||Aeolian dust episodes (ADEs) are emerging disasters occurred from the bare lands of the Kaoping River in southern Taiwan due to typhoons or thermal convections. Four manual sampling sites located along the Kaoping River were conducted to collect PM10 (aerodynamic diameter ≤ 10 μm) with high-volume samplers in an ADE and on regular days in 2012, as well as PM2.5 (aerodynamic diameter ≤ 2.5 μm) and PM2.5–10 (aerodynamic diameter 2.5–10 μm) in six ADEs in 2013. Additionally, soil samples were collected at five potential locations on the surface of bare lands along the Kaoping River Valley. The five soil samples were completely mixed and then sieved with a Tyler 400 mesh (dp < 38 μm) and then resuspended using a dry powder atomizer in a resuspension chamber. Each soil sample could be divided into two independent fractions (i.e., PM2.5 and PM2.5–10). With regard to the ADE and alluvial samples, this study investigated on their chemical contents, including a total of 13 metallic elements, 9 water-soluble ionic species, and 2 carbonaceous species. |
Hourly averaged PM10 concentrations increased drastically from noon to evening, and maximum PM10 concentration levels were reached within 3–4 hours. Sea-salt particles (SSs) in PM10 accounted for 3.56%-5.17% on regular days and 11.66%-16.47% during the ADE. Cl- deficit percentages during the ADE (6.33%-14.12%) were much lower than those on regular days (29.49%-40.38%), indicating acidic particles mainly produced by chemical reactions of acidic aerosols with aeolian dust and SSs. Even alkaline aeolian dust is the dominant source of PM10 during the ADE; the atmospheric particles are attributable to acidic particles in the air. Furthermore, ADEs were clustered by the prevailing wind direction as southern and northwestern types according to four Taiwan Environmental Protection Administration air quality monitoring stations along the Kaoping River in southern Taiwan in 2013. With metallic element analysis and nonparametric statistical methods of Wilcoxon rank-sum test and Kruskal-Wallis test, this study successfully derived the metallic indicators of ADEs. The mass ratios of crustal elements (Fe, Ca, or Al) to reference element (Cd) obtained during the ADEs were much higher than those obtained after the ADEs. High mass ratios of Fe/Cd, Ca/Cd, and Al/Cd in PM2.5-10 were observed over the influenced areas of ADEs. Among them, (Fe/Cd)2.5-10 was proven as the best indicator which can be applied to effectively validate the existence of ADEs and evaluate their influences on ambient air quality. Moreover, PM2.5 concentrations during the ADEs were 3-3.6 fold higher than those after the ADEs. PM2.5 should be a contributor to AD, even though the mass ratios of PM2.5/PM10 ranged from 0.05 to 0.20 during the ADEs. Our findings provide valuable information regarding the characteristics of the AD during the ADEs in the Kaoping River.
The CMB receptor modeling results that aeolian dust and sea-salts in PM10 were major components of atmospheric particles during the cluster ADEs. The contribution of AD emitted from the bare lands to PM10 concentration was in the range of 11.5%-33.1% along the Kaoping River during the ADEs as well as 7.2%-23.0% after the ADEs. A small amount of finer aeolian dust emitted from the bare lands of the riverbed could still be suspended in the ambient air during the ensuing the ADEs. The AD in PM2.5 ranged from 6.2% to 15.7% during the S-type ADEs and ranged from 1.3% to 17.4% during the NW-type ADEs. Both of them were less than that of PM10 during the S- and NW-type ADEs. The AD was mainly enriched in PM2.5-10 rather than PM2.5 since the formation of PM2.5 was directly related to the process that high-temperature vapors and low volatility compound chemical transformed to PM2.5. Additionally, the contribution of biomass burning rose significantly in the range of 7.6% to 13.9% after the S-type ADEs and 5.6% to 13.3% after the NW-type ADEs, suggesting the open burning of agricultural debris is commonly observed along the Kaoping River in summer. Based on the source apportionment of PM10, the wind speed and wind direction were directly relevant to different contributors, so the S- and NW-type ADEs cause the difference of major contributors for different sampling sites.