Please use this identifier to cite or link to this item:
|Title:||Sources for PM air pollution in the Po Plain, Italy: I. Critical comparison of methods for estimating biomass burning contributions to benzo(a)pyrene|
|Authors:||BELIS CLAUDIO; CANCELINHA Jose; DUANE Matthew; FORCINA Vittorio; PEDRONI Valerio; PASSARELLA Rosanna; TANET Gerard; DOUGLAS Kevin; PIAZZALUNGA Andrea; BOLZACCHINI Ezio; SANGIORGI Giorgia; PERRONE Maria Grazia; FERRERO L.; FERMO Paola; LARSEN Bo|
|Citation:||ATMOSPHERIC ENVIRONMENT vol. 45 no. 39 p. 7266-7275|
|Publisher:||PERGAMON-ELSEVIER SCIENCE LTD|
|Type:||Articles in Journals|
|Abstract:||Particle-bound benzo(a)pyrene (B(a)P) constitutes an air pollution problem in many areas of Europe and has been linked to biomass burning (BB). The present study, conducted in 2007 and 2009 at ten stations in the North Italian Po Plain and Valtelline Valley, examines four methods for the quantification of BB contributions to particle-bound B(a)P using data for 61 predictor compounds in more than 700 ambient PM10 and PM2.5 samples. The study was carried out during the heating season – a period of the year with minimal volatilization and atmospheric degradation of B(a)P, which favour source apportionment by receptor modelling. The lowest estimates of the source contribution (SCE) from BB were obtained with the levoglucosan tracer method and multi-linear regression analysis of daily variations in B(a)P concentrations using levoglucosan as the main predictor in combination with a few other predictors including gaseous pollutants and meteorological data. The standard uncertainty of these methods was driven by the uncertainty in the BB emission factor for levoglucosan and mounted to 90% (1 σ). Positive matrix factorization (PMF), using only PAH congeners as predictors, did not produce factors interpretable as emission sources. However, PMF utilizing a broad range of predictor compounds afforded five factors with compositions similar to emission sources. The yielded B(a)P SCEs for BB agreed well with results of chemical mass balance modelling (CMB). Both receptor models gave good predictions (p) of the observed (o) B(a)P concentrations (PMF: p/o = 89 ± 9%, CMB: p/o = 114 ± 17%) with lower uncertainties than the tracer methods (CMB 60%; PMF 54%; 1 σ). The average BB SCEs (mean ± 95% confidence interval) from these models were: 1.0 ± 0.4 ng m−3 at a kerbside in Milan, 1.0 ± 0.2 ng m−3 at six urban background stations in the Po Plain, 0.7 ± 0.3 ng m−3 at two rural background stations in the Po Plain, and 2.1 ± 1.1 ng m−3 at an urban background station in the Valtelline Valley representing 74 ± 32%, 79 ± 18%, 85 ± 33%, and 84 ± 46% of all modelled B(a)P sources, respectively.|
|JRC Institute:||Institute for Environment and Sustainability|
Files in This Item:
There are no files associated with this item.
Items in repository are protected by copyright, with all rights reserved, unless otherwise indicated.