Uncertainty assessment of source attribution of PM2.5 and its water-soluble organic carbon content using different biomass burning tracers in positive matrix factorization analysis — a case study in Beijing, China

https://doi.org/10.1016/j.scitotenv.2015.11.057Get rights and content

Highlights

  • Various biomass burning tracers were compared.

  • PM2.5 source profiles were compared using different biomass burning tracers.

  • Uncertainties in source attribution of PM2.5 and WSOC were quantified.

Abstract

Daily PM2.5 samples were collected at an urban site in Beijing during four one-month periods in 2009–2010, with each period in a different season. Samples were subject to chemical analysis for various chemical components including major water-soluble ions, organic carbon (OC) and water-soluble organic carbon (WSOC), element carbon (EC), trace elements, anhydrosugar levoglucosan (LG), and mannosan (MN). Three sets of source profiles of PM2.5 were first identified through positive matrix factorization (PMF) analysis using single or combined biomass tracers — non-sea salt potassium (nss-K+), LG, and a combination of nss-K+ and LG. The six major source factors of PM2.5 included secondary inorganic aerosol, industrial pollution, soil dust, biomass burning, traffic emission, and coal burning, which were estimated to contribute 31 ± 37%, 39 ± 28%, 14 ± 14%, 7 ± 7%, 5 ± 6%, and 4 ± 8%, respectively, to PM2.5 mass if using the nss-K+ source profiles, 22 ± 19%, 29 ± 17%, 20 ± 20%, 13 ± 13%, 12 ± 10%, and 4 ± 6%, respectively, if using the LG source profiles, and 21 ± 17%, 31 ± 18%, 19 ± 19%, 11 ± 12%, 14 ± 11%, and 4 ± 6%, respectively, if using the combined nss-K+ and LG source profiles. The uncertainties in the estimation of biomass burning contributions to WSOC due to the different choices of biomass burning tracers were around 3% annually and up to 24% seasonally in terms of absolute percentage contributions, or on a factor of 1.7 annually and up to a factor of 3.3 seasonally in terms of the actual concentrations. The uncertainty from the major source (e.g. industrial pollution) was on a factor of 1.9 annually and up to a factor of 2.5 seasonally in the estimated WSOC concentrations.

Introduction

Fine particulate matter with an aerodynamic diameter of less than 2.5 μm (PM2.5) affects human and ecosystem health and also plays an important role in weather modification and climate change (Ali-Mohamed, 1991, Ali-Mohamed and Ali, 2001, Bell et al., 2007). PM2.5 is a complex mixture of chemical compounds mainly composing sulfate, nitrate, ammonium, OC, EC, soil dust and water (Seinfeld et al., 2004, Chan and Yao, 2008, Fang and Liu, 2010, Johansen and Hoffmann, 2004, Kumar et al., 2010, Wang et al., 2002, Zhang et al., 2011a). Water-soluble chemical components, including water-soluble inorganic ions and water-soluble organic matter, play more important roles than insoluble ones on aerosols light scattering and their ability of serving as cloud condensation nuclei (Decesari et al., 2003, Du et al., 2014b, Hecobian et al., 2010, Jung et al., 2011, Zhang et al., 2011b).

Dominant water-soluble inorganic ions (sulfate, nitrate, and ammonium) are mainly secondary aerosols forming from their respective gaseous precursors. However, water-soluble organic matter or water-soluble organic carbon (WSOC) can originate from both primary emissions and secondary organic aerosol (SOA) formations, the latter are through the oxidation processes involving volatile organic compounds (VOCs) (Kalberer et al., 2000, Pathak et al., 2011). Biomass burning rather than fossil fuel combustion was considered to be the major primary source for WSOC (Sullivan et al., 2006, Timonen et al., 2010, Wonaschütz et al., 2011, Gordon et al., 2014, Peltier et al., 2007). Thus, WSOC can be a marker for SOA in the absence of biomass burning (Docherty et al., 2008).

LG is a unique organic tracer for biomass burning because the combustion of other fuels seldom produce LG. The disadvantge of using LG as a tracer is its degradation during transportation process which may affect source apportionment results (Fraser and Lakshmanan, 2000, Popovicheva et al., 2014). The inorganic tracer K+ has also been used extensively to identify biomass burning (Gieré et al., 2006). Although K+ is not an ideal tracer due to its other sources (e.g., soil dust, sea salt, and coal combustion) (Cheng et al., 2000, Duvall et al., 2008, Ninomiya et al., 2004), it does not degrade during transportation. Moreover, seasonal dependent biomass burning types and their respective emission factors for LG and K+ further add difficulties to quantitatively assessing biomass burning contributions to PM2.5 and WSOC. For example, wheat and rape straws are burned in spring, rice and corn straws in autumn and woods in winter (Cheng et al., 2013, Tao et al., 2013, Tao et al., 2014, Wang et al., 2007, Zhang et al., 2015). Therefore, choice of biomass burning tracers could lead to some uncertainties when estimating the contribution of biomass burning to PM2.5 and WSOC.

A previous study using LG as a tracer of biomass burning suggested that biomass burning accounted for 40% of WSOC in Beijing (Du et al., 2014a). However, the uncertainties in source attribution analysis by different biomass burning tracers are unknown. To assess the sources especially biomass burning contributing to PM2.5 and WSOC in Beijing, a comprehensive data set acquired in 2009–2010 is analyzed in the present study making use of previously generated PM2.5 source profiles for this city. The study aims to accomplish the following goals: (1) to systematically characterize WSOC levels on seasonal and annual basis; (2) to identify the biomass burning profiles based on chemical components in PM2.5 using various biomass burning tracers; and (3) to quantify the contributions of biomass burning to PM2.5, WSOC and potential uncertainties in the PMF results due to choices of different biomass burning tracers. Biomass burning tracers nss-K+, LG, and a combination of nss-K+ and LG are applied separately to PMF analysis to quantify the uncertainties in source attribution analysis.

Section snippets

Site description

PM2.5 samples were collected at the Peking University (PKU) (39.99° N, 116.30° E) located in the urban area of Beijing (Fig. 1). Instruments used in this study were installed on the roof (26 m above ground) of an office building of the PKU. This site is located within the educational, commercial, and residential districts, and no main pollution sources exist nearby. Thus, the observations could represent typical pollution conditions in an urban environment of Beijing (Zhang et al., 2013).

Sampling

PM2.5

Characteristics of WSOC in PM2.5

The annual average PM2.5, OC, EC, and WSOC were 135 ± 63 μg m 3, 16.9 ± 10.0 μgC m 3, 5.0 ± 4.4 μgC m 3, and 6.4 ± 3.6 μgC m 3, respectively (Table 1). OC and EC accounted for 13.2 ± 4.7% and 3.5 ± 1.4% (expressed as μgC μg 1), respectively, of PM2.5 mass. WSOC accounted for 38.7 ± 13.1% of OC and 5.1 ± 2.3% of PM2.5 (μgC μg 1). The annual value of WSOC was comparable to that (7.2 ± 5.5 μgC m 3) based on thirteen months measurement in 2010–2011 in Beijing (Du et al., 2014a). However, the value was evidently higher than

Conclusions

To investigate the sources of WSOC in PM2.5, WSOC and other major chemical components were measured in 2009–2010 at an urban site in Beijing. Annual mean concentration of WSOC was 6.4 ± 3.6 μgC m 3 with a summer average (3.2 ± 1.1 μgC m 3) less than half of those in other seasons (6.7 ± 1.8 μgC m 3 to 7.7 ± 5.0 μgC m 3). Source profiles generated from using LG as a biomass burning tracer differed slightly from those using combined tracers of LG and nss-K+; however, they both differed significantly from those

Acknowledgments

This study was financially supported by the Special Scientific Research Funds for Environment Protection Commonwealth Section (No. 200809143), the National Natural Science Foundation of China (No. 41175131), and the Special Scientific Research Funds for National Basic Research Program of China (2013FY112700).

References (47)

  • Y. Ninomiya et al.

    Influence of coal particle size on particulate matter emission and its chemical species produced during coal combustion

    Fuel Process. Technol.

    (2004)
  • S.S. Park et al.

    Investigation of organic aerosol sources using fractionated water-soluble organic carbon measured at an urban site

    Atmos. Environ.

    (2012)
  • R.K. Pathak et al.

    Characteristics of summertime PM2.5 organic and elemental carbon in four major Chinese cities: implications of high acidity for water-soluble organic carbon (WSOC)

    Atmos. Environ.

    (2011)
  • O. Popovicheva et al.

    Physicochemical characterization of smoke aerosol during large-scale wildfires: extreme event of August 2010 in Moscow

    Atmos. Environ.

    (2014)
  • K. Ram et al.

    Day–night variability of EC, OC, WSOC and inorganic ions in urban environment of indo-gangetic plain: implications to secondary aerosol formation

    Atmos. Environ.

    (2011)
  • J. Tao et al.

    Chemical composition of PM2.5 in an urban environment in Chengdu, China: importance of springtime dust storms and biomass burning

    Atmos. Res.

    (2013)
  • M. Viana et al.

    Comparative analysis of organic and elemental carbon concentrations in carbonaceous aerosols in three European cities

    Atmos. Environ.

    (2007)
  • G. Wang et al.

    Characterization of water-soluble species of PM10 and PM2.5 aerosols in urban area in Nanjing, China

    Atmos. Environ.

    (2002)
  • Q. Wang et al.

    Impact of biomass burning on urban air quality estimated by organic tracers: Guangzhou and Beijing as cases

    Atmos. Environ.

    (2007)
  • J.G. Watson et al.

    PM2.5 chemical source profiles for vehicle exhaust, vegetative burning, geological material, and coal burning in Northwestern Colorado during 1995

    Chemosphere

    (2001)
  • T. Zhang et al.

    Water-soluble ions in atmospheric aerosols measured in Xi'an, China: seasonal variations and sources

    Atmos. Res.

    (2011)
  • Z. Zhang et al.

    Characteristics and applications of size-segregated biomass burning tracers in China's Pearl River Delta region

    Atmos. Environ.

    (2015)
  • L. Zhang et al.

    Measurements of reactive oxidized nitrogen at eight Canadian rural sites

    Atmos. Environ.

    (2008)
  • Cited by (85)

    View all citing articles on Scopus
    View full text