Source Identification of Volatile Organic Compounds in Houston, Texas pdf

In this paper, a novel factor analysis model, where the normal chemical mass balance model was augmented by a parallel equation that accounted for wind speed and direction, temperature,

Source Identification of Volatile

Organic Compounds in Houston,

Texas

W E I X I A N G Z H A O ,†

P H I L I P K H O P K E , *, † A N D T H O M A S K A R L‡

Department of Chemical Engineering, Clarkson University,

Box 5708, Potsdam, New York 13699-5708, and The National

Center for Atmospheric Research, Atmospheric Chemistry

Division, P.O Box 3000, Boulder, Colorado 80307

The complexity of the volatile organic compound (VOC)

mixture in the Houston area makes studies of the air quality

in that area very challenging In this paper, a novel

factor analysis model, where the normal chemical mass

balance model was augmented by a parallel equation that

accounted for wind speed and direction, temperature,

and weekend/weekday effects, was fitted with a multilinear

engine (ME) to provide identification and apportionment

of the VOC sources at the La Porte Municipal Airport site

in Houston during the Texas Air Quality Study (TexAQS)

2000 The analysis determined the profiles and contributions

of nine sources and the corresponding wind speed,

wind direction, temperature, and weekend factors The

reasonableness of these results not only suggests the high

resolving power of the expanded factor analysis model

for source apportionment but also provides the novel and

effective auxiliary information for more specific source

identification In addition, a new approach to estimate the

measurement uncertainty and the details of determining

the source number and dealing with missing values are also

presented as important parts of the data analysis process.

This study demonstrates the feasibility of the expanded

model to identify sources in complex VOC systems and

extract useful information for locating VOC emitters

and controlling their emissions in the Houston area.

Introduction

Volatile organic compounds (VOC) are organic chemicals

that easily vaporize at room temperature Many VOCs have

been found to have adverse effects on air quality and human

health (1) For example, long time exposure to benzene will

increase the risk of leukemia, and reactive VOCs such as

primary olefins are important in the formation of tropospheric

ozone However, motor vehicle exhaust, chemical

manu-facturing, paints, solvents, biogenic emissions, and many

other sources create exposure to VOCs Because identification

of the potential sources of VOCs is a prerequisite for

controlling VOCs’ emissions and protecting air quality and

public health, it has been paid more and more attention (2).

To identify the number of sources and their profiles,

receptor models are widely used (3, 4) There are two principal

approaches to receptor modeling If the number and the

profiles of sources are known, chemical mass balance (CMB) can be used to estimate the contribution of each source to

the pollution (5) where regression methods are used to

provide quantitative results However, in many cases, source information is unknown a priori, so factor analysis (multi-variate analysis) needs to be used to extract the sources information

Hopke and co-workers (6), Heidam (7), Henry (8), and Barrie and Barrie (9) applied principal component analysis (PCA) to source identification, but Paatero and Tapper (10, 11) showed that PCA cannot provide a true minimal variance

solution since they are based on an incorrect weighting In view of the limitations of PCA, a new technique, positive matrix factorization (PMF), was developed for sources

identification and apportionment (12) The distinct

advan-tages of PMF over PCA are that non-negative constraints are built in PMF models and PMF does not rely on the information from the correlation matrix but utilizes a

point-by-point least-squares minimization scheme (12) It has been reported (13) that the source profiles produced by PMF are

better and more reasonable at describing the source structure than those by PCA Over the past few years, PMF has been applied to a number of particle composition data sets (e.g.,

14, 15).

Recently, the PMF analysis can be expanded by using a

more general model (16), and a new analysis tool called the multilinear engine (ME) was developed (17) to solve such

problems ME is very flexible and provides a general

framework for fitting any of the multilinear model (18, 19),

so it becomes possible to obtain not only the sources profiles but also other interesting parametric factors that may be important for source identification and pollution control and planning For example, wind directional information can help

locate the potential sources It was reported (16, 19) that in

some cases the expanded factor models could determine more sources than PMF

The coexisting system of VOCs is complex A small change

in environmental conditions (e.g., temperature) may result

in changes on VOC concentrations, and also some VOCs may be involved in chemical reactions during the trans-portation Meanwhile, as a consequence of high density of petroleum refineries, synthetic organic chemical plants, and various mobile sources, the formation rate and concentration

of ozone in the Houston area are extremely high, and propene

becomes a dominant reactive hydrocarbon (20) The specific

VOC mixture in the Houston area represents a specific air

quality problem compared with other metropolitans (21),

which makes studies of the air quality in Houston very

challenging Henry et al (22, 23) made some studies on VOC

source identification, one of which was also for the Houston area with the data for the period June through November

1993, but these studies did not provide any information about the influences of environmental parametric factors (e.g., wind speed, wind direction, and temperature) on the observed pollutant concentrations, and only three sources in the Houston area were identified Therefore, in the present study,

an expanded factor analysis model will be used to identify the VOCs sources in the Houston area with the goals of (1) checking the feasibility of the expanded modeling to VOC sources, (2) observing the influences of environmental parametric factors on the observed concentrations, and (3) supplying convincing information that will be useful for air pollution control in the Houston area ME will be used as an optimization tool for data fitting in this study since it has

proved to be effective in model fitting (16, 18, 19).

* Corresponding author phone: (315)268 3861; fax: (315)268 4410;

e-mail: hopkepk@clarkson.edu

†Clarkson University

‡The National Center for Atmospheric Research

Environ Sci Technol.2004,38,1338-1347

13389 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL 38, NO 5, 2004 10.1021/es034999c CCC: $27.50  2004 American Chemical Society

Expanded Factor Analysis Model

In general, the ordinary bilinear receptor model can be written

as

where X is the matrix of VOCs’ concentrations, F is the matrix

of source profiles, G is the matrix of source contributions,

and E is the residuals matrix Their elements x ij , f jp , and g ip

can be respectively understood as the concentration of

compound j measured in sample i, the concentration of

compound j in the emission of source p, and the strength

of source p on sample i (16, 19).

In this section, wind direction and wind speed will be

used to illustrate the construction of the expanded factor

analysis model (16, 19) In the bilinear model, the contribution

of source p to the concentration of compound j in sample

i, u ijp , is represented by u ijp ) g ip f jp In the expanded model,

a parallel equation is developed where the contribution u ijp

can be represented by another form

where d i and s iare the values of wind direction and wind

speed of sample i The ranges of wind direction and wind

speed are divided into a series of subranges having similar

numbers of samples in each Then each wind direction/

speed belongs to a specific range Thus, D(d i ,p), an element

of D, represents the action of source p on pollution in the

wind direction range of d i For example, if source 3 strongly

affects the observed concentration at the wind direction of

80°, D(4,3) (the first index, 4, corresponds to 80°if the wind

direction range 0-360°are evenly divided into 18 subranges)

should be a relatively larger value S(s i ,p) has a similar

definition for wind speed Thus, z ipcan be considered as a

multiplier that represents the comprehensive action of wind

direction and speed on the observed pollution Obviously,

in different physical models, z ipcan correspond to different

expressions In this study, z ipcorresponds to the factors for

wind speed, wind direction, temperature, and weekday/

weekend

The expanded receptor model can then be expressed by

where W(w i ,p) denotes the action of weekdays or weekends

by source p on the observed concentration, I is the number

of samples, and J is the number of measured chemical species.

By fixing the weekday coefficient at unity, W(w i ,p) is a vector

with n p (the number of sources) elements T(t i ,p) represents

the action of temperature t i for source p The task of solving

this expanded PMF model is to determine the values of F,

G, D, S, W, and T to fit the data as well as possible The

optimization problem can be defined as

in which e ij and e′ij are determined by eqs 3a and 3b and σ ij

and σ′ijare the error estimates, which can be considered as

special weights Clearly, z ip is the combination of all

influencing factors such as wind speed, temperature, and weekday/weekend factors, so with ME an expanded factor analysis model can provide us not only the source profiles and contributions but also the strength of other factors affecting the observed concentrations A prerequisite to applying this model is that the action (contribution) of considered parametric factors can be expressed in linear terms As an optimization method for factor analysis, ME has two problems to be solved, that is, how to determine the number of factors and how to avoid local optimal In the section of Results and Discussion, the methods for solving these two problems for this case will be described in detail Because eq 3b will generate a poorer fit to the data than

eq 3a, the error estimate for eq 3b, σ′ij, must be (much) larger

than that for eq 3a, σ ij (16, 19) In this study σ′ijis 8 times of

σ ij and σ ijis represented as

where c1denotes the uncertainty of measurement and c3is

a constant Here c3is valued at 0.2 Because of the complex VOC mixture in the ship channel area and the potential interference at low concentration, the experimental uncer-tainties obtained by the measurement technique used here were hard to access An approach using the fast Fourier transformation (FTT) was applied to solve this problem The procedure can be briefly described as follows Xylene will be used as an example from the species being studied in this

paper Let c be the concentration series of xylene, which has

7292 measurements Thus, the key steps to estimate the measurement uncertainty from the measurement series are

as follows:

(a) Generate a random series r with the same length as

c (7292 elements) and variance ν r2) 1

(b) Perform FFT on c and r, and calculate their magnitude spectra and call them mfc and mfr.

(c) Plot mfc and mfr, respectively It can be seen from Figure 1 that mfc consists of two parts; one with low

frequencies represents the useful information while the other

with high frequencies represents the noise, and mfr does

not show two different parts because it is generated by a random series

(d) Select an interval of noise in mfc Although it is difficult

to determine the exact starting and ending points of the noise interval, the selected noise range should be sufficiently long

to reflect the noise information In this example, the selected

interval was from mfc(1000) to mfc(6000) Then, calculate

the mean value of the selected interval and name it m_mfc.

X ) GFT+ E (1)

x ij)∑

p)1

N

x ij)∑

p)1

N

z ip f jp + e′ij)

∑

p)1

N

D(d i ,p)S(s i ,p)W(w i ,p)T(t i ,p)f jp + e′ij (3b)

i)1

I

∑

j)1

J (e ij /σ ij)2+∑

i)1

I

∑

j)1

J (e′ij /σ′ij)2 (4)

FIGURE 1 Illustration of FFT-based uncertainty estimation (a) The concentration series of xylene, c; (b) the magnitude spectrum of concentration series, mfc; (c) the random data series, r; (d) the magnitude spectrum of random data series, mfr.

σ ij ) c1+ c3x ij (5)

(e) Select the same range in mfr, and calculate the mean

value of this range and name it m_mfr Actually, it is also

feasible to use the whole range of mfr since there is only

noise in mfr.

(f) Calculate νcaccording to

and consider it as the estimation of the uncertainty of the

xylene concentration series

Although this estimation could not be guaranteed to be

fully accurate because, strictly speaking, each measurement

should have its own uncertainty, it is practical since it

produces satisfactory analysis results Finally, the estimated

uncertainties for the compounds in this study, acrylonitrile,

isoprene, benzene, toluene, styrene, c8-benzenes (with the

dominant component: xylene), c7-ketone, c9-benzenes,

c10-benzenes, c13-c10-benzenes, M43, M61, and M87, are 0.074, 0.102,

0.114, 0.147, 0.033, 0.114, 0.070, 0.082, 0.047, 0.016, 0.835,

0.360, and 0.089, respectively Here M43, M61, and M87

denote the classes of compounds with the mass/charge values

of 43, 61, and 87, respectively In this study, the dominant

components for them are propene, acetic acid, and vinyl

acetate, respectively

Data Preprocessing

As part of the TexAQS 2000, a proton transfer reaction mass

spectrometer (PTR-MS) from the University of Innsbruck

was placed in an air-conditioned trailer situated next to a

10-m sampling tower at the southwest side of the municipal

airport at La Porte, TX, to identify and quantify the VOC

mixture in that area A map showing the sampling site is

presented in Figure 2 The PTR-MS technique has been

previously described in detail (24), so only a brief description

is given here The principle of the PTR-MS is the reaction of

organic species in ambient air with H3O+ions, generated

from the hollow cathode discharge of water vapor, to produce

the protonated organic species (RH+) The concentration of

the product ions can be calculated from a reaction dynamic

equation (24) Only organic species with a proton affinity

greater than that of water can be detected by the mass

spectrometer More details about the sampling procedure

can be found in ref 20.

The sampling period for the data in this study was from 08/20/00 to 09/08/00, and the most sampling frequencies were about 1/4-6 min-1, but the frequencies for some periods were 1 min-1 All the samples were used for analysis to ensure

a sufficiency of samples The concentrations of 14 VOCs (methanol, acrylonitrile, isoprene, benzene, toluene, styrene, c8-benzenes (xylenes), c7-ketone, c9-benzenes, c10-ben-zenes, c13-benc10-ben-zenes, M43 (propene), M61 (acetic acid), and M87 (vinyl acetate)) were selected for this study, with the detection limits being 100 pptv, 60 pptv, 20 pptv, 70 pptv, 70 pptv, 30 pptv, 70 pptv, 60 pptv, 30 pptv, 30 pptv, 30 pptv, 1 ppbv, 1 ppbv, and 200 pptv, respectively Some of the reasons for this selection are benzene, toluene, and xylene (BTX) compounds are usually considered of high importance for urban VOC reactivity/air quality, propene is one of the dominant reactive hydrocarbons in Houston, which makes Houston a special case when compared to other U.S cities, and the toxicity of acrylonitrile directly affects the air quality

in the vicinity of an emitter The meteorological data like temperatures were from the NOAA Aeronomy Lab and winds were measured next to the VOC inlet at 10 m above ground Because there were some missing and below detection limits values in the concentration measurements and meteorologi-cal data, 7292 samples were retained for analysis following the pretreatment below Consecutive missing values (for example, c7-ketone has 2240 consecutive missing values) were deleted and the values below detection limits were

replaced with half of the detection limits (25).

Additionally, the data for wind direction, wind speed, and temperature were divided approximately evenly into 31, 5, and 5 levels, respectively, so in each factor, the effect of any level will not be overwhelmed by any of the others

Results and Discussion

Due to the long lifetime and multiple sources, methanol proves to be a ubiquitous compound The initial trials including methanol showed that methanol had a significant contribution in each source profile, suppressing other compounds such as vinyl acetate and c13-benzenes Such behavior typically suggested that there was a high variability

in the amounts of methanol associated with its sources In this case, an increase of the uncertainty of methanol could not resolve this problem Thus, methanol was excluded from the final analysis Figure 3 shows the concentration time series for all compounds except methanol

The determination of the number of sources is one of the major problems in any factor analysis In this study, three rules were applied to decide the proper source number that

(1) the resolved source profiles should be explainable, (2) Q

value defined in eq 4 is expected to show a change in slope with the number of sources from rapid to slow at the point

of the decided number, and (3) there should be a satisfactory fit between the predicted concentrations and the measured values In detail, when the source number increased from 6

to 7, 7 to 8, 8 to 9, and 9 to 10, the decreases in Q were 6907,

8325, 5723, and 4915, respectively Clearly, there is a change

in the slope at 9 sources

In addition, there was a better fit between the predicted concentrations and the measured values at that source number The fit between the predicted c13-benzenes con-centration and the measured values increased exceptionally quickly when the source number changed from 9 to 10 (The correlation coefficient for 10 sources was 0.932 while that for

9 sources was 0.58.) However, actually more than 75% of the c13-benzenes concentration measurements were below the detection limit and replaced by half of the detection limit,

so the exceptionally good fit might suggest that there was overfitting of c13-benzene for the case of 10 sources Neither poor fit nor over fit is acceptable So the number of sources for this analysis was chosen to be 9 During the experiments,

FIGURE 2 Sampling site (La Porte Municipal Airport) in Houston,

Texas.

vc

the candidate cases had at least three runs to avoid local

optima, and finally the case of 9 sources (excluding methanol)

was selected as the best solution The results are discussed

below

The profiles and time-resolved contributions of nine

sources are shown in Figures 4 and 5 To describe the

contribution variation of source i between weekdays and

weekends in quantity, a ratio called KD is defined as eq 7:

The plots of the wind direction factor, wind speed factor,

temperature factor, and weekend factor of 9 sources are

shown in Figures 6, 7, 8, and 9, respectively In the wind

directional plots, each column of matrix D is displayed in a

polar plot to represent the factor values for the different wind

directions (i.e., the longer the radius is, the bigger the

contribution at that direction) In addition, the emitter

location plots of acrylonitrile, toluene-xylene, benzene,

styrene, and propene are presented in Figures 10-14 The

plot showing emitter locations was produced by

superim-posing the wind directional plot (blue area) onto the map

where the corresponding emitters in the observed area were

displayed as circles, squares, or triangles Emission rates for

2000 that were obtained from the Toxic Release Inventory

(26) were shown on the plots with the corresponding colors,

if available The size of blue area in each plot does not

represent the distance between receptor site and emitter

but denotes the strengths of the identified source on the

pollution at different wind directions

Activity can change considerably from weekdays to

weekends Some production factories do not operate on

weekends, so the emissions of these sources vary accordingly

In addition, the pattern of motor vehicle use also changes

as fewer people commute to work and fewer heavy-duty diesel trucks will be operated on weekends Thus, the weekend factor should reflect changes in the human activities However, in this study, the number of weekend samples (there are only 3 weekends in this study and moreover they contain many missing values) might not be sufficient enough to obtain

FIGURE 3 Concentration series of each compound.

FIGURE 4 Profiles of the identified VOC sources in La Porte, Houston.

FIGURE 5 Time-resolved contribution of each identified VOC source

in La Porte, Houston.

KDi)mean{g i,j |j ∈ weekends}

a general conclusion, but they can provide us an initial

estimate of the influence of weekend factors on pollution

It can be seen from Figure 9 (the value for the source without

weekend effect should be around 1) that sources 3 and 9

have significant weekend influence while sources 1, 2, 4, and

8 show only a weak weekend influence The possible reasons

for weak weekend effect might be (1) although relatively little

isoprene in source 2 is biogenerated, its emission should be

independent of weekend/weekday and (2) the c9-benzene,

c10-benzene, toluene, xylene, and other compounds in these

4 sources are from refineries that are usually operated in

continuous mode, and their emission rates on weekends may

overwhelm the negative effect caused by the decreased

number of motor vehicles

Source 1 contains mainly acrylonitrile Figure 10 shows

that most acrylonitrile emitters are located to the northwest

and south of the sampling site Likely, these emitters include

the boilers, dryer stacks, aeration tanks, ponds, and waste

gas processing equipments of chemical or rubber plants (27).

The wind directional plot for this source agrees with the

emitter locations as it shows a large contribution from the

northwest and a peak at about 150° The high peaks of the

contribution plot for this source correspond to the nighttime

period when southerly winds dominate No information is

available on the diurnal patterns of the source In addition,

the KD value of this source in Table 1 is 0.404, so this source

is expected to have a significant weekend influence However

such an influence does not agree with the result in Figure

9 As mentioned above, the limited number of weekend measurements for analysis may not be sufficient, which may

be the cause of this disagreement Better results for the weekend factor might be obtained if a larger data set were available

Source 2 shows isoprene and M87 (vinyl acetate) Isoprene

is a typical biogenic VOC (28), but the contribution of biogenic

isoprene is small in the immediate proximity of the La Porte

site (29) A number of anthropogenic isoprene emitters (likely,

rubber industry) are located to the north and south of the

sampling site (20, 27) For M87, there are a number of vinyl

acetate emitters to the north and south of the sampling site, and especially several large vinyl acetate emitters are located

to the north (27) These emitters are most likely the storage

tank and other equipment of chemical plants The wind directional plot for this source in Figure 6 largely confirms the location of these emitters, as it shows some convexes in

FIGURE 6 Wind direction factor plots.

TABLE 1 Ratio of Mean Contribution of Weekend Samples to That of Weekday Samplesa

no 1 2 (M) 3 (M) 4 (M) 5 (M) 6 (M) 7 (M) 8 9 (M)

KD 0.404 0.964 0.599 1.039 0.579 0.662 0.738 1.410 0.493

a The (M) after the number denotes the KD value of this source is identical to the corresponding weekend factor result.

FIGURE 7 Wind speed factor plots.

FIGURE 8 Temperature factor plots.

the south and a broad contribution from the north The

contribution of biogenic isoprene is small, but a number of

anthropogenic isoprene emitters are located at similar

directions as the vinyl acetate emitters This might be one

of the reasons why isoprene and M87 occur in the same source The KD value of this source is 0.964, which agrees with the result of weekend factor that there is a very weak weekend effect for this source

Source 3 is characterized by c7-ketone There are some possible emitters (synthetic organic manufacturing plants)

to the southeast of the sampling site (20), so the wind

FIGURE 9 Weekend factor plot.

FIGURE 10 Locations of acrylonitrile emitters The circles denote

the acrylonitrile emitters The blue area corresponds to the wind

direction plot for this source The red × at the center of the blue

area is the sampling site.

FIGURE 11 Location of toluene-xylene emitters The red squares

and blue circles denote xylene and toluene emitters, respectively.

The blue area corresponds to the wind direction plot for this source.

The red × at the center of the blue area is the sampling site.

FIGURE 12 Location of benzene emitters The green triangles denote benzene emitters The blue area corresponds to the wind direction plot for this source The red × at the center of the blue area is the

sampling site.

FIGURE 13 Location of styrene emitters The black circles denote styrene emitters The blue area corresponds to the wind direction plot for this source The red × at the center of the blue area is the

sampling site.

FIGURE 14 Location of propene The circles denote the propene emitters The blue area corresponds to the wind direction plot for this source The red × at the center of the blue area is the sampling

site.

directional plot shows a broad contribution from that

direction There are a number of peaks in the corresponding

contribution plot, but none of these peaks were on the

weekend In addition, the KD value of this source is 0.599

and identical to the result of weekend factor

Source 4 contains toluene and xylene These emitters are

operation units and equipment of the chemical and refining

industry (e.g., tanks, boilers, reactors, pyrolysis furnaces) (27).

Figure 11 shows the emitters are mainly located to the

northwest and south of the sampling The wind directional

plot shows a large contribution from the north and a sharp

spike in the south, in agreement with the locations of the

major emitters In addition, toluene and xylene can be

generated by motor vehicles (highway 225 is just to the north

and highway 146 to the east and southeast of the sampling

site) This can be another cause of the shape of the wind

directional plot There are many peaks almost evenly

distributed in the contribution plot and the KD value of this

source is 1.039, which agrees with the weak weekend effect

in Figure 9 In addition, the time corresponding to the peaks

was mainly in the morning (6:00-8:00) and night

(22:00-24:00)

Source 5 is characterized by benzene mostly from

chemical plants or refineries Figure 12 shows that most

benzene emitters are distributed to the north and south of

the sampling site (27) Particularly, one benzene source is

located on Bay Area Blvd (in the direction of 150°) (20) In

addition, the motor vehicles on highways 225 and 146 may increase the concentration of benzene The location infor-mation of the emitters is supported by the wind directional plot that shows a broad contribution from the north and a spike in the direction at about 150° The contribution plot for this source shows a number of peaks, none of which were

on the weekends, and the KD value of this source is 0.579 This agrees with the result of weekend factor The significant variation between weekends and weekdays suggests that the contribution of mobile sources on weekdays might have a greater impact on benzene emission

Source 6 is characterized by styrene Figure 13 shows the location of the styrene emitters, most of which are likely to

be the units of petrochemical plants There are many styrene emitters around the sampling site and most of them are

located to the northwest and south of the sampling site (27).

Particularly there is a large styrene emitter at about 210° The corresponding wind directional plot in Figure 6 agrees with the location information as it shows a broad contribution from the north and also a sharp spike at the direction of 210° Most of the high peaks in the contribution plot were at nighttime when the southerly wind was dominant The KD

FIGURE 15 Distribution plot for scaled residual errors.

value of this source is 0.662, which agrees with the weekend

factor result that this source has a weekend effect

Source 7 is represented by M61 whose kernel component

is acetic acid Acetic acid is a typical photochemical reaction

product (30), so the wind direction plot shows a relatively

smooth shape Meanwhile, a number of anthropogenic acetic

acid emitters (e.g., acetic acid storage tanks, boilers, and

exhausted liquid tanks) are located to the north and south

of the sampling site (27) The KD value of this source is 0.738

and identical to the weekend factor result The photochemical

sources should not have weekend effect, so the variation

between weekday and weekend may be due to the changes

of the emission rates of the anthropogenic acetic acid sources

Source 8 is characterized by c9-benzenes and

c10-benzenes A number of c9- and c10-benzenes emitters (e.g.,

the operation units of chemical or petrochemical plants) are

located to the north of the sampling site (27) The wind

directional plot for this source shows a broad contribution

from the northwest and a spike in the south The motor

vehicles on highways 225 and 146 and Spencer Hwy may

increase the concentrations of c9- and c10-benzenes and

may be a reason for the spike in the south The peaks in the

contribution plot are distributed over both weekdays and

weekends (August 27 and September 2), but the KD value of

this source is 1.410 Therefore, a weekend effect is expected

However, the weekend factor result in Figure 9 shows only

a weak weekend influence As before, this discrepancy may

arise from the limited weekend data

Source 9 is represented by M43 (propene) Propene is

most likely emitted by the refineries along the ship channel

(20) Figure 14 shows a number of large emitters are located

to the north and northeast of the sampling site, which is

supported by the wind directional plot with a large

contri-bution from the northeast Most high peaks in the

contribu-tion plot correspond to daytime periods when the dominant

wind is a northerly wind The KD value of this source is 0.493,

so this source seems to have a significant weekend effect,

which agrees with the weekend factor result

The 13 VOCs, except c13-benzene which only appears in

small amounts, are distributed into reasonable source

profiles, and the corresponding contribution and directional

patterns are in general agreement with known source

information One reason for the absence of c13-benzenes

can be that almost 75% of this compound’s measurements

were below the detection limit and were replaced with half

of the detection limit Therefore, it may not be possible to

make any quantitative attributions for this compound

Because of the same reason, the scaled residual errors for

c13-benzenes in Figure 15 are not satisfactory while the others

have a reasonable distribution Although the weekend data

are not sufficient enough to make a correct conclusion on

weekend effect for each source, the weekend factors of most

sources (7 out of 9) are identical with the defined KD values

These results suggest the feasibility of including the weekend

effect analysis

Wind speed and temperature are two potentially

impor-tant meteorological factors that can help interpret the

observed VOC concentrations Figure 7 shows the wind speed

factor For most factors, the wind speed factor values decrease

with increasing wind speed This trend suggests a dilution

effect that the same emitted mass is released into a larger

volume of air as wind speed increases; the concentration

therefore decreases (16) However, the factors of sources 2

and 9 increase with increasing wind speed and source 3 shows

an almost flat curve The possible reasons for these

phe-nomena might be (1) for these sources that may be composed

of point emitters (e.g., high-concentration storage tank), there

may be more coherent plume effect at higher wind speed

(higher wind speed makes these emitted VOCs gathered

together rather than dispersed) and (2) high-speed wind may enhance the evaporations of some VOCs

The influence of temperature on pollutants is more complex than that of wind speed because increasing tem-perature will not only speed up the vaporization of VOCs but also change the chemical properties of VOCs and enhance the reactions between VOCs and oxidants in the air It is relatively difficult to summarize the action of temperature

on the observed concentration For some sources (e.g., Nos

2, 3, and 7), the temperature factor values in Figure 8 increase with temperature This trend might be the result of increased vaporization Another explanation for source 7 is likely that the increase in temperature enhances the rates of the photochemical reactions However, the temperature factors for other sources do not show increasing trends

The source identification of VOCs in the La Porte Airport has been successfully performed using an expanded factor analysis model and the corresponding optimization tool, ME The profiles and contributions of the 9 identified sources proved to be reasonable Besides, wind direction, wind speed, temperature, and weekend factors were also determined The information on wind directions appears to agree with the known emission inventories and the wind speed factors for most sources suggest a dilution effect For many sources, weekend and temperature factors help in interpreting their influences on the observed concentrations It is not clear that this model is the best representation of the physical and chemical influences of such factors on the observed con-centrations However, the results do suggest this is a feasible direction for such a study In addition, the results suggest that the error estimation obtained through the FFT is reasonable in term of finding an interpretable solution It appears that expanded modeling is feasible for not only identifying VOC sources in complex systems such as the air system in Houston but also revealing the various important features of these sources

Acknowledgments

This work was supported by the United States Environmental Protection Agency through cooperative agreement number R-82806201 under a subcontract to Clarkson University by The University of Texas at Austin (UT) Although the research described in this article has been funded wholly or in part

by the United States Environmental Protection Agency, it has not been subjected to the Agency’s required peer and policy review and, therefore, does not necessarily reflect the views of the Agency and no official endorsement should be inferred The meteorological data for this study were supplied

by NOAA Aeronomy Lab

Literature Cited

(1) Ilgen, E.; Karfich, N.; Levsen, K.; Angerer, J.; Schneider, P.;

Heinrich, J.; Wichmann, H.; Dunemann, L.; Begerow, J Atmos.

Environ 2001, 35, 1235-1252.

(2) Fujita, E M.; Watson, J G.; Chow, J C.; Magliano, K L Atmos.

Environ 1995, 29, 3019-3035.

(3) Hopke, P K Trends Anal Chem 1985, 4, 104-106.

(4) Hopke, P K Receptor Modeling for Air Quality Management;

Elsevier: Amsterdam, 1991

(5) Cooper, J A.; Watson, J G.; Huntzicker, J J Atmos Environ.

1984, 8, 1347-1355.

(6) Hopke, P K.; Gladney, E S.; Gordon, G E.; Zoller, W H.; Jones,

A G Atmos Environ 1976, 10, 1015-1025.

(7) Heidam, N Z Atmos Environ 1981, 15, 1421-1427 (8) Henry, R C Atmos Environ 1987, 21, 1815-1820.

(9) Barrie, L A.; Barrie, M J J Atmos Chem 1990, 11, 211-226 (10) Paatero, P.; Tapper, U Chemom Intell Lab Syst 1993, 18,

183-194

(11) Paatero, P.; Tapper, U Environmetrics 1994, 5, 111-126 (12) Paatero, P Chemom Intell Lab Syst 1997, 37, 23-35 (13) Huang, S.; Rahn, K A.; Arimoto, R Atmos Environ 1999, 33,

2169-2185

(14) Lee, E.; Chan, C K.; Paatero, P Atmos Environ 1999, 33,

3201-3212

(15) Kim, E.; Larson, T V.; Hopke, P K.; Slaughter, C.; Sheppard, L

E.; Claiborne, C Atmos Res 2003, 66, 291-305.

(16) Paatero, P.; Hopke, P K Chemom Intell Lab Syst 2002, 60,

25-41

(17) Paatero, P J Comput Graphical Stat 1999, 8, 854-888.

(18) Xie, Y.; Hopke, P K.; Paatero, P.; Barrie, L A.; Li, S Atmos Environ.

1999, 33, 2549-2562.

(19) Kim, E.; Hopke, P K.; Paatero, P.; Edgerton, E S Atmos Environ.

2004, in press.

(20) Karl, T.; Jobson, T.; Kuster, B.; Williams, E.; Stutz, J.; Goldan, P.;

Fall, R.; Fehsenfeld, F.; Lindinger, W J Geophys Res 2003, 108

(D16): Art No 4508

(21) Goldan, P D.; Parrish, D D.; Kuster, W C.; Trainer, M.; McKeen,

S A.; Holloway, J.; Jobson, B T.; Sueper, D T.; Fehsenfeld, F

C J Geophys Res 2000, 105, 9091-9105.

(22) Henry, R C.; Spiegelman, C H.; Collins, J F.; Park, E Proc Natl.

Acad Sci U.S.A 1997, 94, 6596-6599.

(23) Henry, R C.; Lewis, C W.; Collins, J F Environ Sci Technol.

1994, 28, 823-832.

(24) Salisbury, G.; Williams, J.; Holzinger, R.; Gros, V.; Mihalopoulos,

N.; Vrekoussis, M.; Sarda-Este˘ve, R.; Berresheim, H.; von

Kuhlmann, R.; Lawrence, M.; Lelieveld, J Atmos Chem Phys.

2003, 3, 925-940.

(25) Polissar, A V.; Hopke, P K.; Malm, W C.; Sisler, J F J Geophys.

Res 1998, 103, 19045-19057.

(26) Toxic Release Inventory, U.S Environmental Protection Agency, http://www.epa.gov/tri/ (accessed July/August 2000) (27) Texas Commission on Environmental Quality (TCEQ), 2001 Development of Source Speciation Profiles from the TNRCC Point Source Database (Final Report)

(28) Guenther, A.; Baugh, W.; Davis, K.; Hampton, G.; Harley, P.; Klinger, L.; Vierling, L.; Zimmerman, P.; Allwine, E.; Dilts, S.;

Lamb, B.; Westberg, H.; Baldocchi, D.; Geron, C.; Pierce, T J.

Geophys Res 1996, 101, 18555-18567.

(29) Wiedinmyer, C.; Guenther, A.; Estes, M.; Strange, I W.; Yardwood,

G.; Allen, D T Atmos Environ 2001, 35, 6465-6477.

(30) Seinfeld, J H.; Pandis, S N Atmospheric Chemistry and Physics: From Air Pollution to Climate Change; John Wiley & Sons: New

York, 1998

Received for review September 11, 2003 Revised manuscript received December 8, 2003 Accepted December 18, 2003.

ES034999C