Pollution source apportionment using a priori information and positive matrix factorization

Lingwall, Jeff; Christensen, William F.

doi:10.1016/j.chemolab.2007.03.007

Cited by 23 publications

(23 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The agreement between the simulated and modeled contributions and profiles was evaluated for each data set by calculating the average absolute error (AAE) (Christensen and Gunst, 2004;Christensen et al, 2006;Lingwall and Christensen, 2007). Unlike other model performance statistics, such as the root mean square error (RMSE), the AAE gives the amplitude of average error without being pulled by the large one (Javitz et al, 1988;Willmott and Matsuura, 2005).…”

Section: Discussionmentioning

confidence: 99%

“…(1)) was generated by using average contributions (Table S1 in the Supporting Information) that were available in the literature (Zheng et al, 2002;Lee et al, 2003;Kim et al, 2005;Marmur et al, 2005;Alastuey et al, 2006;Hopke et al, 2006;Kulkarni et al, 2007;Song et al, 2008;Yuan et al, 2009;Chan et al, 2011;Guo et al, 2011). Next, hourly fluctuations were added by assuming random variability and a lognormal distribution (Lingwall and Christensen, 2007) for each source, except for vehicle exhaust. Time series data of carbon monoxide (a classic tracer for traffic) collected at an air quality monitoring station in an urban area were used to approximate the diurnal and weekly vehicle exhaust variations.…”

Section: Data Simulationmentioning

confidence: 99%

“…The factor profile matrix F (i.e., f pj in Eq. (1)) was derived from the US EPA SPECIATE database (Lingwall and Christensen, 2007;Simon et al, 2010), which included 18 VOCs and 17 PM 2.5 species (Fig. S1 in the Supporting Information).…”

Section: Data Simulationmentioning

confidence: 99%

“…This assessment was achieved by using synthetic data as in several previous studies (Miller et al, 2002;Brinkman et al, 2006;Christensen et al, 2006;Lingwall and Christensen, 2007;Hemann et al, 2009;Vedal et al, 2009;Habre et al, 2011).…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Evaluation of a Modified Receptor Model for Solving Multiple Time Resolution Equations: A Simulation Study

Liao

Kuo

Hopke

et al. 2013

Aerosol Air Qual. Res.

View full text Add to dashboard Cite

This study was conducted to evaluate the performance of an improved source apportionment model that is suitable for incorporating data with multiple time resolutions. This evaluation was achieved by using synthetic data sets that simulated environmental concentrations of volatile organic compounds (VOCs) and fine particulate matter (PM 2.5 ) from the five following sources: petroleum refinery, vehicle exhaust, industrial coating, coal combustion, and natural gas. Hourly VOCs and speciated PM 2.5 data were simulated for a one-week period. The PM 2.5 data were further averaged every twelve hours to generate data sets with mixed temporal resolutions. The Multilinear Engine program was applied to resolve the source profiles and contributions. A series of sensitivity analyses was conducted to examine how uncertainties in the profile variation, measurement error, and source collinearity affected the model performance. The resolved factor profiles closely matched the input profiles, and the measurement error had a larger impact on the modeling results than the profile variation. In the most comprehensive data set that contained all three types of uncertainty, the R 2 values between the input and retrieved source contributions were between 0.87 and 0.95. The estimated percentage contributions were also comparable with the input ones, demonstrating the applicability and validity of this improved model.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Data Simulationmentioning

confidence: 99%

Section: Data Simulationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Evaluation of a Modified Receptor Model for Solving Multiple Time Resolution Equations: A Simulation Study

Liao

Kuo

Hopke

et al. 2013

Aerosol Air Qual. Res.

View full text Add to dashboard Cite

show abstract

“…Positive matrix factorization does not guarantee uniquely identified estimates of and F, but merely reduces the degree of nonidentifiability. Lingwall and Christensen (2007) evaluate the use of PMF in pollution receptor models via simulation studies in the presence and absence of a priori information. As PMF is currently the most commonly used estimation technique, the methods developed in this article will be compared to the results based on a PMF analysis.…”

Section: Introductionmentioning

confidence: 99%

Incorporating Time-Dependent Source Profiles Using the Dirichlet Distribution in Multivariate Receptor Models

2010

Self Cite

View full text Add to dashboard Cite

Multivariate receptor modeling is used to estimate profiles and contributions of pollution sources from concentrations of pollutants such as particulate matter in the air. The majority of previous approaches to multivariate receptor modeling assume pollution source profiles are constant through time. In an effort to relax this assumption, this article uses the Dirichlet distribution in a dynamic linear receptor model for pollution source profiles. The receptor model developed herein is evaluated using simulated datasets and then applied to a physical dataset of chemical species concentrations measured at the U.S. Environmental Protection Agency's St. Louis-Midwest supersite. Supplemental materials to this articles are available online.

show abstract

Pollution source direction identification: embedding dispersion models to solve an inverse problem

2011

Self Cite

View full text Add to dashboard Cite

We develop a Bayesian method for identifying pollution source directions that combines deterministic and stochastic models. We frame the source direction identification as an inverse problem, embedding the deterministic dispersion model American Meteorological Society/United States Environmental Protection Agency Regulatory Model (AERMOD) directly into the likelihood function. AERMOD's fast computation time allows us to run the model at each iteration of the Markov chain Monte Carlo (MCMC), thereby creating a simulated likelihood function and obviating the need for an emulator. The method is flexible enough to identify multiple source directions for cases in which a species or source type of interest is emitted at more than one location, and reversible jump MCMC is used to evaluate the appropriate number of sources. Source direction identification is an important part of the pollution source apportionment problem, which entails identifying and describing pollution sources and their contributions.

show abstract

Pollution source apportionment using a priori information and positive matrix factorization

Cited by 23 publications

References 6 publications

Evaluation of a Modified Receptor Model for Solving Multiple Time Resolution Equations: A Simulation Study

Evaluation of a Modified Receptor Model for Solving Multiple Time Resolution Equations: A Simulation Study

Incorporating Time-Dependent Source Profiles Using the Dirichlet Distribution in Multivariate Receptor Models

Pollution source direction identification: embedding dispersion models to solve an inverse problem

Contact Info

Product

Resources

About