0.75 ± 0.27) were calculated by considering samples with fbio higher than 0.90, which can be used for source apportionment studies. Such studies are crucial in assessing the effects of PRB on regional air quality and climate.