理乐志

chihchao's blog

0%

文献阅读笔记：Srivastava-2021-ACP

发表于 分类于

正矩阵分解（PMF）解析北京城市和乡村PM2.5来源。

Information for the paper

Title: Insight into PM2.5 sources by applying positive matrix factorization (PMF) at urban and rural sites of Beijing

Author: Deepchandra Srivastava

Year: 2021

Journal: Atmospheric Chemistry and Physics

URL: https://doi.org/10.5194/acp-21-14703-2021

Introduction

  1. Receptor models are widely used for source apportionment of PM2.5. These methods include positive matrix factorization (PMF), principal component analysis (PCA), chemical mass balance (CMB), and UNMIX.
  2. These source apportionment studies have predominantly used OC (organic carbon), EC (elemental carbon), water-soluble ions, and metals as the input data matrix to explore the co-variances between species and their associated sources, but to the best of our knowledge, the use of organic markers in PMF has not been explored extensively in Beijing. The use of organic molecular markers in PMF has enhanced our understanding of the PM fraction as they can be source specific and could potentially offer a clearer link between factors and sources.
  3. In addition, one or two punches of PM2.5 filter sample were also analysed offline using AMS to investigate the water-soluble OA (WSOA) mass spectra following the procedure explained previously (Qiu et al., 2019).

Overview of PM sources in Beijing based on the current source apportionment study

  1. Based on the factor profiles, we identified traffic emissions, biomass burning, road dust, soil dust, coal combustion, oil combustion, and secondary inorganics.
  2. Overall, secondary inorganics, biomass burning, and coal combustion sources were the main contributors to the total PM2.5 mass during winter.
  3. Secondary inorganics, road dust, and coal combustion showed the highest contribution during summer at PG, while PM2.5 particles were mainly composed of soil dust and secondary inorganics at IAP.
PMF

Coal combustion

  1. Coal combustion was identified based on it accounting for a high proportion of PAHs (27%–78%), especially picene (78%) as a specific marker of coal combustion, together with significant amounts of OC (45 %) and EC (29 %).
  2. This factor also made a substantial contribution to n-alkanes (28%–58%), stearic acid (64%), and hopanes (53%–56%), as these compounds are also abundant in coal smoke.
  3. Due to a lack of infrastructure at the rural site PG, the residents still used coal for cooking and heating purposes at the time of sampling.
  4. Levoglucosan, a major pyrolysis product of cellulose, has been proposed as a molecular marker of biomass burning aerosols (Simoneit, 1999).
  5. A study conducted in China suggested that residential coal combustion can also contribute significantly to levoglucosan emissions, based on both source testing and ambient measurements (Yan et al., 2018).
  6. High concentrations of this source and levoglucosan were observed at low wind speeds, indicating the significant role of local activities.

Oil combustion

  1. V and Ni are widely used markers for oil combustion in residential, commercial, and industrial applications.
  2. The V/Ni ratio obtained in this study was 0.9, close to the previously obtained ratio for residual oil used in power plants.
  3. Results suggest this source might be attributed to residual oil combustion linked to industrial activities as a large number of highly polluting industries are still located in the Beijing neighbourhood.

Biomass burning

  1. The biomass burning factor was characterized by high contributions to Cl- (74%), K+ (27%), and levoglucosan (25%). This factor also made significant contributions to PAHs and followed a clear seasonal variation with a higher contribution in winter.
  2. It accounted for 36% and 30% of the PM2.5 mass during the wintertime at IAP and PG, respectively, while the contribution during the summertime was extremely low.
  3. Biomass burning is an important natural source of NH3 (Zhou et al., 2020) which rapidly reacts with HNO3 to form NH4NO3 aerosols. The presence of NH4NO3 aerosols in biomass burning plumes has also been reported previously (Paulot et al., 2017; Zhao et al., 2020).
  4. Cl- can be emitted from both coal combustion and biomass burning, especially during the cold period in Beijing (Sun et al., 2006).

Secondary inorganics

  1. This factor showed a temporal variation, with remarkably high concentrations observed during the period of high RH and low ozone concentration in the winter.
  2. The heterogeneous reactions on pre-existing particles in the polluted environment under high-RH and low-ozone conditions have been shown to play a key role in the formation of secondary aerosols compared to gas-phase photochemical processes (Sun et al., 2013; Niu et al., 2016; Ma et al., 2017b). Therefore, aqueous-phase processes may be the major formation pathway for secondary inorganic aerosols in Beijing during the study period.

Traffic emmisions

  1. The traffic emissions factor showed relatively high contributions to metallic elements, such as Zn (47%), Pb (57%), Mn (27%), and Fe (22%).
  2. Zn is a major additive to lubricant oil. Zn and Fe can also originate from tyre abrasion, brake linings, lubricants, and corrosion of vehicular parts and tailpipe emission (Pant and Harrison, 2012, 2013; Grigoratos and Martini, 2015; Piscitello et al., 2021).
  3. As the use of Pb additives in gasoline has been banned since 1997 in Beijing, the observed Pb emissions may be associated with wear (tyre/brake) rather than fuel combustion (Smichowski et al., 2007).
  4. We noticed that OC and EC contribution in this factor is relatively low, while it may be higher in traffic emissions. However, given the modern gasoline fleet in Beijing (Jing et al., 2016), it is not unexpected to observe low OC and EC contributions.
  5. Metallic elements such as Mn, Fe, and Zn were also used previously to indicate industrial activities.
  6. These observations suggest that indeed this factor contains traffic aerosols, though a significant influence of industrial emissions cannot be ruled out.

Road dust

  1. This factor makes a major contribution to crustal species, such as Na+, Al, and Fe (60%, 48%, and 34% of species in this factor, respectively), suggesting this factor may represent the characteristics of a dust-related source as reported previously.
  2. Na is a major element of sea salt, sea spray, and marine aerosols (Viana et al., 2008) and has also been found to be enriched in fine particulates from coal combustion (Takuwa et al., 2006). However, the significant influence of marine activities was not expected as Beijing is far away from the sea.
  3. High concentrations of Zn and Pb have also been reported for particles emitted from asphalt pavement.
  4. In addition, the ratio of Fe/Al observed in the factor chemical profile was 1.26, much higher than the value observed in the earth’s crust (0.6), suggesting an anthropogenic origin of some Fe (Sun et al., 2005).
  5. These metals (Fe and Al) can also have industrial sources as already reported in the Beijing area.
  6. Fe is a characteristic component of iron and steel industry emissions (Li et al., 2019), while Al may also come from metal processing (Yu et al., 2013). However, disentangling the influence of industrial emissions would require further investigation.

Soil dust

  1. This factor mainly represents wind-blown soils and was typically characterized by a high contribution to crustal elements, such as Ti (63%), Ca2+ (41 %), Fe (27 %), and Al (17 %).
  2. In addition, the contributions to Mn and Zn in the factor profile suggest that the given source also included resuspended road dust but probably to a lesser extent.
  3. In addition, other previous studies (Yu et al., 2013; Zhang et al., 2013) also reported a significant contribution of soil dust to PM2.5 mass, suggesting that soil dust is an important contributor to PM2:5 mass in the Beijing area.
  4. It is also expected as Beijing is in a semi-arid region and there are sparsely vegetated surfaces both within and outside the city.

Comparison of filter-based PMF results with other receptor modelling approaches on the same dataset

  1. This suggests that the biomass burning factor in PMF may contain a substantial amount of aged aerosols since carbon emitted during biomass burning is in some cases oxygenated and water soluble (Lee et al., 2008b) and is subject to rapid oxidation in the atmosphere.
  2. The formation of both secondary inorganic aerosol and oxygenated organic aerosol is dependent upon largely the same set of oxidant species, notably but not solely the hydroxyl and nitrate radicals.
  3. In both cases there are also both homogeneous and heterogeneous (aqueous-phase) pathways, so conditions which promote the formation of oxidized organic aerosol will also favour formation of secondary inorganic aerosol, and hence a correlation is to be expected and is often observed (Hu et al., 2016; Zhang et al., 2011).
  4. In addition, both oxygenated fractions were also found to be correlated with biomass burning aerosols. This further highlights a potentially important role of biomass burning activity in SOA formation at IAP.
  5. High Cl associated with fine aerosols in winter is a distinctive feature in Beijing and even around inland China, which is ascribed to coal combustion (Wang et al., 2008).
  6. For example, Zn, Cu, and Pb including sometimes EC were most often used to characterize traffic emissions among all previous studies.
  7. Thus, it is clear that these metals could belong to several sources, and their proper assignment to respective sources is difficult in the complex environment.

总结与思考

  1. 作者使用有机标记物对正矩阵分解（PMF）源解析的方法进行改进
  2. 离线采样的滤膜同样可以用在气溶胶质谱仪的分析中。
  3. 左旋葡聚糖一般被认为是生物质燃烧的特征物，然而在中国进行的最新研究表明，居民燃煤同样会排放左旋葡聚糖。
  4. 高湿度、低臭氧浓度的条件下，在已有颗粒物上发生的多相反应比气相光化学反应在次生气溶胶形成过程中扮演了更重要的角色。

扩展阅读