Source apportionment of PM2.5 organic carbon in the San Joaquin Valley using monthly and daily observations and meteorological clustering. (June 2018)

• Record Type: Journal Article
• Authors: Skiles, Matthew J.; Lai, Alexandra M.; Olson, Michael R.; Schauer, James J.; de Foy, Benjamin
• Is Part Of: Environmental pollution. Volume 237(2018)
• Page Start: 366
• Page End: 376
• Publication Date: 2018-06
• Subjects: cmb -- San joaquin valley -- Organic molecular markers -- Meat smoke -- Meteorological cluster analysis; Pollution -- Periodicals; Pollution -- Environmental aspects -- Periodicals; Environmental Pollution -- Periodicals; Pollution -- Périodiques; Pollution -- Aspect de l'environnement -- Périodiques; Pollution -- Effets physiologiques -- Périodiques; Pollution; Pollution -- Environmental aspects; Periodicals; Electronic journals; 363.73
• DOI: 10.1016/j.envpol.2018.02.055
• Languages: English
• ISSNs: 0269-7491

Abstract

Abstract: Two hundred sixty-three fine particulate matter (PM2.5 ) samples collected on 3-day intervals over a 14-month period at two sites in the San Joaquin Valley (SJV) were analyzed for organic carbon (OC), elemental carbon (EC), water soluble organic carbon (WSOC), and organic molecular markers. A unique source profile library was applied to a chemical mass balance (CMB) source apportionment model to develop monthly and seasonally averaged source apportionment results. Five major OC sources were identified: mobile sources, biomass burning, meat smoke, vegetative detritus, and secondary organic carbon (SOC), as inferred from OC not apportioned by CMB. The SOC factor was the largest source contributor at Fresno and Bakersfield, contributing 44% and 51% of PM mass, respectively. Biomass burning was the only source with a statistically different average mass contribution (95% CI) between the two sites. Wintertime peaks of biomass burning, meat smoke, and total OC were observed at both sites, with SOC peaking during the summer months. Exceptionally strong seasonal variation in apportioned meat smoke mass could potentially be explained by oxidation of cholesterol between source and receptor and trends in wind transport outlined in a Residence Time Analysis (RTA). Fast moving nighttime winds prevalent during warmer months caused local emissions to be replaced by air mass transported from the San Francisco Bay Area, consisting of mostly diluted, oxidized concentrations of molecular markers. Good agreement was observed between SOC derived from the CMB model and from non-biomass burning WSOC mass, suggesting the CMB model is sufficiently accurate to assist in policy development. In general, uncertainty in monthly mass values derived from daily CMB apportionments were lower than that of CMB results produced with monthly marker composites, further validating daily sampling methodologies. Strong seasonal trends were observed for biomass and meat smoke OC apportionment, and monthly mass averages had lowest uncertainty when derived from daily CMB apportionments. Graphical abstract: Monthly averaged mass concentrations with uncertainty bars for a) Mobile Sources, b) Meat Smoke, c) CMB Other, and d) Biomass Burning obtained from three distinct CMB averaging methodologies for the Fresno site. CMB Method 3 results were derived from daily mass apportionments averaged to obtain monthly mass values with uncertainty defined as the standard error of values in the average. CMB Method 1 and CMB Method 2 results were derived from monthly averaged molecular markers, run through the CMB model with uncertainty inputs defined as the standard error and standard deviation of marker data included in the mean, respectively.Image 1 Highlights: ∙∼300 days of organic tracers from Bakersfield and Fresno were analyzed with MM-CMB. ∙Monthly mass averages derived from daily CMB results had the lowest uncertainties. ∙High wintertime mass averages were driven by biomass burning and meat smoke events. ∙Meat smoke tracer seasonality was impacted by seasonal oxidation and transport.