Estimating primary vehicular emission contributions to PM2.5 using the Chemical Mass Balance model: Accounting for gas-particle partitioning of organic aerosols and oxidation degradation of hopanes☆
Introduction
Mitigation of airborne fine particulate matter (PM2.5) through controlling vehicular emissions (VE) has shown to be effective in curbing PM2.5 pollution in many urban areas. Despite the substantial progress, it remains important to sustain the VE control effort for several reasons, including the demonstrated health impacts of PM2.5 even at low concentrations (Di et al., 2017; Liu et al., 2019), increasing population residing in urban areas (United Nations, 2018), and a growing number of on-road vehicles in the coming decades (Frey, 2018). The dependence on motor vehicles calls for a continuous and reliable assessment of vehicular PM2.5 (PMvehicle) contributions, which is essential for further control of air quality with timely strategies.
A growing body of atmospheric research has been using receptor modeling to quantify the contributions to PM2.5 loading from primary sources, including VE. Among the various receptor modeling approaches, the Chemical Mass Balance (CMB) and Positive Matrix Factorization (PMF) models coupled with highly source-specific species (e.g. organic tracers) have gained wide utility for their ability to resolve relatively more detailed and robust source contribution estimates (SCEs) (e.g. Schauer et al., 1996; Wang et al., 2017). In addition to elemental carbon (EC), hopanes and polycyclic aromatic hydrocarbons (PAHs) are two potent tracer groups for tracking VE contribution, but the use of PAHs is often complicated by the presence of multiple combustion sources. CMB drawing on EC and hopanes has received broader recognition by its effectiveness to obtain separate SCEs for diesel and gasoline VE (e.g. Subramanian et al., 2006; Wong et al., 2019a; Zheng et al., 2006). PMF is less effective in this regard due to the limit posed by the similarity in temporal emission patterns of the two vehicle types (Shrivastava et al., 2007; Wang et al., 2017). The ability to resolve vehicle type-specific SCEs by CMB is a significant advantage, especially for formulating vehicle control measures.
In CMB, the SCEs can be calculated on a sample-by-sample basis as opposed to PMF, which requires a large data set for model execution. While this feature is conducive to acquiring source contribution information in the face of sample size constraint, CMB-derived SCEs are sensitive to the weight fractions of source tracers (e.g. hopane-to-PM2.5 ratio) in input source profiles (Subramanian et al., 2006). Such a strong dependency presents great challenge to achieving meaningful SCEs, and highlights the importance of using representative profiles. By the same logic, any modification of source profiles by atmospheric processing should be accounted for in CMB. Multiple lines of evidence have shown that source profiles would undergo modification during atmospheric transport, however, the impacts on CMB-derived SCEs have remained largely uncharacterized.
Photooxidation degradation of hopanes could play a key role in modifying VE source profiles (Lambe et al., 2009; Robinson et al., 2006; Wong et al., 2019b). It has been proposed gasoline PMvehicle contribution could be underestimated substantially in CMB due to omission of this process (Weitkamp et al., 2008; Yu et al., 2011). This issue was addressed in our previous work, where the oxidation of individual hopane homologues were incorporated in CMB by adopting a set of homologue-specific rate constants to obtain an optimized solution (Wong et al., 2019b). The results showed that neglecting the oxidation process would underestimate CMB-derived PMvehicle by half.
One additional source of uncertainty to the CMB-apportioned PMvehicle is gas-particle partitioning of organic aerosols (OA). PMvehicle is primarily composed of EC and OA (Chow et al., 2011; Kleeman et al., 2000). The latter are largely semivolatile and thus their mass varies continuously with ambient temperature and OA concentration (Robinson et al., 2010). Shifting the paradigm from treating primarily emitted OA (POA) as nonvolatile to semivolatile, Robinson et al. (2007) significantly improved the description of vehicular OA contributions by 3D-air quality model. By logic extension, such an updated framework treating POA needs to be considered in receptor modeling as well. Xue et al. (2019) introduced an approach to account for the gas-particle partitioning of POA and the oxidative loss of organic tracers (including tracers for VE, biomass burning, and cooking emissions, etc.) in CMB. The technique was applied to apportion the ultrafine particles in California cities.
In this work, we focus on the apportioning of PMvehicle by CMB and for the first time integrate the approach of Xue et al. (2019) for considering gas-particle partitioning and the approach of Wong et al. (2019b) for treating hopane oxidation. The objective of this work is to present how the aggregate consideration of gas-particle partitioning of POA and hopane oxidation would affect PMvehicle estimation by CMB. CMB analyses are conducted on PM2.5 samples collected from an urban roadside site and a general urban site in Hong Kong. Separate diesel and gasoline VE profiles were developed from the roadside measurements for the CMB modeling. The revised CMB methodology is evaluated by comparing the roadside results with recent source apportionment studies conducted at the same location. Then, the method is extended to the general urban site with the focus to examine the impacts of ambient temperature and OA loading on CMB-apportioned PMvehicle. Finally, the limitations of this work and ways toward improving PMvehicle estimation by CMB are discussed.
Section snippets
Data sets for source apportionment
Two ambient data sets of speciated PM2.5 were analyzed by CMB in this work. The first data set is derived from an ad hoc air quality monitoring campaign at the Mong Kok Air Quality Monitoring Station (MK AQMS). The campaign was intended for characterizing the influence of cooking emissions on local air quality (Cheng & Yu, 2020). Yet, the strong influence of VE provides an opportunity to derive representative VE source profiles for incorporation into CMB analysis. Furthermore, vehicle
Ambient characteristics of EC and hopanes
Table S7 lists the concentration statistics in different diel time segments for PM2.5 and their major components (OC, EC, sulfate, nitrate and ammonium), and C29–C30 hopanes during the field campaigns, providing an overview of the general pollution characteristics. Hong Kong is a coastal city facing the South China Sea to the south and mainland China to the north. Generally, local emissions are the predominant sources in summer due to the prevailing southerlies carrying pristine air mass from
Conclusions
In this study, we demonstrated an implementation to incorporate gas-particle partitioning of OA and oxidation degradation of hopanes into CMB analysis to improve the accuracy of PMvehicle source apportionment. Two speciated data sets from Hong Kong were analyzed, with the first collected in summer from the MK AQMS, a downtown roadside station and the second set in summer and winter from the YL AQMS, a general station located in a residential district. Clear evidence for atmospheric hopane
Author statement
Yee Ka Wong: Conceptualization, Methodology, Formal analysis, Writing – original draft. X. H. Hilda Huang: Conceptualization, Writing – review & editing. Yuk Ying Cheng: Investigation. Jian Zhen Yu: Conceptualization, Writing – review & editing, Supervision.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments and Data
This work is supported by the Hong Kong Environmental Protection Department (HKEPD) (tender ref. 14–05593) and Hong Kong Research Grants Council (16305418 and R6011-18). Chemical composition data used for the CMB and PMF analyses in this study can be downloaded from https://doi.org/10.14711/dataset/RHRN4N.
References (41)
- et al.
Chemically-speciated on-road PM2.5 motor vehicle emission factors in Hong Kong
Sci. Total Environ.
(2010) - et al.
PM2.5 source profiles for black and organic carbon emission inventories
Atmos. Environ.
(2011) - et al.
In-injection port thermal desorption and subsequent gas chromatography-mass spectrometric analysis of polycyclic aromatic hydrocarbons and n-alkanes in atmospheric aerosol samples
J. Chromatogr. A
(2004) - et al.
Seasonal characteristics and regional transport of PM2.5 in Hong Kong
Atmos. Environ.
(2005) - et al.
Gas-particle partitioning of primary organic aerosol emissions: (1) Gasoline vehicle exhaust
Atmos. Environ.
(2013) - et al.
Source apportionment of airborne particulate matter using organic compounds as tracers
Atmos. Environ.
(1996) - et al.
Sources of organic aerosol: positive matrix factorization of molecular marker data and comparison of results from different source apportionment models
Atmos. Environ.
(2007) - et al.
Contribution of motor vehicle emissions to organic carbon and fine particle mass in Pittsburgh, Pennsylvania: effects of varying source profiles and seasonal trends in ambient marker concentrations
Atmos. Environ.
(2006) - et al.
The effective variance weighting for least squares calculations applied to the mass balance receptor model
Atmos. Environ.
(1984) - et al.
Estimating contributions of vehicular emissions to PM2.5 in a roadside environment: a multiple approach study
Sci. Total Environ.
(2019)
Incorporating hopane degradation into chemical mass balance model: improving accuracy of vehicular source contribution estimation
Atmos. Environ.
Abundance and seasonal characteristics of elemental and organic carbon in Hong Kong PM10
Atmos. Environ.
Minimizing contamination from plastic labware in the quantification of C16 and C18 fatty acids in filter samples of atmospheric particulate matter and their utility in apportioning cooking source contribution to urban PM2.5
Atmosphere
Characterization of particulate matter emissions from on-road gasoline and diesel vehicles using a soot particle aerosol mass spectrometer
Atmos. Chem. Phys.
Air pollution and mortality in the Medicare population
N. Engl. J. Med.
Trends in onroad transportation energy and emissions
J. Air Waste Manag. Assoc.
Atmospheric stability of levoglucosan: a detailed laboratory and modeling study
Environ. Sci. Technol.
Contribution of vehicular carbonaceous aerosols to PM2.5 in a roadside environment in Hong Kong
Atmos. Chem. Phys.
Size and composition distribution of fine particulate matter emitted from motor vehicles
Environ. Sci. Technol.
Effective rate constants and uptake coefficients for the reactions of organic molecular markers (n-alkane, hopanes, and steranes) in motor oil and diesel primary organic aerosols with hydroxyl radicals
Environ. Sci. Technol.
Cited by (8)
PM2.5 source apportionment using organic marker-based chemical mass balance modeling: Influence of inorganic markers and sensitivity to source profiles
2023, Atmospheric EnvironmentCitation Excerpt :A detailed intercomparison using data from Beijing concluded that CMB gave the most detailed and plausible results compared to PMF (Xu et al., 2021b). The CMB has been used for source apportionment of PM at many locations, worldwide (Zheng et al., 2002; Perrone et al., 2012; Yin et al., 2015; Wu et al., 2020; Wong et al., 2021). Multicollinearity, arising when two different sources have similar profiles, often disturbs the estimation of the CMB modeling.
Eliminating collinearity observed in chemical mass balance analysis using multi linear regression and synthetic receptor source profile dataset
2024, Sadhana - Academy Proceedings in Engineering SciencesAccurate traceability of air pollution in AQP area for multi-source data
2023, Huanjing Kexue Xuebao/Acta Scientiae CircumstantiaeImplementation of air pollution traceability method based on IF-GNN-FC model with multiple-source data
2023, International Journal of Data Science and Analytics
- ☆
This paper has been recommended for acceptance by Pavlos Kassomenos.