Chang et al. 2018 - Shanghai atmospheric trace elements

Chang and colleagues continuously measured 18 trace elements in PM2.5 at an urban Shanghai receptor site for one year, then used positive matrix factorization to apportion the elements to traffic, shipping, nonferrous smelting, coal combustion, and ferrous smelting. This is atmospheric source-attribution evidence, not mineral-water occurrence evidence: it identifies which emission sectors load metals into the shared environment.

Key numbers

The study measured hourly PM2.5 elemental concentrations from 1 March 2016 to 28 February 2017. The PMF model used n=1265 measurements after excluding missing data. The abstract states that mass concentrations determined by Xact ranged from nominal detection limits of 0.1 to 20 ng m-3 up to 15 ug m-3, and that element-related oxidized species accounted for 8.3% of PM2.5 on average.

The 1 h minimum detection limits in ng m-3 were Si 17.80, K 1.17, Ca 0.30, V 0.12, Cr 0.12, Mn 0.14, Fe 0.17, Ni 0.10, Cu 0.27, Zn 0.23, As 0.11, Se 0.14, Ag 1.90, Cd 2.50, Au 0.23, Ba 0.39, Hg 0.12, and Pb 0.13. The paper reports total elemental As, Hg, and Cr by XRF, not inorganic arsenic, methylmercury, or Cr(VI).

Across the full study period, mean +/- 1 sigma elemental concentrations were Si 640 +/- 1010 ng m-3, Fe 410 +/- 390 ng m-3, K 390 +/- 330 ng m-3, Ca 190 +/- 380 ng m-3, Zn 120 +/- 130 ng m-3, Mn 32 +/- 39 ng m-3, Pb 27 +/- 26 ng m-3, Ba 24 +/- 25 ng m-3, V 13 +/- 15 ng m-3, Cu 12 +/- 11 ng m-3, Cd 10 +/- 4 ng m-3, As 7 +/- 7 ng m-3, Ni 6 +/- 5 ng m-3, Cr 5 +/- 6 ng m-3, Ag 4 +/- 2.6 ng m-3, Se 2.6 +/- 2.9 ng m-3, Hg 2.2 +/- 1.7 ng m-3, and Au 2.2 +/- 3.4 ng m-3.

The five-factor PMF solution apportioned PM2.5 trace elements to traffic-related sources, shipping, nonferrous metal smelting, coal combustion, and ferrous metal smelting. The abstract reports relative contributions of 46% for traffic-related sources, 6% for shipping, 15% for nonferrous metal smelting, 18% for coal combustion, and 15% for ferrous metal smelting.

The traffic-related factor was the largest source, accounting for 46% or 680 ng m-3 of the measured elemental mass in PM2.5. Factor 1 was characterized by Ca, Fe, Ba, and Si, explaining 90%, 50%, 77%, and 63% of the concentration respectively. The factor’s average rush-hour concentration was more than 2 times higher than during nighttime.

The shipping factor contributed 5.9% of the trace elements in the Shanghai urban center. V and Ni were the marker elements: V was 100% assigned to the shipping factor, and Ni was 74% assigned to it. The average measured V/Ni ratio was 3.2, close to the heavy-oil combustion ratio of 3.4 cited for Shanghai port aerosols from ocean-going ship engines.

The nonferrous-metal-smelting factor contributed 15% or 220 ng m-3 to the total measured elemental mass in PM2.5. It was dominated by Au, Cd, Ag, and Hg: Au was 100% assigned to the factor, Cd 65%, Ag 63%, and Hg 37%.

The coal-combustion factor was dominated by As, Se, Pb, Hg, Cu, and K. The paper states that As, Se, Pb, and Hg explained 56% to 95% of the concentration in factor 4, K contributed 53%, and coal combustion contributed 280 ng m-3 or 19% of PM2.5 trace elements during the study period. Seasonal coal-combustion-related PM2.5 trace-element concentrations were 410 ng m-3 in winter, 270 ng m-3 in spring, 150 ng m-3 in summer, and 210 ng m-3 in fall.

The ferrous-metal-smelting factor contributed 220 ng m-3 or 15% of the PM2.5 trace elements in Shanghai. Factor 5 was distinguished by Cr, Mn, and Zn, representing 100%, 56%, and 52% of the concentration respectively.

The precipitation analysis identified 12 precipitation events, with duration times from 7 to 55 h and accumulated rainfall from 26 to 220 mm. The authors found that precipitation could effectively lower traffic-related and coal-combustion-related trace-element concentrations, while effects varied by species and source.

Methods (brief)

Hourly elemental measurements used an Xact 625 multi-metals monitor drawing air through a PM2.5 cyclone inlet at 16.7 L min-1 onto reel-to-reel Teflon filter tape, with non-destructive energy-dispersive X-ray fluorescence analysis. Daily PM2.5 samples were also collected and analyzed by ICP-MS as a reference method; the paper reports good agreement for several elements but poorer data quality for As, Cd, and Ba. Source apportionment used US EPA PMF version 5.0 with 3-10 factor solutions examined, then conditional probability function and bivariate polar plot analyses to interpret source regions.

Implications

Certification: Do not use these atmospheric PM2.5 concentrations as food, water, or consumer-product occurrence data. They are upstream atmospheric-deposition and source-apportionment evidence.

Courses: Strong source-attribution case study because it separates traffic, shipping, nonferrous smelting, coal combustion, and ferrous smelting signatures, including V/Ni for shipping and As/Se/Pb/Hg for coal.

App: Context only. The source supports ambient/shared-liability explanations for regional metal loading, not product-level contamination estimates.

Wiki pages this source may touch

• atmospheric-deposition
• source-attribution-environmental-burden-apportionment
• lead
• cadmium
• arsenic-total
• nickel
• chromium
• mercury-total
• vanadium
• zinc

Verification notes

Recovered from skip:not-food-occurrence under the 2026-06-10 inclusion-by-default rule. The old skip treated the paper as out of scope because it did not measure mineral water or a finished product, but it is in-scope a3 atmospheric source-attribution evidence.

The Pb-query and Cd-query PDFs have identical extracted text and the same DOI. The Cd-query worklist row is covered as a recovered duplicate by this source page. Numbers were checked against the extracted PDF text, especially the abstract, Methods, Section 3.1, Table 1, PMF source-analysis sections 3.2.1-3.2.5, and the precipitation-effect section. Species are preserved as total elemental measurements; the source does not report inorganic arsenic, methylmercury, or Cr(VI).

Page history

The five most recent substantive edits to this page. The full version history lives in git; when DOI minting comes online (see schema docs), each entry below will also link to a version-pinned DataCite DOI.