2024-03-28T16:23:45Zhttp://digital.csic.es/dspace-oai/requestoai:digital.csic.es:10261/631752021-12-15T12:59:37Zcom_10261_35com_10261_5col_10261_288
Comparison of receptor models for source apportionment of the PM10 in Zaragoza (Spain)
Callén Romero, Mª Soledad
Cruz Eiriz, M. T. de la
López Sebastián, José Manuel
Navarro López, María Victoria
Mastral Lamarca, Ana María
Gobierno de Aragón
Consejo Superior de Investigaciones Científicas (España)
PM10
Receptor modelling
PCA-APCS
Unmix
PMF
Receptor models are useful to understand the chemical and physical characteristics of air pollutants by identifying their sources and by estimating contributions of each source
to receptor concentrations. In this work, three receptor models based on principal component analysis with absolute principal component scores (PCA-APCS), Unmix and Positive Matrix Factorization (PMF) were applied to study for the first time the
apportionment of the airborne particulate matter less or equal than 10 m (PM10) in Zaragoza, Spain, during one year sampling campaign (2003-2004). The PM10 samples were characterized regarding their concentrations in inorganic components: trace elements and ions and also organic components: polycyclic aromatic hydrocarbons (PAH) not only in the solid phase but also in the gas phase. A comparison of the three receptor models was carried out in order to do a more robust characterization of the PM10. The three models predicted that the major sources of PM10 in Zaragoza were related to natural sources (60%, 75% and 47% respectively for PCA-APCS, Unmix and PMF) although anthropogenic sources also contributed to PM10 (28%, 25% and 39%). With regard to the anthropogenic sources, while PCA and PMF allowed high discrimination in the sources identification associated with different combustion sources such as traffic and industry, fossil fuel, biomass and fuel oil combustion, heavy traffic and evaporative emissions, the Unmix model only allowed the identification of industry and traffic emissions, evaporative emissions and heavyduty vehicles. The three models provided good correlations between the experimental
and modelled PM10 concentrations with major precision and the closest agreement between the PMF and PCA models.
Authors would like to thank Aula Dei-CSIC (R. Gracia) for providing the meteorological data, the Government of Aragon (DGA) for the grant to M.T.C, the CSIC for the JAE post-doct contract to J.M.L and the Spanish Government for the
Ramón y Cajal contract to M.V.N.
Peer reviewed
2012-12-18T13:05:36Z
2012-12-18T13:05:36Z
2009-05-13
artículo
http://purl.org/coar/resource_type/c_6501
Chemosphere 76(8): 1120-1129 (2009)
0045-6535
http://hdl.handle.net/10261/63175
10.1016/j.chemosphere.2009.04.015
http://dx.doi.org/10.13039/501100003339
http://dx.doi.org/10.13039/501100010067
en
Postprint
http://dx.doi.org/10.1016/j.chemosphere.2009.04.015
open
Elsevier