Search

Source apportionment studies have struggled to quantitatively link secondary organic aerosols (SOAs) to their precursor sources due largely to instrument limitations. For example, aerosol mass spectrometer (AMS) provides quantitative measurements of the total SOA fraction but lacks the chemical resolution to resolve most SOA sources. In contrast, instruments based on soft ionisation techniques, such as extractive electrospray ionisation mass spectrometry (EESI, e.g. the EESI time-of-flight mass spectrometer, EESI-TOF), have demonstrated the resolution to identify specific SOA sources but provide only a semi-quantitative apportionment due to uncertainties in the dependence of instrument sensitivity on molecular identity. We address this challenge by presenting a method for positive matrix factorisation (PMF) analysis on a single dataset which includes measurements from both AMS and EESI-TOF instruments, denoted "combined PMF" (cPMF). Because each factor profile includes both AMS and EESI-TOF components, the cPMF analysis maintains the source resolution capability of the EESI-TOF while also providing quantitative factor mass concentrations. Therefore, the bulk EESI-TOF sensitivity to each factor can also be directly determined from the analysis. We present metrics for ensuring that both instruments are well represented in the solution, a method for optionally constraining the profiles of factors that are detectable by one or both instruments, and a protocol for uncertainty analysis. As a proof of concept, the cPMF analysis was applied to summer and winter measurements in Zurich, Switzerland. Factors related to biogenic and wood-burning-derived SOAs are quantified, as well as POA sources such as wood burning, cigarette smoke, cooking, and traffic. The retrieved EESI-TOF factor-dependent sensitivities are consistent with both laboratory measurements of SOA from model precursors and bulk sensitivity parameterisations based on ion chemical formulae. The cPMF analysis shows that, with the standalone EESI-TOF PMF, in which factor-dependent sensitivities are not accounted for, some factors are significantly under- or overestimated. For example, when factor-dependent sensitivities are not considered in the winter dataset, the SOA fraction is underestimated by ∼25 % due to the high EESI-TOF sensitivity to components of primary biomass burning such as levoglucosan. In the summer dataset, where both SOA and total OA are dominated by monoterpene oxidation products, the uncorrected EESI-TOF underestimates the fraction of daytime SOA relative to nighttime SOA (in which organonitrates and less oxygenated C x H y O z molecules are enhanced). Although applied here to an AMS and EESI-TOF pairing, cPMF is suitable for the general case of a multi-instrument dataset, thereby providing a framework for exploiting semi-quantitative, high-resolution instrumentation for quantitative source apportionment. [ABSTRACT FROM AUTHOR]

A new methodology for performing long-term source apportionment (SA) using positive matrix factorization (PMF) is presented. The method is implemented within the SoFi Pro software package and uses the multilinear engine (ME-2) as a PMF solver. The technique is applied to a 1-year aerosol chemical speciation monitor (ACSM) dataset from downtown Zurich, Switzerland. The measured organic aerosol mass spectra were analyzed by PMF using a small (14 d) and rolling PMF window to account for the temporal evolution of the sources. The rotational ambiguity is explored and the uncertainties of the PMF solutions were estimated. Factor–tracer correlations for averaged seasonal results from the rolling window analysis are higher than those retrieved from conventional PMF analyses of individual seasons, highlighting the improved performance of the rolling window algorithm for long-term data. In this study four to five factors were tested for every PMF window. Factor profiles for primary organic aerosol from traffic (HOA), cooking (COA) and biomass burning (BBOA) were constrained. Secondary organic aerosol was represented by either the combination of semi-volatile and low-volatility organic aerosol (SV-OOA and LV-OOA, respectively) or by a single OOA when this separation was not robust. This scheme led to roughly 40 000 PMF runs. Full visual inspection of all these PMF runs is unrealistic and is replaced by predefined user-selected criteria, which allow factor sorting and PMF run acceptance/rejection. The selected criteria for traffic (HOA) and BBOA were the correlation with equivalent black carbon from traffic (eBC tr) and the explained variation of m/z 60, respectively. COA was assessed by the prominence of a lunchtime concentration peak within the diurnal cycle. SV-OOA and LV-OOA were evaluated based on the fractions of m/z 43 and 44 in their respective factor profiles. Seasonal pre -tests revealed a non-continuous separation of OOA into SV-OOA and LV-OOA, in particular during the warm seasons. Therefore, a differentiation between four-factor solutions (HOA, COA, BBOA and OOA) and five-factor solutions (HOA, COA, BBOA, SV-OOA and LV-OOA) was also conducted based on the criterion for SV-OOA. HOA and COA contribute between 0.4–0.7 µgm-3 (7.8 %–9.0 %) and 0.7–1.2 µgm-3 (12.2 %–15.7 %) on average throughout the year, respectively. BBOA shows a strong yearly cycle with the lowest mean concentrations in summer (0.6 µgm-3 , 12.0 %), slightly higher mean concentrations during spring and fall (1.0 and 1.5 µgm-3 , or 15.6 % and 18.6 %, respectively), and the highest mean concentrations during winter (1.9 µgm-3 , 25.0 %). In summer, OOA is separated into SV-OOA and LV-OOA, with mean concentrations of 1.4 µgm-3 (26.5 %) and 2.2 µgm-3 (40.3 %), respectively. For the remaining seasons the seasonal concentrations of SV-OOA, LV-OOA and OOA range from 0.3 to 1.1 µgm-3 (3.4 %–15.9 %), from 0.6 to 2.2 µgm-3 (7.7 %–33.7 %) and from 0.9 to 3.1 µgm-3 (13.7 %–39.9 %), respectively. The relative PMF errors modeled for this study for HOA, COA, BBOA, LV-OOA, SV-OOA and OOA are on average ±34% , ±27% , ±30% , ±11% , ±25% and ±12% , respectively. [ABSTRACT FROM AUTHOR]

A new methodology for performing long-term source apportionment (SA) using positive matrix factorization (PMF) is presented. The method is implemented within the SoFi Pro software package and uses the multilinear engine (ME-2) as a PMF solver. The technique is applied to a one-year aerosol chemical speciation monitor (ACSM) dataset from downtown Zurich, Switzerland. The measured organic aerosol mass spectra were analyzed by PMF using a small (14 days) and rolling PMF window to account for the temporal evolution of the sources. The rotational ambiguity is explored and the uncertainty of the PMF solutions were estimated. Factor/tracer correlations for averaged seasonal results from the rolling window analysis are higher than those retrieved from conventional PMF analyses of individual seasons, highlighting the improved performance of the rolling window algorithm for long-term data. In this study four to five-factors were tested for every PMF window. Factor profiles for primary organic aerosol from traffic (HOA), cooking (COA) and biomass burning (BBOA) were constrained. Secondary organic aerosol was represented by either the combination of semi-volatile and low-volatility organic aerosol (SV-OOA and LV-OOA, respectively), or by a single OOA when this separation was not robust. This scheme leads to roughly 40 000 PMF runs. Full visual inspection of all these PMF runs is unrealistic and is replaced by predefined user-selected criteria, which allow factor sorting and PMF run acceptance/rejection. The selected criteria for traffic (HOA) and biomass burning (BBOA) were the correlation with equivalent black carbon (eBCtr) and the explained variation of m/z 60, respectively. COA was assessed by the prominence of a lunchtime concentration peak within the diurnal cycle. SV-OOA and LV-OOA were evaluated based on the fraction of m/z 43 and m/z 44 in their respective factor profiles. Seasonal pre-tests revealed a non-continuous separation of OOA into SV-OOA and LV-OOA, in particular during the warm seasons. Therefore, a differentiation between four-factor solutions (HOA, COA, BBOA and OOA) and five-factor solutions (HOA, COA, BBOA, SV-OOA and LV-OOA) was also conducted based on the criterion for SV-OOA. HOA and COA contribute between 0.4-0.7 μg m-3 (7.8-9.0 %) and 0.7-1.2 μg m-3 (12.2-15.7 %) on average throughout the year, respectively. BBOA shows a strong yearly cycle with the lowest mean concentrations in summer (0.6 μg m-3, 12.0 %), slightly higher mean concentrations during spring and fall (1.0 and 1.5 μg m-3, or 15.6 and 18.6 %, respectively), and highest mean concentrations during winter (1.9 μg m-3, 25.0 %). In summer, OOA is separated into SV-OOA and LV-OOA, with mean concentrations of 1.4 μg m-3 (26.5 %) and 2.2 μg m-3 (40.3 %), respectively. For the remaining seasons the seasonal concentrations of SV-OOA, LV-OOA and OOA range from 0.3-1.1 μg m-3 (3.4-15.9 %), 0.6-2.2 μg m-3 (7.7-33.7 %) and 0.9-3.1 μg m-3 (13.7-39.9 %), respectively. The relative PMF errors modelled for this study for HOA, COA, BBOA, LV-OOA, SV-OOA and OOA are on average ±34 %, ±27 %, ±30, ±11 %, ±25 % and ±12 %, respectively. [ABSTRACT FROM AUTHOR]