Estimating Background Secondary Organic Aerosol in the Southeastern United States from a Regionally Representative Site
Supplemental Information
Michael Link1, Yong Zhou1, Brett Taubman1*, JamesSherman2, HadiMorrow1, Ian Krintz2, Luke Robertson2, Ryan Cook1, Justine Stocks1, Matthew West2, Barkley C. Sive1
[1]Department of Chemistry, Appalachian State University, Boone, NC 28608, USA
[2]Department of Physics, Appalachian State University, Boone, NC 28608, USA
Corresponding author E-mail ; phone +1 (828) 262-7847
Seasonal variability of total aerosol measured at AppalAIR by the AMS
Total measured PM2.5 by the quadrupole AMS include sulfate, nitrate, organics, ammonium, and chloride. The current study was primarily concerned with characterizing regional influences and abundances of trace gases and secondary organic aerosol at the AppalAIR site and the implications of these measurements for organic aerosol chemistry in the SEUS. However, to demonstrate the importance of secondary organic aerosol to total ambient aerosol measured at the AppalAIR site the seasonal variability of total measured aerosol by the AMS will be discussed here briefly for the summer of 2012, winter of 2013, and summer of 2013.
Secondary organic aerosol dominated total aerosol loading over all recorded times of measurement from the AMS at the AppalAIR site as shown in Figure S-1.
Figure S-1. Time series of ammonium, organics, nitrate, and sulfate as measured by the AMS for the (a) summer of 2012, (b) summer of 2013, and (c) winter of 2013. Relative contributions of each measured species to total measured aerosol are displayed in the pie charts.
PMF factor Spectra from winter dataset.
Figure S-2. Time Series and mass spectra for resolved PMF factors from December 19, 2012–March 28, 2013; BBOA (top panel), and LV-OOA (bottom panel).
Estimation of the Planetary Boundary Layer (PBL)
Radiosonde measurements of meteorological conditions such as temperature and humidity were acquired for the months of June, July and August during the summer of 2013. PBL heights were calculated from these measurements following the methods of Seidel et al. (2010) and Compton et al. (2013). The results of these calculations are displayed in Figure S-3.
Figure S-3. Measured PBL from the Virtual Temperature (red) and Specific Humidity (blue) methods and categorized by month. In the box and whisker plots the solid black line represents the median, the top and bottom of the boxes represent the 90th and 10thpercentiles, and the top and bottom whiskers represent the 95th and 5th percentiles, respectively.
Air mass aging using RONO2 kinetics.
Briefly, the formation and removal of alkyl nitrates in the atmosphere can be described by a simplified reaction (R1):
(R1)
If it is assumed that the primary formation and removal pathways are photochemical, where RH is the parent hydrocarbon of the alkyl nitrate, RONO2 is the alkyl nitrate,is the rate of formation of the alkyl nitrate, andis the rate for the removal of the alkyl nitrate. Based on this simplified reaction scheme a mathematical relationship between freshly emitted hydrocarbons and secondary alkyl nitrate products that evolve with time can be described by equation 1:
[RONO2]/[RH] describes the ratio of the alkyl nitrate to the parent hydrocarbon at a particular time, and the individual rate variables are based on the unique rate constants for the photochemical conversions and removal of individual alkyl nitrate species. For this study, the initial alkyl nitrate to parent hydrocarbon ratio[RONO2]0/[RH]0, was assumed to be zero [Roberts et al., 1998; Simpson et al., 2003]. Concentrations of alkyl nitrates above the theoretical photochemical curve, described from equation 1, suggest that there are additional sources of the alkyl nitrates, and concentrations below the curve suggest there are sinks, or other unrecognized processes taking place. The primary assumption used when performing this calculation is that the primary reaction controlling alkyl nitrate formation is the reaction of the peroxy radical with NO [Atkinson et al., 1982]. When NO concentrations are low the probability that an alkylperoxy radical will react with other alkylperoxy radicals increases with increasing carbon number of the parent hydrocarbon [Bertman et al., 1995]. This serves as an additional sink within the simplified reaction described by the rate.
The distribution of the EtONO2 and n-PrONO2 data above the pure photochemistry curve has been observed in many studies performing these calculations [Bertman et al., 1995; Roberts et al., 1998; Russo et al., 2010], and has been attributed to the thermal decomposition of larger (>C5) carbon number alkoxy radicals which eventually form lower carbon number (C1-C3) alkyl nitrates. Most of the 2-PrONO2 data is also distributed above the photochemistry curve suggesting that thermal decomposition of alkoxy radicals may also contribute to the enhancement of this alkyl nitrate ratio. The data for the C5 alkyl nitrate species, 2-pentyl nitrate and 3-pentyl nitrate, were observed to be distributed below the photochemistry curve. This suggests that the alkylperoxy radicals of these species may be reacting with one another at a rate that competes with the reaction between the alkylperoxyradical with NO. Although NO was not measured in this experiment, a lack of significant local alkyl nitrate processing, in addition to the minimal influence of measured anthropogenic VOCs suggests that NO concentrations may be low enough for alkylperoxy radical self-reaction to be a significant sink for the C5alkylperoxy radical thus lowering the ratio of C5 alkyl nitrates to their parent hydrocarbons.
Assessment of regional representation.
Table S-1. Latitudinal and longitudinal ranges, correlation values for ln(AOD) versus temperature, and fitting parameters for ln(AOD) versus temperature for pixels in SEUS (Figure 8).
Pixel / Latitude Range (0N) / Longitude Range (0W) / a / aSlope β / aIntercept α1 / 37.787-35.632 / 83.566-86.689 / 0.57 / 0.077 / -3.32
2 / 37.787- 35.632 / 80.443-83.566 / 0.60 / 0.085 / -3.52
3 / 37.787- 35.632 / 77.32-80.443 / 0.61 / 0.085 / -3.43
4 / 35.632-33.477 / 85.223-88.346 / 0.50 / 0.067 / -3.11
5 / 35.632-33.477 / 82.100-85.223 / 0.58 / 0.082 / -3.46
6 / 35.632-33.477 / 78.977-82.100 / 0.64 / 0.091 / -3.52
7 / 33.477-31.322 / 87.637-90.79 / 0.59 / 0.081 / -3.41
8 / 33.477-31.322 / 84.514-87.637 / 0.61 / 0.089 / -3.53
9 / 33.477-31.322 / 81.391-84.514 / 0.62 / 0.089 / -3.53
bBoone / 36.259-36.169 / 81.638-81.748 / 0.60 / 0.097 / -3.97
aResults from the linear regression of ln(AOD) versus.
bRegional representation was accessed through consistency of correlation values and fitting parameters with the values obtained from the area above Boone.