scispace - formally typeset
Search or ask a question
Journal ArticleDOI

A Comparison of Multiple Combined Models for Source Apportionment, Including the PCA/MLR-CMB, Unmix-CMB and PMF-CMB Models

01 Jan 2014-Aerosol and Air Quality Research (Taiwan Association for Aerosol Research)-Vol. 14, Iss: 7, pp 2040-2050
TL;DR: In this paper, multiple combined models, including the PCA/MLR-CMB, Unmix-cMB, and PMF-cmb models, were developed and employed to analyze the synthetic datasets, in order to understand 1) the accuracies of the predictions by multiple combined model; 2) the effect of Fpeak-rotation on the predictions of the PMF CMB model; and 3) the relationship between the extracted mixed source profiles (in the first stage) and the final predictions.
Abstract: A combined models was developed and applied to synthetic and ambient PM datasets in our prior works. In this study, multiple combined models, including the PCA/MLR-CMB, Unmix-CMB and PMF-CMB models, were developed and employed to analyzed the synthetic datasets, in order to understand 1) the accuracies of the predictions by multiple combined models; 2) the effect of Fpeak-rotation on the predictions of the PMF-CMB model; and 3) the relationship between the extracted mixed source profiles (in the first stage) and the final predictions. 50 predictions based on different combined model solutions were obtained and compared with the synthetic datasets. The average absolute errors (AAE), cluster analysis (CA), and PCA plots were applied to evaluate the precision of the predictions. These statistical methods showed that the predictions of the PCA/MLR-CMB and PMF-CMB model (with Fpeaks from 0 to 1.0) were satisfactory, those of the Unmix-CMB model were instable (some of them closely approached the synthetic values, while other them deviated from them). Additionally, it was found that the final source contributions had good correlation with their marker concentrations (obtained in the first stage), suggesting that the extracted profiles of the mixed sources can determine the final predictions of combined models.

Content maybe subject to copyright    Report

Citations
More filters
Journal ArticleDOI
TL;DR: The location of point pollution sources and prevailing wind direction were found to be important factors in the spatial distribution of heavy metals and there was significant enrichment of Pb, Zn, Co, Cu and Cr based on geo-accumulation index value.

231 citations

Journal ArticleDOI
TL;DR: Cluster and PSCF results indicated that local as well as long-transported PM2.5 from the north-west India and Pakistan were mostly pertinent, and re-confirmed that secondary aerosols, soil/road dust, vehicular emissions, biomass burning, fossil fuel combustion, and industrial emission were dominant contributors to PM 2.5 in Delhi.
Abstract: The present study investigated the comprehensive chemical composition [organic carbon (OC), elemental carbon (EC), water-soluble inorganic ionic components (WSICs), and major & trace elements] of particulate matter (PM2.5) and scrutinized their emission sources for urban region of Delhi. The 135 PM2.5 samples were collected from January 2013 to December 2014 and analyzed for chemical constituents for source apportionment study. The average concentration of PM2.5 was recorded as 121.9 ± 93.2 μg m−3 (range 25.1–429.8 μg m−3), whereas the total concentration of trace elements (Na, Ca, Mg, Al, S, Cl, K, Cr, Si, Ti, As, Br, Pb, Fe, Zn, and Mn) was accounted for ∼17% of PM2.5. Strong seasonal variation was observed in PM2.5 mass concentration and its chemical composition with maxima during winter and minima during monsoon seasons. The chemical composition of the PM2.5 was reconstructed using IMPROVE equation, which was observed to be in good agreement with the gravimetric mass. Source apportionment of PM2.5 was carried out using the following three different receptor models: principal component analysis with absolute principal component scores (PCA/APCS), which identified five major sources; UNMIX which identified four major sources; and positive matrix factorization (PMF), which explored seven major sources. The applied models were able to identify the major sources contributing to the PM2.5 and re-confirmed that secondary aerosols (SAs), soil/road dust (SD), vehicular emissions (VEs), biomass burning (BB), fossil fuel combustion (FFC), and industrial emission (IE) were dominant contributors to PM2.5 in Delhi. The influences of local and regional sources were also explored using 5-day backward air mass trajectory analysis, cluster analysis, and potential source contribution function (PSCF). Cluster and PSCF results indicated that local as well as long-transported PM2.5 from the north-west India and Pakistan were mostly pertinent.

110 citations


Cites result from "A Comparison of Multiple Combined M..."

  • ...…al. 2007; Gildemeister et al. 2007; Zheng et al. 2007; Olson and Norris 2008; Begum et al. 2010; Harrison et al. 2011; Gugamesetty et al. 2012; Wang et al. 2012; Lelpo et al. 2014; Shi et al. 2014) and our previous publications (Sharma et al. 2014b, c; Sharma et al. 2015; Sharma et al. 2016a, b)....

    [...]

Journal ArticleDOI
TL;DR: In this article, the authors conducted a measurement of volatile organic compounds (VOCs) during November 2014 in the suburban area of Beijing, China covering the period of the Asia-Pacific Economic Cooperation (APEC) meeting period.

104 citations

Journal ArticleDOI
TL;DR: In this paper, the authors used principal component analysis combining multiple linear regression (PCA-MLR), UNMIX and Positive Matrix Factorization (PMF) to identify the sources of PM2.5.

70 citations

References
More filters
Journal ArticleDOI
TL;DR: In this paper, a new variant of Factor Analysis (PMF) is described, where the problem is solved in the weighted least squares sense: G and F are determined so that the Frobenius norm of E divided (element-by-element) by σ is minimized.
Abstract: A new variant ‘PMF’ of factor analysis is described. It is assumed that X is a matrix of observed data and σ is the known matrix of standard deviations of elements of X. Both X and σ are of dimensions n × m. The method solves the bilinear matrix problem X = GF + E where G is the unknown left hand factor matrix (scores) of dimensions n × p, F is the unknown right hand factor matrix (loadings) of dimensions p × m, and E is the matrix of residuals. The problem is solved in the weighted least squares sense: G and F are determined so that the Frobenius norm of E divided (element-by-element) by σ is minimized. Furthermore, the solution is constrained so that all the elements of G and F are required to be non-negative. It is shown that the solutions by PMF are usually different from any solutions produced by the customary factor analysis (FA, i.e. principal component analysis (PCA) followed by rotations). Usually PMF produces a better fit to the data than FA. Also, the result of PF is guaranteed to be non-negative, while the result of FA often cannot be rotated so that all negative entries would be eliminated. Different possible application areas of the new method are briefly discussed. In environmental data, the error estimates of data can be widely varying and non-negativity is often an essential feature of the underlying models. Thus it is concluded that PMF is better suited than FA or PCA in many environmental applications. Examples of successful applications of PMF are shown in companion papers.

4,797 citations


"A Comparison of Multiple Combined M..." refers methods in this paper

  • ...The detailed introductions of the principle and applications for CMB, PCA/MLR, Unmix and PMF models have been presented in literature (Watson, 1984; Paatero and Tapper, 1994; Lee et al., 1999; Watson and Chow, 2001; Song et al., 2006; Chen et al., 2007; Zheng et al., 2007; Begum et al., 2010;…...

    [...]

  • ...The detailed introductions of the principle and applications for CMB, PCA/MLR, Unmix and PMF models have been presented in literature (Watson, 1984; Paatero and Tapper, 1994; Lee et al., 1999; Watson and Chow, 2001; Song et al., 2006; Chen et al., 2007; Zheng et al., 2007; Begum et al., 2010; Harrison et al., 2011; Gugamsetty et al., 2012) and our prior publications (Shi et al....

    [...]

Journal ArticleDOI
TL;DR: In this paper, the 24-hour PM2.5 samples were taken at 6-day intervals at five urban and rural sites simultaneously in Beijing, China for 1 month in each quarter of calendar year 2000.

541 citations


"A Comparison of Multiple Combined M..." refers background in this paper

  • ...In order to reduce the PM pollution, understanding the potential source categories and their contributions (source apportionment) is necessary (Zheng et al., 2005)....

    [...]

Journal ArticleDOI
TL;DR: In this article, an advanced algorithm called positive matrix factorization (PMF) in receptor modeling was used to identify the sources of respirable suspended particulates (RSP) in Hong Kong.

441 citations


"A Comparison of Multiple Combined M..." refers methods in this paper

  • ...The detailed introductions of the principle and applications for CMB, PCA/MLR, Unmix and PMF models have been presented in literature (Watson, 1984; Paatero and Tapper, 1994; Lee et al., 1999; Watson and Chow, 2001; Song et al., 2006; Chen et al., 2007; Zheng et al., 2007; Begum et al., 2010; Harrison et al., 2011; Gugamsetty et al., 2012) and our prior publications (Shi et al....

    [...]

  • ...…of the principle and applications for CMB, PCA/MLR, Unmix and PMF models have been presented in literature (Watson, 1984; Paatero and Tapper, 1994; Lee et al., 1999; Watson and Chow, 2001; Song et al., 2006; Chen et al., 2007; Zheng et al., 2007; Begum et al., 2010; Harrison et al., 2011;…...

    [...]

Journal ArticleDOI
TL;DR: In this paper, the authors analyzed 52-year historical surface measurements of haze data in the Chinese city of Guangzhou, and showed that the dramatic increase in the occurrence of air pollution events between 1954 and 2006 has been followed by a large enhancement in the incidence of lung cancer.

293 citations


"A Comparison of Multiple Combined M..." refers background in this paper

  • ...Long-term exposure to particulate matter air pollution can result in increased risk of human mortality (Ozkaynak and Thurston, 1987; Russell, 2009; Yan et al., 2009; Tie et al., 2009; Habre et al., 2011; Shen et al., 2011; Cheng et al., 2012; Amodio et al., 201; Vernile et al., 2013)....

    [...]

Journal ArticleDOI
TL;DR: In this paper, the authors present a review of the improvements in the factor analysis methods that are applied in receptor modeling as well as easier application of trajectory methods for airborne particulate matter detection.
Abstract: Receptor modeling is the application of data analysis methods to elicit information on the sources of air pollutants. Typically, it employs methods of solving the mixture resolution problem using chemical composition data for airborne particulate matter samples. In such cases, the outcome is the identification of the pollution source types and estimates of the contribution of each source type to the observed concentrations. It can also involve efforts to identify the locations of the sources through the use of ensembles of air parcel back trajectories. In recent years, there have been improvements in the factor analysis methods that are applied in receptor modeling as well as easier application of trajectory methods. These developments are reviewed. Copyright © 2003 John Wiley & Sons, Ltd.

292 citations


"A Comparison of Multiple Combined M..." refers background in this paper

  • ...Among several receptor models, two main classes of models have been employed widely over the world (Hopke, 2003; Andriani et al., 2011; Pant and Harrison, 2012)....

    [...]

  • ...The strengths and weaknesses for the two classes of receptor models have been summarized in literature (Hopke, 2003; Pant and Harrison, 2012)....

    [...]

  • ...The first class of models need both the input data of receptor and the source profiles; while the later class of models extracts source profiles and their contributions over sets of receptor samples (Hopke, 2003)....

    [...]