Global_Environmental_Research_Vol.27No.1
51/80

5.2 Isotopic Compositions as a Source Signature 5.3 Application of Receptor Models Bangi, Malaysia tended to be higher than that during a non-haze period, some exceptions to this trend were observed. Employing data on carbon fractions following the IMPROVE_A protocol, OP/OC4 in TSP and (OC2 + OC3 + OP)/(EC2 + EC3) in PM2.5 were suggested as indicators of IPFs by Fujii et al.56 and Tham et al.89, respectively. OP indicates pyrolyzed OC. Even for PM2.5 samples collected in Malaysia, Fujii et al.50, 67 confirmed the effectiveness of OP/OC4 as an indicator for IPFs. However, in Singapore, there were no significant differences in OP/OC4 between IPF-dominant smoke (2.1 ± 1.2) and non-smoke-dominant samples (2.3 ± 1.6). We calculated (OC2 + OC3 + OP)/(EC2 + EC3) by using data on carbon fractions in TSP collected in Bangi, Malaysia during strong haze, light haze and non-haze periods56. The results showed significant differences between haze (strong haze: 9.07 ± 0.29, light haze: 3.72 ± 0.61) and non-haze samples (2.04 ± 0.69) even in Malaysia. The usefulness of this indicator is demonstrated at source and receptor sites in the dataset. Although further validation of this indicator is needed, (OC2 + OC3 + OP)/(EC2 + EC3) seems to be useful compared to OP/OC4. MN/LG has been used in the source assignment of specific biofuels (e.g., Alves et al., 2010). Based on a report by Fabbri et al. (2009), differences in MN/LG between hardwood burning (0.042 – 0.077) and softwood burning (0.15 – 0.26) can be shown106. MN/LG in PM at IPF sources44, 106 ranged widely from 0.021 to 0.095. Whether of Indonesian26 or Malaysian peat83, 90, these ratios were less than 0.088. Thus, IPFs can be roughly categorized as hardwood burning. Based on data on LG and MN in TSP collected in Bangi, Malaysia during haze and non-haze periods reported by Fujii et al.56, LG/MN was calculated. The results showed MN/LG during strong haze, light haze and non-haze periods to be 0.12, 0.113 ± 0.015 and 0.108 ± 0.014, respectively, showing no obvious differences between haze and non-haze samples. Other IPF indicators have been suggested (e.g., SA/LG106 and SA/VA49) based on data taken at IPF sources. The effectiveness of these indicators, however, has not been fully verified at receptor sites. The above-mentioned indicators are summarized in Tables S6 and S7 in the Supplementary Information, based on the dataset. obtained from Some studies have been conducted to investigate the isotopic composition of haze from IPFs as a potential signature of these fires7, 15, 62, 80 and wind-blown mineral dust particles during IPF events118. For PM2.1 samples collected in Singapore during the 1997 IPF and in Indonesia after the major fires had died out, the carbon isotropic ratio of TC (δ13CTC, defined in the “Appendix” sheet in the dataset) showed a decrease from about −25.5‰ to −27.5‰ with an increase in TC Chemical Properties of the Southeast Asian Haze from Indonesian Peatland Fires burning laboratory concentration (9 – 81 µgC m-3)7. The results suggest an addition of OA from the combustion of C3 plants whose δ13C is lower. The δ13C of individual PAHs in the haze (TSP) in Malaysia from the 1997 IPF ranged from −17.7 to −27.9‰15. The isotopic ratios were similar to those of soot PAHs extracted from gasoline and diesel vehicles; therefore the δ13C of the individual PAHs in the haze cannot serve as a signature of IPFs in Malaysian urban air polluted by automobile exhaust. Wiggins et al.80 estimated the age and origin of IPF-emitted PM2.5, utilizing the radiocarbon content (14C) of carbonaceous PM2.5 collected in Singapore during haze and non-haze periods80. A Keeling plot (Mouteva et al., 2015) was used separate urban background contributions, and the Δ14C for the haze period was significantly lower than atmospheric levels (−76 ± 51‰ vs. 25 ± 3‰), corresponding to a carbon pool of combusted organic matter with a mean turnover time of 800 ± 420 years estimated by a radiocarbon box model80. Under an assumption about TC EF and peat Δ14C profiles, a Monte Carlo analysis indicated that the smoke plumes reaching Singapore originated primarily from peat burning (∼85 %), and not from deforestation fires or waste burning80. Isotopic ratios of Pb (206Pb/207Pb and 208Pb/207Pb) in TSP samples collected during the IPF-derived haze period in Singapore in 2013 were 1.169 ± 0.012 and 2.448 ± 0.013, respectively118. There was a statistically significant difference between the two ratios during haze and non-haze periods (1.159 ± 0.005 and 2.438 ± 0.006). The difference in the Pb isotope ratios could be explained by mixing 10% – 70% of mineral dust (coarse) particles from suspension of crustal soil during fire events118. The contribution of such mineral particles to TSP during the haze period was supported by enhancement of the Al/Pb ratio (average: 4,721) comparable to the ratio in crustal material, with the average ratio during non-haze periods being 669118. In addition to chemical characterization, quantitative knowledge of IPF source contribution to ambient PM in SEA provides valuable information to policy makers for mitigating IPF-haze pollution67. In this section, the results of IPF source apportionment by chemical mass balance (CMB) and positive matrix factorization (PMF) models are described. Detailed information on CMB and PMF models, as well as other receptor models, are provided in 209Po and 210Po activities were determined for three extracted fractions of Fe and Mn oxides, organic matter and residuals in PM10 (marine aerosol) sampled for two years along the east coast of Malaysia 62. During the southwest monsoon period when strong transboundary haze affects this area, a significant amount of excess 210Po in fractions bound to organic matter (119 ± 85 μBq m-3, cf. 54 ± 55 of ambient values), which are regulated by seasonal changes62, could be an IPF signature. to 45

元のページ  ../index.html#51

このブックを見る