A new methodology to assess the performance and uncertainty of source apportionment models in intercomparison exercises

A new methodology to assess source apportionment model performance in intercomparison exercises, encompassing the preparation of real-world and synthetic datasets and the evaluation of the source apportionment results reported by participants, is described. The evaluation consists of three types of tests: complementary tests, preliminary tests, and performance tests. The complementary tests provide summary information about the source apportionment results as a whole. The preliminary tests check whether source/factors belong to a given source category. Three types of indicators: Pearson correlation (Pearson), standardized identity distance (SID), and weighted difference (WD) are used to test factor/ source chemical profiles, while factor/source time series and contribution-to-species values are tested only using the Pearson. The performance tests, based on international standards for proficiency testing, are targeted at evaluating whether the reported biases in the quantification of the factor/source contribution estimates (SCEs) and uncertainties are consistent with previously established quality standards in a fitness-for-purpose approach. Moreover, the consistency of the SCE time series is evaluated using a variant of the RMSE normalised by the reference standard uncertainty. The described methodology facilitates a thorough evaluation of the source apportionment output. The new indicator to compare source or factor profiles presented in this study (SID) is more robust and provides additional information compared to the existing ones. © 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license

[1]  P. Paatero,et al.  PM source apportionment and health effects: 1. Intercomparison of source apportionment results , 2006, Journal of Exposure Science and Environmental Epidemiology.

[2]  Michael Thompson,et al.  The International Harmonized Protocol for the proficiency testing of analytical chemistry laboratories (IUPAC Technical Report) , 2006 .

[3]  Anna Pederzoli,et al.  Performance criteria to evaluate air quality modeling applications , 2012 .

[4]  Ralph Morris,et al.  Particulate Matter Source Apportionment Technology (PSAT) in the CAMx Photochemical Grid Model , 2007 .

[5]  P. Paatero,et al.  Understanding and controlling rotations in factor analytic models , 2002 .

[6]  Peter A. Scheff,et al.  Fine particulate source apportionment using data from the USEPA speciation trends network in Chicago, Illinois: Comparison of two source apportionment models , 2007 .

[7]  Robert K. Stevens,et al.  Overview of the Mathematical and Empirical Receptor Models Workshop (Quail Roost II) , 1984 .

[8]  G. N. Lance,et al.  Computer Programs for Hierarchical Polythetic Classification ("Similarity Analyses") , 1966, Comput. J..

[9]  B. Claudio,et al.  European guide on air pollution source apportionment with receptor models , 2013 .

[10]  M. Thompson,et al.  The international harmonized protocol for the proficiency testing of (chemical) analytical laboratories (Technical Report) , 1993 .

[11]  C Belis,et al.  Enhancing source apportionment with receptor models to foster the air quality directive implementation , 2012 .

[12]  S. Friedlander,et al.  Chemical element balances and identification of air pollution sources. , 1973, Environmental science & technology.

[13]  Gary A. Norris,et al.  Methods for estimating uncertainty in factor analytic solutions , 2013 .

[14]  M. Viana,et al.  Inter-comparison of receptor models for PM source apportionment: Case study in an industrial area , 2008 .

[15]  Philip K. Hopke,et al.  The Application of Receptor Modeling to Air Quality Data , 2010 .

[16]  Judith C. Chow,et al.  Source Apportionment: Findings from the U.S. Supersites Program , 2008, Journal of the Air & Waste Management Association.

[17]  James J. Schauer,et al.  Source apportionment of airborne particulate matter using organic compounds as tracers , 1996 .

[18]  Christian George,et al.  Inter-comparison of source apportionment models for the estimation of wood burning aerosols during wintertime in an Alpine city (Grenoble, France) , 2010 .

[19]  C Belis,et al.  Determination of Local Source Profile for Soil Dust, Brake Dust and Biomass Burning Sources , 2010 .

[20]  P. Hopke,et al.  Source apportionment of particulate matter in Europe: A review of methods and results , 2008 .

[21]  Sönke Szidat,et al.  Comparison of several wood smoke markers and source apportionment methods for wood burning particulate mass , 2008 .

[22]  Philip K. Hopke,et al.  Source identification of airborne PM2.5 at the St. Louis‐Midwest Supersite , 2006 .

[23]  Philip K. Hopke,et al.  Critical review and meta-analysis of ambient particulate matter source apportionment using receptor models in Europe , 2013 .

[24]  Min-Suk Bae,et al.  Positive matrix factorization (PMF) analysis of molecular marker measurements to quantify the sources of organic aerosols. , 2007, Environmental science & technology.

[25]  Philip K. Hopke,et al.  Rotational tools for factor analytic models , 2009 .

[26]  J. Kindle,et al.  Summary diagrams for coupled hydrodynamic-ecosystem model skill assessment , 2009 .

[27]  C Belis,et al.  Sources for PM air pollution in the Po Plain, Italy: II. Probabilistic uncertainty characterization and sensitivity analysis of secondary and primary sources , 2012 .

[28]  Gary A. Norris,et al.  The Krakow receptor modelling inter-comparison exercise , 2008 .