SMOKY MOUNTAIN AMBIENT HALOCARBON AND HYDROCARBON
MONITORING, SEPTEMBER 21-26, 1978
by
Dagmar R. Cronn
David E. Harsch
Air Pollution and Resources Section
Chemical Engineering Department
Washington State University
Pullman, Washington 99164
R0804033-03-2
Project Officer
Will iam Wilson
Regional Field Studies Office
Environmental Sciences Research Laboratory
Research Triangle Park, NC 27711
U.S. ENVIRONMENTAL PROTECTION AGENCY
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
RESEARCH TRIANGLE PARK, NC 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
-------
ABSTRACT
A cooperative field project was conducted in the Smoky Mountains
National Park, Tennessee during September, 1978. Participants included
the Environmental Protection Agency (Regional Field Studies Office and
Atmospheric Instrumentation Branch) and Washington State University (Air
Pollution Research Section of the Department of Chemical Engineering).
The purpose of the study was two-fold. The principal goal was to study the
relative importance of biogenic emissions on aerosol burden. In addition,
the field work served as preparation for the U.S. scientific delegation to
the U.S./U.S.S.R. Joint Abastumani Forest Aerosol Experiment of 1979, as
well as an adjunct to the VISTTA program (Visibility Impairment due to Sulfur
Transportation and Transformation in the Atmosphere).
WSU was responsible for ascertaining the extent to which anthropogenic
emissions impacted the study site. This objective was met by operation
of a continuous, automatic gas chromatograph with electron capture detection
for analysis of CF2C12, CFC13, C^CC^, CC14 and NgO. With the exception
of a few hours during a rainstorm, regional pollution buildup as well as
excursions due to advection of air across more localized sources was contin-
uously observed.
A second WSU responsibility was the measurement of the biogenic hydro-
carbons which are postulated as the precursors of the secondary, natural
aerosol. Whole air samples were collected and analyzed by flame ionization
IV
-------
detection gas chromatography for hydrocarbons. Isoprene and several terpenes
were identified. Isoprene accounted for 4% to 13% of the total non-methane
hydrocarbons (TNMHC) measured in the sample. The terpenes and isoprene
typically accounted for less than 30% of the TNMHC, thereby constituting a
minor amount of the hydrocarbon components. However, due to the diffuseness
of the biogenic source and the short expected lifetime of biogenic hydrocar-
bons, their contribution to rural aerosols might not be insignificant.
This report was submitted in partial fulfillment of Grant No. R0804033
by Washington State University under the sponsorship of the U.S. Environmen-
tal Protection Agency. This report covers the period April 15, 1978, to
April 15, 1980, and work was completed as of April 15, 1980.
-------
CONTENTS
Foreword iii
Abstract iv
Figures vli
Tables ixi
Acknowledgements x
1. Introduction 1
2. Conclusions 3
3. Instrumentation and Experimental Procedure 5
Continuous halocarbon measurements 5
Hydrocarbon analysis 13
4. Analytical Results 21
Halocarbon results 21
Aircraft samples 22
Hydrocarbon results 23
5. Discussion 26
How remote was the study site? 26
Meteorology as a determinant of halocarbon behavior . 34
Terpenes and other gaseous hydrocarbons 36
References 41
Appendices
A. Halocarbon continuous monitoring data 43
B. Halocarbon mixing ratios vs. time for Smoky Mountain and
Washington State . 48
VI
-------
FIGURES
Number Page
1 Map showing the Tennessee study site in the Great Smoky
Mountains National Park .................... 6
2 Schematic diagram of system for automatic, continuous
monitoring of N20, F-12, F-ll, CHaCCla and CC14 ........ 7
3 A typical set of chromatograms for the Smoky Mountain site
for 9/24/78 at 2300 EOT. Quantisation for F-ll and F-12
was obtained by use of integrated peak areas ......... 9
4 Vacuum system for sample injection for light and heavy
hydrocarbon analyses ..................... 16
5 Chromatogram of the light hydrocarbons ............. 18
6 Chromatogram of heavy hydrocarbons ............... 19
B-l Nitrous oxide mixing ratio as a function of time for the
Smoky Mountain site ...................... 48
B-2 Nitrous oxide mixing ratio as a function of time for the
eastern Washington state site ................. 49
B-3 Fluorocarbon 12 mixing ratio as a function of time for the
Smoky Mountain site ...................... 50
B-4 Fluorocarbon 12 mixing ratio as a function of time for the
eastern Washington state site ................. 51
B-5 Fluorocarbon 11 mixing ratio as a function of time for the
Smoky Mountain site ...................... 52
B-6 Fluorocarbon 11 mixing ratio as a function of time for the
eastern Washington state site ................. 53
B-7 Carbon Tetrachloride mixing ratio as a function of time for the
Smoky Mountain site ...................... 54
B-8 Carbon Tetrachloride mixing ratio as a function of time for the
eastern Washington state site ................. 55
-------
Number Page
B-9 Methyl chloroform mixing ratio as a function of time for the
Smoky Mountain site 56
B-10 Methyl chloroform mixing ratio as a function of time for the
eastern Washington state site 57
-------
TABLES
Number Page
1 Precision of Replicate Standard Measurements 10
2 Halocarbon Standard (MKS-1) Mixing Ratios (vs. KIT Series,
6/12-13/78) 11.
3 WSU Standard vs. EPA Standard (11:00, 9/25/78) 12
4 Blank Run on WSU Halocarbon Sampling and Analysis System
(11:00, 9/22/78) 13
5 Hydrocarbon Sample Collection and Storage Times 14
6 Comparison of Results of Equivalent Samples Analyzed by
WSU and EPA 20
7 Smoky Mountain "Background" Mixing Ratios 21
8 Tennessee Aircraft Sample Results 22
9 Hydrocarbon Levels, /^g/m^ 24
10 Diurnal Variation of Nitrous Oxide 28
11 Time of Occurrence and Maximum Levels Attained for Major
Methyl Chloroform Peaks 32
12 Halocarbon Levels During Rainstorm 35
13 Ambient "Background" Mixing Ratios for Western Versus
Eastern United States - September, 1978 35
14 TNMHC and Isoprene Results for Hydrocarbon Grab Samples ... 37
15 Lifetimes for OH Attack 39
16 Lifetimes for Ozone Attack 40
A-l Halocarbon Continuous Monitoring Data 43
-------
ACKNOWLEDGEMENTS
It is with deep regret that we report the death of David E. Harsch.
Mr. Harsch was killed in the crash of the Washington State University air
pollution monitoring plane on August 23, 1979 in New York while on an air
research mission.
We wish to thank Mr. Robert Watkins, a research technologist in the
air pollution section at WSU, for performing the hydrocarbon analyses.
We appreciated the help and support of the EPA field personnel,
under the supervision of Mr. Robert K. Stevens, during the course of the
field work.
-------
SECTION I
INTRODUCTION
Under the auspices of the EPA Regional Field Studies Office, a field
program was conducted in the fall of 1978 which included the participation
of the Air Pollution Research Section at Washington State University. The
field study served two purposes. It was an adjunct to the VISTTA program
(Visibility Impairment due to Sulfur Transportation and Transformation in
the Atmosphere). The field work also served as preparation for the U.S.
delegation to the U.S./U.S.S.R Joint Science Exchange Program, the
Abastumani Forest Aerosol Experiment - 1979. Thus, the principal objective
of the field work was to study the occurrence of natural aerosols as they
relate to visibility and total particulate loading.
From September 21 through September 26, 1978, the Air Pollution and
Resources Section at Washington State University (WSU) participated in this
ambient monitoring program in the Great Smoky Mountain National Park,
Tennessee. During this period, WSU provided continuous ambient monitoring
(one discrete sample each 40 minutes) of nitrous oxide (N20), dichlorodi-
fluoromethane (CCl2F2)> fl uorotrichloromethane (CC13F), methyl chloroform
(CH3CC13), and carbon tetrachloride (CCl/j). WSU also collected whole air
samples for hydrocarbon analyses (12 six-liter stainless steel sample can-
isters) and air shipped the samples to Pullman, Washington for analysis.
The purpose of the halocarbon measurements was to ascertain when the air
-------
mass at the the study site was influenced by anthropogenic emissions. The
gaseous hydrocarbons were measured because the naturally occurring gaseous
hydrocarbon plant emissions (such as isoprene and the terpenes) are postu-
lated as the precursors of the secondary natural aerosol.
Personnel from the Environmental Protection Agency Atmospheric Instru-
mentation Branch, headed by Mr. Robert K. Stevens, also participated in the
monitoring program. EPA equipment was housed in an instrumented van with
experiments consisting of various particulate matter collections, hydrocar-
bon grab samples in Teflon bags, wind speed and direction, carbon monoxide,
ozone, nitrogen oxides, and light scattering (nephel ometer).
The WSU instrumentation, experimental procedures, and analytical re-
sults for both the continuous measurements of ^0 and halocarbons and the
discrete gaseous hydrocarbon determinations are reported and discussed.
Interpretations of halocarbon "events" or excursions above the general
background are discussed in context with air mass trajectory data and 63,
bscat> and CO measurements acquired by EPA. The hydrocarbon data are com-
pared to similar samples collected by Ms. Sarah Meeks, which were shipped
to Houston for analysis by EPA personnel under the supervision of Mr.
William Lonneman. The complete WSU data set for halocarbons and ^0 are
included in Appendices A and B. Appendix A contains the tabulated mixing
ratios for halocarbons and ^0, while Appendix B presents the halocarbon
and N20 data graphically. Appendix B also presents continuous halocarbon
monitoring data obtained during the same time period from an ambient moni-
toring station in eastern Washington state. Table 9 presents the hydro-
carbon data.
-------
SECTION 2
CONCLUSIONS
The ambient air of the Smoky Mountains National Park in Tennessee was
monitored for one week during September, 1978 for four halocarbons (CF2C12.
CFC13, CC1 4 and CH3CC13), nitrous oxide and gaseous hydrocarbons including
isoprene and several terpenes. A continuous, automatic halocarbon and
nitrous oxide gas chromatograph with an electron capture detector was
successfully field tested during this time. Twelve whole air samples were
analyzed for gaseous hydrocarbons. Comparison with the results of samples
collected and analyzed simultaneously by EPA was good. Two samples were
collected via aircraft above the site and verified the halocarbon results
measured concurrently at the ground monitoring site.
Nitrous oxide levels remained stable during the study period with no
observations of localized sources nor diurnal behavior consistent with
ground uptake or release. The average value of 328 ppb was consistent
with tropospheric values measured worldwide by WSU since 1975.
F-12 (CF2C12) did show excursions during the study period. The average
lowest mixing ratio observed at the Tennessee site (^305 ppt) was higher
than concurrent measurements at a clean air, background site in eastern
Washington state (^282 ppt). This observation is explained by regional air
pollution typically observed for the eastern half of the United States.
F-ll and CH3CC13 also exhibited elevated "background" levels of ^182 ppt
-------
and 140 ppt, respectively, compared to 172 ppt and 122 ppt, respectively, in
eastern Washington state. F-ll often experienced excursions above the
182 ppt level while CH3CC13 was almost constantly exhibiting extreme eleva-
tions. Carbon tetrachloride suffered the least number and magnitude of
excursions of any of the measured halocarbons. The lowest levels of each of
the halocarbons correlated with a rainstorm. However, the excursions of the
compounds did not correlate with each other and the specific sources contri-
buting to these excursions were not identified. There appeared to be a
gradual increase in the "low" levels observed during the course of the next
few days following the rainstorm. Specific excursions could not be accomo-
dated with twenty-four hour airmass trajectory data, presumably because
localized sources were so diverse and widespread. The halocarbon monitoring
did show that the sampling site was under continuous influence of anthropo-
genic sources except for a brief few hours one night.
Isoprene was observed in each of the hydrocarbon samples ranging from
a trace (O.I yg/m3) to 11.6 yg/m3. Four terpenes (a-pinene, 3-pinene,
myrcene and A3-carene) were observed intermittently. The four terpenes
identified in these samples have been identified as the major emissions
from various pine species. Isoprene contributed from less than 1% to
13% of the total non-methane hydrocarbons observed in the samples. The
low concentrations observed for the biogenic hydrocarbons do not
necessarily imply a small contribution to the aerosol burden because of
the diffuse nature of the emissions coupled with the short lifetimes
expected against ozone and hydroxyl radical attack.
-------
SECTION 3
INSTRUMENTATION AND EXPERIMENTAL PROCEDURE
CONTINUOUS HALOCARBON MEASUREMENTS
The EPA and WSU monitoring site was established in a clearing near the
f~! ".
Elktnont campground in the Great Smoky Mountain National Park (^35.7° N lat-
itude, ^83.4°W longitude),-seven miles southwest of Gatlinburg, Tennessee.
Gatlinburg is approximately 40 miles southeast of Knoxville. See map in
Figure 1. The EPA equipment was housed in an air-conditioned mobile van,
and certain of the aerosol collection equipment was assembled in the field
nearby. WSU equipment was housed in a tent ^10 m from the EPA van at a
location which was not generally downwind of the EPA equipment. However,
the coolant used in the van air-conditioning equipment was "Freon 22" (CHF2C1),
which was not being measured. The F-12 used to calibrate the nephel ometer
was removed from the site after use. A small stream ran through the area,
separating the site from a moderately-traveled paved road to the east. One
of the major highways through the park was about one mile to the north.
Prevailing winds were generally light from the northwest quadrant, and the
weather was generally fair with a rain shower during the night of September
22.
The WSU equipment used for continuous halocarbon and N20 sampling and
analysis is depicted in Figure 2. A 1/4" copper tube extended out the top
of the tent terminating approximately 3 m above ground. Ambient air was
continuously pumped via a Metal Bellows Model MB-41 pump, into one inlet
-------
TENNESSEE
[NOXVILLE
SEVIERVILLE
PITTMAN CENTER
GATLINBURGy
ELKMONT
CALDERWOODx^
DAM ~~^
GREAT SMOKY MOUNTAINS
NATIONAL PARK
y
SCALE
0 5
MILES
YSON DAM
Figure 1. Map showing the Tennessee study site in the Great Smoky Mountains National Park.
-------
1. 1/4" copper tubing.
2. Metal Bellows Model MB41 pump.
3. Bypass vent.
4. Carle Model 2011 4-port switching valve, one port blocked.
5. Carle Model 4700 Valve Minder selects either sample or standard by switch-
ing the 2011 valve.
6. Carle Model 4100 Valve Minder controls the load/inject switching
of the two 5518 valves.
7. Carle Model 4200 Valve Actuators.
8. Halocarbon standard cylinder (MKS-I).
9. Matheson Model 3100 regulator.
10. Carrier gas (95% Ar/5% CH4).
11. Gas pressure regulator.
12. Perkin-Elmer Model 3920B dual channel ECD gas chromatograph.
13. Carle Model 5518 sampling valves.
14. 1/4" by 10' stainless steel column of 10% SF-96 on 100/120 mesh
Chromosorb W, 5-ml sample loop.
15. 1/8" by 6' stainless steel column of Porasil B (100/120 mesh),
1-ml sample loop.
16. 63Ni electron capture detectors.
17. Hewlett Packard Dual Channel Model 7128 strip chart recorder.
18. Hewlett Packard Model 3380 Recording Integrator.
Figure 2. Schematic diagram of system for automatic, continuous monitoring
of N20, F-12, F-ll, CH3CC13 and CC14-
-------
of the four-port switching valve. Air from the secondary standard cylinder
was allowed to slowly bleed into the other inlet of the valve. A valve
actuator selected either ambient air or standard to flush the sample loops
of the gas chromatograph. Sample air and standard were run alternately
with a 20-minute analysis time per injection or a sample run every 40 min-
utes.
A Carle Model 4700 valve minder on a 20-minute cycle time controlled
when an injection was made onto the GC columns. The valve minder signaled
two Carle Model 4201 valve actuators, each of which operated a Carle Model
5518 six-port gas sampling valve. The 20-minute cycle time consisted of
two minutes in the "load" position to allow sample (or standard, depending
on the stream selected ten minutes earlier by the Carle 2011 switching
valve) to flush through the gas sample loops; and then 18 minutes in the
"inject" position to allow carrier gas to flush the contents of the sample
loops through the columns of the gas chromatograph.
The carrier gas utilized was 95% argon/5% methane, with an isothermal
oven temperature of 50°C and with the electron capture detectors operating
at 350°C at a standing current of 3.0 nanoamperes. The two columns used
were a 1/8" OD x 6 ft stainless steel column of 100/120 mesh Porasil B for
separation of ^0, F-12, and F-ll, using a 1 ml sample loop volume; and a
1/4" OD x 10 ft stainless steel column of 10% SF-96 on 100/120 mesh Chromo-
sorb W for F-12, F-ll, CH3CC13, and CC14, using a 5 ml sample loop volume.
A typical strip chart trace from the dual-pen recorder is shown in Figure
3. Because of the relatively rural location of the sampling site, other
short-lived anthropogenic halocarbon species such as trichloroethylene
(CHCl=CCl2) and tetrachloroethylene (CCl2=CCl2)» which commonly appear
-------
20-
16-
12-
8-
4-
Inj.-
10% SF-96 on 100/120
MESH CHROMOSORB
W
IOftxl/4in(O.D.)
COLUMN
SAMPLE SIZE, 5ml
PORASIL B
100/120 MESH
*F-I2 339ppt
in sufficient amount to quantitate. The duplicate outputs of F-12 and F-l
results served as a back-up in case of recording equipment failure and as
verification for the "event" excursions above baseline mixing ratio. For
the first two days of the monitoring period, a failure of the N20 amplifie
precluded the collection of NgO, F-12, or F-ll data from the Porasil colun
Outputs from the gas chromatograph were recorded in duplicate. The
halocarbon signal was fed to one channel of a dual-pen Hewlett-Packard Moc
7127 strip chart recorder and also to a Hewlett-Packard Model 3380 recordi
integrator. The NgO electrometer output was displayed by the second chanr
-------
of the recorder and also fed into a Perkin-Elmer Model M-l computing inte-
grator. The integrators were switched on and off by electrical signals
from the Carle 4700 valve minder. Precision of the recorded responses for
a series of standard runs are given in Table 1. Much of the variation is
likely due to changes in detector sensitivity with time. ^0 quantisation
was obtained from peak areas provided by the M-l integrator. Mixing ratios
for the remaining halocarbons were calculated from the 7127 strip chart
peak heights of the Chromosorb column, except peak areas were used when
baseline excursions, recorder failure or offscale peaks during "event"
episodes precluded use of peak heights.
TABLE 1. PRECISION OF REPLICATE STANDARD MEASUREMENTS
Compound
F-12
F-ll
CH3CC13
CC14
Measurement
3380
3380
7127
3380
3380
7127
3380
3380
7127
3380
3380
7127
peak
peak
peak
peak
peak
peak
peak
peak
peak
peak
peak
peak
height*
area*
height"*"
height*
area*
height"1"
height*
area*
height*
height*
area*
height*
N
25
25
7
25
25
7
25
25
25
24
25
25
X
42.5
25744
128.3
66.5
82780
197.4
5.9
19000
38.3
15
57916
93
mm
0
S
.8
au11 879
mm
mm
au
mm
mm
au
mm
mm
au
mm
1
0
1
3
0
1
1
0
1
.4
.5
602
.6
.3
6
.8
«S/X
mm
au
mm
mm
au
mm
mm
au
mm
.295mm
274
au
0.594mm
1.
3.
1.
0.
1.
1.
5.
8.
4.
2.
2.
0.
9
4
1
8
9
8
8
2
8
0
2
6
* 1540 EST on 9/22/78 to 0940 EST on 9/23/78
f 1200 EST on 9/22/78 to 1500 EST on 9/22/78
11 area units
10
-------
The secondary halocarbon and ^0 standard used was a cylinder (#MKS-I)
of ambient air collected cryogenically in eastern Washington state. It was
standardized by comparison to six long-term ambient air standard cylinders;
these six cylinders having been standardized by means of static dilution of
commercial gas mixtures. Table 2 gives the halocarbon and nitrous oxide
mixing ratios determined for the standard. Sample air mixing ratios were
calculated by comparison of area or peak height responses to those of the
standard runs immediately preceding and following the sample run. The
accuracy of the standards is estimated to be +10% for F-ll and F-12 and ±20%
for Creels, CC14 and N20 (Rasmussen and Pierotti, 1978; Rasmussen, 1978).
TABLE 2. HALOCARBON STANDARD (MKS-I) MIXING RATIOS
(vs. KLT SERIES. 6/12-13/78)
Compound Mixing Ratio
F-12 265.2 ppt
F-ll 159.6 ppt
CH3CC13 117.8 ppt
CC14 139.2 ppt
N20 325.2 ppb
A standard supplied by EPA was analyzed by WSU to compare the quanti-
tation between laboratories. The results, given in Table 3, show agreement
within 4% for F-12, a 20% discrepancy for F-ll, an EPA contamination prob-
lem for CH3CC13, and a 9% discrepancy for N20.
11
-------
TABLE 3. USU STANDARD vs. EPA STANDARD (11:00. 9/25/78)
Sample
WSU Standard (MKS-I)
EPA Standard (claimed)
Analysis #1 of EPA Standard
Analysis #2 of EPA Standard
EPA/WSU (%)
F-12
(ppt)
265.2
320
333
332
96
F-ll
(ppt)
159.6
245
306
308
80
Mixing Ratio
CH3CC1 3
(ppt)
117.8
28600
32100
--
CC14
(ppt)
139.2
<2
<2
N20
(ppb)
325.2
300
331
328
91
Two types of blank checks were made of the system. The Metal Bellows
MB-41 sampling pump was blanked in the laboratory prior to the study and
found to be clean for all of the hal ocarbons of interest. The only major
contaminant found in the pump was Fluorocarbon-113 (CC12F-CC1F2) and only
when the pump was warm. Sampling system blanks were also run by attaching
a pressurized can of "zero gas", containing negligible amounts of the spec-
ies being checked, to the sample inlet line (pump not running). This air
was allowed to flow through the pump and analytical system at a typical
sampling flow rate (100 ml/min). A halocarbon analysis was performed using
this gas to fill the sample loops. Table 4 presents results of a blank run
on the Smoky Mountain sampling system. The lower limits of detection (2X
the baseline noise) of the method for the halocarbons and for N20 are also
given.
12
-------
TABLE 4. BLANK RUN ON WSU HALOCARBON SAMPLING AND ANALYSIS SYSTEM (11:00,
9/22/78)
Mixing Ratio
Sample
Blank
Detection Limit
F-12
(ppt)
<3
3
F-ll
(ppt)
<2
2
CH3CC13
(ppt)
3
3
CC14
(ppt)
<2
2
N20
(ppb)
<2
2
HYDROCARBON ANALYSIS
Twelve hydrocarbon whole air samples were collected in six-liter
stainless steel canisters pressurized to about 20 psig and air shipped to
WSU laboratories in Pullman, Washington, for analysis. The samples were
filled by means of a Metal Bellows MB-41 pump at a site about 50 feet, east
of a power pole, which stood next to the tent at the Elkmont Smoky Mountain
site. This location was about 20 feet west of the edge of the clearing and
about 50 feet from the stream. A 5-foot vertical piece of 1/4" stainless
steel tubing was used as a sample inlet and the pump was connected to the
can by means of a short length of 1/4" stainless steel tubing. Care was
taken to avoid trampling of vegetation in the area during the sample proce-
dure.
WSU has utilized the stainless steel canisters for several halocarbon
and hydrocarbon sampling studies in the past with excellent stability for
halocarbons and acceptable stability demonstrated for most hydrocarbon
species (Harsch, 1979). The canisters are equipped with two inlet valves
so they can be flushed with several exchange volumes of sample prior to
filling. Canisters are filled with a clean or "zero" gas prior to field
sampling to verify cleanliness for the species of interest, and transported
13
-------
to the sampling site with a positive pressure of this gas to eliminate con-
tamination from outside air.
The interior surfaces of the canisters are electropolished utilizing
the patented "Summa" process of the Molectrics Corporation. Saturated hy-
drocarbon species are, in general, very stable in the canisters. Unsatur-
ated and more reactive hydrocarbons, including the terpene compounds, are
not completely stable and will show decreases with time. For this reason,
the samples were air shipped immediately after collection and analyzed as
soon as possible after they reached Pullman. Table 5 presents data on the
hydrocarbon samples which were taken during the study and includes elapsed
time between sampling and analysis. Samples were typically analyzed within
TABLE 5. HYDROCARBON SAMPLE COLLECTION AND STORAGE TIMES
Canister
Sample Taken
Sample Analyzed Elapsed Sample
Number
66
38
42
113
99
247
164
107
246
51
173
57
Time (EOT) Date Time (PDT) Date Time, Hours
1615
1620
1335
0640
1650
0640
1650
0635
1515
0650
1245
0815
9/21/78
9/22/78
9/23/78
9/24/78
9/24/78
9/25/78
9/25/78
9/26/78
9/26/78
9/27/78
9/27/78
9/28/78
16:00
10:00
12:00
14:00
14:00
16:00
14:00
16:00
14:00
12:30
15:30
17:00
9/26/78
9/25/78
9/26/78
9/26/78
9/27/78
9/28/78
9/29/78
9/29/78
9/30/78
9/30/78
9/30/78
9/30/78
117
69
74
58
96
85
96
85
98
81
78
60
Al iquot, ml
500
300
400
220
400
400
500
200
500
500
500
500
14
-------
3 to 4 days of sampling. While not optimum, degradation should not have
been significant during that time. Nevertheless, terpene concentrations
reported should be considered as lower limits to the actual ambient concen-
trations. A light hydrocarbon analysis (the C2-Cs fraction) and a heavy
hydrocarbon analysis (the CS-CIQ fraction) were performed on each sample
canister.
Light Hydrocarbon Analysis
A Hewlett-Packard Model 5700 gas chromatograph with a flame ionization
detector was used for the analysis. The column was a 20 ft x 1/16" O.D.
stainless steel tube packed with Durapak N-octane with helium carrier gas
at 6 ml/min. Integration was performed by a Hewlett-Packard Model 3380A
integrator. Sample transfer into an external freezeout loop was made with
a vacuum system which is connected to the gas chromatograph by means of a
6-port Carle Model 5518 sampling valve. Figure 4 shows the vacuum system
used to pull the samples through the freezeout loop, which is immersed in
liquid oxygen. Typically, 500 ml of sample were transferred. The pump was
used to evacuate the vacuum chamber. The vacuum sample inlet line was
attached to the Carle valve exit port on the GC. The vacuum gauge metered
the sample volume through the loop. Oven conditions were -70°C to +65°C at
16°/min with no initial temperature hold.
Heavy Hydrocarbon Analysis
A Perkin-Elmer Model 990 gas chromatograph with a flame ionization de-
tector was used for the analysis. The column was a 30 M SE54 glass capil-
lary with helium carrier gas at 1 ml/min. Oven conditions were -50°C to
+80°C at 4°/min without initial temperature hold. Integration was made with
15
-------
VACUUM SYSTEM
for sample injection
OFF
SAMPLE
FLOW ADJ.
FRONT VIEW
MPLE
BACK VIEW
Figure 4.
analyses.
Vacuum system for sample injection for light and heavy hydrocarbon
16
-------
a Perkin-Elmer Model-1 integrator. Sample transfer was by the same means as
with the light hydrocarbon analysis.
Identification for lights and heavies was by retention time comparison
with known standards. Calibration was achieved by measuring instrument
response to a concentration of neo-hexane (2,2-Dimethylbutane) in air
(0.2 ppm).
As an example of the results of the hydrocarbon analysis techniques,
a typical ^2 to ^6 "light" hydrocarbon analysis (see Figure 5) and a
typical C$ to C]Q "heavies" analysis (see Figure 6) are shown for one of
the samples collected in the Smoky Mountains. Column conditions and
temperatures are given in the figures.
-------
UJ
CO
z
o
Q.
CO
UJ
ir
TENNESSEE
I CAN 42
1335 EOT
9-23-78
500 ml
l/8"x 10'PARAPAK n-OCTANE
(60/80 MESH)
FLOW RATE 6ml/min He
-70°C TO 65°C l6°C/min
WITH 2min HOLD
14
0
2 <
\ 6
8
10 12
_l L_
14 16
18 20
22 24
26 28
30 32
34 36
i -i
38 41
i 1
TIME (MIN)
Figure 5. Chromatogram of the light hydrocarbons. 1 = ethane; 2 = ethylene;
3 = acetylene; 4 = propane; 5 = propylene; 6 = isobutane; 7 = n-butane; 8 =
1-butene + propyne; 9 = 1,3-butadiene + isobutane; 10 = isopentane + t-2-butene;
11 = n-pentane; 12 = 2-methylpentane + t-2-pentene + 3-methyl-1-butene; 13 =
3-methylpentane; 14 = isoprene.
18
-------
14
TENNESSEE
CAN 42
1335 EOT
9-24-78
400 ml
SE30 GLASS CAPILLARY COLUMN (30m)
FLOW RATE 0.9 ml/min He
-50°C TO 80°C
-------
Table 6 compares the results of two pairs of concurrently collected
samples; one of each pair collected and analyzed by WSU and the other
collected and analyzed by EPA. The data in Table 6 concerning the EPA
results have been adapted from a recent report (Arnts and Meeks, 1979).
There is passable comparability between the two collection and analysis
procedures.
TABLE 6. COMPARISON OF RESULTS OF EQUIVALENT SAMPLES ANALYZED BY WSU AND
EPA*.
Laboratory:
Collection Time (EOT):
WSU EPA
0640 0600-0800
WSU EPA
1650 1600-1620
1 Ethane
2 Ethyl ene
3 Acetylene
4 Propane
5 Propylene
6 Isobutane
7 n-Butane
8 1-Butene & propyne
9 1 ,3-Butadiene & isobutene
10 Isopentane & t-2-butene
11 n-Pentane
12 2-Methylpentane & t-2-pentene
& 3-methyl-l-butene
13 3-Methylpentane
14 Isoprene
15 Toluene
16 a-Pinene
17 B-Pinene
18 Myrcene
19 A^-Carene
Other Identified Compounds
Subtotal
Unidentified Components
TNMHC
6.6
0.3
]6. 5
0.5
1.2
2.6
t
0.7
1.7
1.1
t§
t
2.6
8.7
1.4
1.6
35.5
19.5
55
6.8
1.3
1.9
6.1
0.3
1.8
2.5
NR1I
3.5
2.2
1.2
0.7
1.6
1.6
NR
NR
NR
20.9
52.4
5.8
58.2
7.5
1.4
7.0
2.1
0.5
1.2
2.7
0.2
0.3
2.1
1.4
0.5
t
4.6
1.8
0.8
10.6
44.7
30.3
75
7.5
0.9
1.7
4.9
0.3
1.1
2.4
NR
2.5
0.3
0.5
0.5
3.3
11.1
0.9
NR
NR
NR
21.9
59.8
16.5
76.3
*Sample Collection Date: 9/25/78, Concentrations: yg/m3, EPA Results
Adapted From Arnts and Meeks, 1979.
'''Below detection limit.
%ot reported
§Trace amount
20
-------
SECTION 4
ANALYTICAL RESULTS
HALOCARBON RESULTS
The entire set of halocarbon and N20 data collected during the field
study is listed in Appendix A. No nitrous oxide data were available for
the first two days of the study period because the detector amplifier
failed.
Table 7 presents the average mixing ratios observed for periods when
the levels were nearly the lowest observed during the monitoring project.
These average values are representative of what was judged to be ambient
"background" mixing ratios for the halocarbon compounds at the study site.
A further discussion of the implications of this table follows in Section
5.
TABLE 7. SMOKY MOUNTAIN "BACKGROUND" MIXING RATIOS
Compound Time Sampled "X (ppt) S.D. N %S/I
F-12 1420 to 2020 EOT on 9/22/78 305 3.6 10 1.2
F-ll 1420 to 2020 EOT on 9/22/78 183 2.8 10 1.5
CH3CCl3 1620 to 2020 EOT on 9/21/78 180 4.7 7 2.6
CC14 1620 to 2340 EOT on 9/21/78 148 2.8 12 1.9
N20 9/23/78-9/26/78 328 (ppb) 4.1 87 1.2
21
-------
Appendix B contains plots of mixing ratios at the Smoky Mountain study
site as a function of time during the monitoring period for ^0 (Figure B-l),
F-12 (Figure B-3), F-ll (Figure B-5), CCU (Figure B-7), and CH3CCl3
(Figure B-9). During the same time period a similar ambient halocarbon
monitoring station was in operation in Eastern Washington State. Appendix B
also contains concurrently collected data from this site for comparison to
the Smoky Mountain site for each compound (^0, Figure B-2; F-12, Figure
B-4; F-ll, Figure B-6; CC14, Figure B-8; and C^CCla, Figure B-10).
AIRCRAFT SAMPLES
Two aircraft samples were collected on 9/26/78 courtesy of Dr. Tom
Stephens of Virginia Polytechnic Institute. The Cessna aircraft used a
Metal Bellows Model 41 pump to fill the stainless steel canisters directly
above the ground site. Table 8 shows the results of these analyses and com-
pares them with the mixing ratios measured concurrently at ground level. Both
of the samples were collected within the mixing layer. We had hoped to col-
lect samples above the mixing layer in order to compare with the lowest
mixing ratios observed at ground level. We would have expected high altitude
samples above the site to more closely agree with the average values measured
TABLE 8. TENNESSEE AIRCRAFT SAMPLE RESULTS
Canister
Number
93
28
Altitude
(ft MSL)
MOOO
^2800
ground
Time
(EOT)
1420
1430
1420
F-12
(ppt)
312
314
322
F-ll
(ppt)
__
197
205
CH3CC13
(ppt)
201
207
205
CC14
(Ppt)
165
167
160
22
-------
in eastern Washington state. Because they were from the mixed layer the
aircraft samples were similar to the ground level samples within the preci-
sion of analysis. This is not unexpected.
HYDROCARBON RESULTS
The complete analysis record for the hydrocarbons in the twelve sample
canisters is contained in Table 9. Nineteen separate components were ident-
ified and individually quantified. The amount of the remaining gaseous
hydrocarbons (the difference between TNMHC and the individually identified
components) is also indicated as unidentified components. The nineteen
components typically accounted for about 2/3 of the TNMHC with a range from
54 to 83%.
23
-------
TABLE 9. HYDROCARBON LEVELS, yg/m3'
Date
Collection Time (EOT)
Storage Time (Hrs. )
Can No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Subtotal
Unidenti
TNMHC
Ethane
Ethyl ene^'
Acetyl ene
Propane
Propylene
Isobutane
n-Butane
1 -Butene
& Propyne
1 ,3-Butadiene &
Isobutene
Isopentane &
t-2-Butene
n-Pentane
2-Methyl pentane &
t-2-Pentene &
3-Methyl-l-butene
3 -Methyl pentane
Isoprene
Tol uene
a-Pinene
3-Pinene
Myrcene
A-3-Carene
fied Components
9/21
1615
117
66
3.9
3.6
2.6
2.2
1.3
1.0
3.1
0.2
0.4
3.3
1.6
1.9
0.5
10.1
3.0
5.8
2.1
1.4
2.5
50.5
26.5
77
9/22
1620
69
38
3.4
1.4
]3.0
1.4
0.4
1.1
0.5
0.7
0.9
0.1
t§
t
3.5
4.7
1.9
___
3.5
26.5
25
9/23
1335
74
42
9.5
2.9
6.7
3.3
0.9
2.5
6.0
0.2
0.5
4.9
2.5
1.5
1.0
11.6
3.5
5.3
1.9
1.9
66.6
32.4
99
9/24
0640
58
113
...t
1.1
]13.7
0.7
1.0
2.0
0.2
0.2
2.0
1.1
0.5
0.4
3.5
5.4
_
13.9
42.6
24.4
67
9/24
1650
96
99
1.2
]4.1
0.5
0.8
2.0
0.1
0.4
1.8
1.1
0.3
0.2
3.2
1.6
1.4
2.0
1.6
22.3
12.7
35
9/25
0640
85
247
6.6
0.3
]6.5
0.5
1.2
2.6
_ _ _
0.7
1.7
1.1
t
t
2.6
8.7
1.4
1.6
_
35.5
19.5
55
continued
24
-------
SECTION 5
DISCUSSION
HOW REMOTE WAS THE STUDY SITE?
In evaluating the sampling results of this Smoky Mountain program the
most informative conclusions occur for the halocarbons when the values are
considered in terms of a low "baseline" concentration and excursions above
this baseline. Plots of the halocarbon sampling data are shown in Figures
B-3, B-5, B-7 and B-9. In these figures the majority of the sampling values
fall in the lower part of the concentration range covered by the sampling
data (rather than being evenly distributed around a central value, for
example). We have empirically approximated a baseline concentration for the
five continously recorded compounds. These were given in Table 7: F-12,
305 ppt; F-ll, 183 ppt; CH3CC13, 180 ppt; CC14, 148 ppt; and N20, 328 ppb.
While these values are useful for this present experiment, no claim can be
made for their being applicable to a longer time period or a regional base-
line or background estimate.
Nitrous Oxide
There were no excursions in the observed levels of nitrous oxide during
the study period. Table A-l lists the actual N20 data set while Figure B-l
shows the plot of N20 versus time. Eighty-seven measurements gave an average
N20 mixing ratio of 328 ppb with a standard error (standard deviation of the
26
-------
mean) of 0.4 ppb. The values ranged from 318 ppb to 341 ppb. The data were
not corrected for differences in water vapor content, a correction which has
been recommended by Goldan, et al. (1978). This average value is the same
as that measured by WSU in tropospheric air samples all over the world during
the last four years (see for example Cronn, et al., 1976; Cronn, et al.,
1977a; Cronn, et al., 1977b; Cronn and Robinson, 1979; Rasmussen, et al.,
1976). Figure B-2 shows NgO mixing ratio versus time data for the long-term
continuous halocarbon and nitrous oxide monitoring site, which is set up in
eastern Washington state. The same time period is depicted as for the Smoky
Mountain data in Figure B-l. There is no significant difference in the
behavior or average values at the two sites ("x = 330 ppb with a 0.4 standard
error for the Washington state site from 9/21/78 through 9/26/78). The lack
of any observed elevated values of N20 indicates that no substantial sources
such as the burning of coal and fuel oil or automotive catalytic converters,
as identified by Weiss and Craig (1976), had an influence on the site. This
is in spite of the observation on at least one evening of the smell of camp-
fires from the nearby campsite.
Since the sample inlet line was approximately 3 m above the soil, it
was not expected that a diurnal variation would be observed due to uptake
or release of NgO by the soil. However, a check of this possibility has
been made by averaging the data points between the local hours of 0000-0600,
0600-1200, 1200-1800, and 1800-2400. This information is tabulated in Table
10. The average values decrease from the 0000-0600 data set through the
1800-2400 set. The differences are not always significant, but the 0000-0600.
average is higher than either the 1200-1800 or the 1800-2400 periods. These
results for samples collected 3 m above the soil surface are in keeping with
27
-------
the results of Cicerone, et al. (1978) for a sample inlet 0.5 m above the
surface, showing that the 1200-1800 hour time period sees significantly
lower N20 concentrations than does the period 0000-0600 hours.
TABLE 10. DIURNAL VARIATION OF NITROUS OXIDE
Time Period
(EOT)
0000-0600
0600-1200
1200-1800
1800-2400
Number of
Samples
20
19
21
27
Average
(ppb)
329.7
328.5
327.8
326.6
Standard
Error
0.8
0.9
1.0
0.7
Percentage Standard
Deviation
1.0
1.2
1.5
1.1
Fluorocarbon 12
In contrast to N20, F-12 did show excursions at various times during the
study period. Table A-l tabulates the F-12 data and Figure B-3 visualizes
the behavior at the sampling site as a function of time. There were 166
data points. The lowest F-12 mixing ratio calculated was 280 ppt. The high-
est value recorded was 532 ppt at 0900 EST on 9/26/78. The six values in
the 400-500 ppt range were 486, 428, 419, 414, 412 and 405 ppt. The average
of all F-12 data was 323 ppt (S.D. = 33 or 10%). As mentioned previously
in Table 7, however, a better estimate of the generally lowest F-12 values
was about 305 ppt. The lowest two quartiles of data gave an average of
304 ppt (S.E. = 0.85).
There was a significant difference between the average lowest mixing
ratio observed at the Smoky Mountain site and the average F-12 mixing ratio
28
-------
TABLE 9. (continued)
Date
Collection Time (EOT)
Storage Time (Mrs.)
Can No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Subtotal
Unidentified
TNMHC
Ethane
Ethyl ene^i
Acetyl ene
Propane
Propyl ene
Isobutane
n-Butane
1-Butene
& Propyne
1 ,3-Butadiene &
Isobutene
Isopentane &
t-2-Butene
n-Pentane
2-Methyl pentane &
t-2-Pentene &
3-methyl-l-butene
3-Methyl pentane
Isoprene
Tol uene
a-Pinene
3-Pinene
Myrcene
A^-Carene
Components
9/25
1650
96
164
7.5
1.4
7.0
2.1
0.5
1.2
2.7
0.2
0.3
2.1
1.4
0.5
t
4.6
1.8
0.8
10.6
44.7
30.3
75
9/26
0635
85
107
7.3
0.5
]7.4
0.4
1.1
2.3
t
0.3
1.7
0.9
t
t
0.3
1.2
3.3
26.7
5.3
32
9/26
1515
98
246
6.0
1.4
]«2
0.6
1.3
3.2
0.2
1.1
2.6
1.4
0.5
0.3
4.0
4.8
1.2
1.6
0.7
31.7
21.9
59
9/27
0650
81
51
4.4
0.5
]12.0
0.9
1.2
2.6
0.1
0.2
2.0
1.2
0.6
0.1
t
1.3
1.2
1.1
2.0
31.4
9.6
41
9/27
1245
78
173
4
1.8
1.9
3.9
0.6
0.8
2.2
0.1
0.2
2.0
1.1
0.4
t
1.6
3.5
1.0
1.0
26.1
16.9
43
9/28
0815
60
57
6
4.8
2.2
14.7
1.2
1.8
5.0
0.3
0.4
6.7
4.0
2.9
1.8
2.4
0.8
1.3
1.3
57.6
40.4
98
* Multiply by 1.5 to convert to ppbC.
^ Below detection limit.
^ Questionable due to water interference.
§ Trace amount.
25
-------
observed concurrently at the clean-air, background site in eastern Washington
state. The Smoky Mountain low average of about 304 ppt (S.E. = 0.85) was
significantly above the approximately 282 ppt (S.E. = 1.0) observed at the
so-called Klemgard site near Pullman, Washington. This observation is in
keeping with the usual expectation that regional air pollution impacts the
eastern half of the United States. Furthermore, while there were numerous
excursions above 304 ppt in the Smokies, Figure B-4 shows very little vari-
ation in the level of F-12 in the Pacific Northwest.
The increases above 304 ppt indicated periods when air at the study
site had been previously carried across a more localized source of F-12.
Periods of very high F-12 excursions occurred from 1940 EOT to midnight on
9/21/78; from 2040 EOT on 9/23/78 through 0940 on 9/24/78; and from 0820 EOT
to 0940 EOT on 9/26/78. The two values of 405 and 419 ppt on 9/22/78 are
attributed to F-12 release during calibration of the EPA nephelometer. Two
other periods with less dramatic fluctuations occurred between 0100 EOT and
1720 EOT on 9/23/78 and between 2220 EOT and 2340 EDT on 9/24/78.
Fluorocarbon 11
Like F-12, F-ll showed excursions at various times during the study
period above a "baseline" level. The distribution of F-ll as a function of
time is shown in Figure B-5. The range of values was from 165 ppt to 327 ppt
with a median value of 188 ppt. The next highest values observed were 309,
299, 293, 273, and 256 ppt. Table A-l lists the entire data set of 169
points. The average of all points was 197 ppt with a standard deviation
of 25 (13%). However, the average of the bottom two quartiles was 181 ppt
with a standard deviation of 5.6 or 3% and a standard error of 0.6. This is
very close to the "background" level reported in Table 7 of 183 ppt (S.E., =
29
-------
0.9), but is higher than the 172 ppt average from the 112 data points plotted
for the Washington state site in Figure B-6. The standard error of the
Klemgard data points from 9/21/78 through 9/26/78 was 0.5 which is a conse-
quence of the absence of significant elevations of F-ll at the Washington
state site. The difference between 172 ppt at the background monitoring
site in the western United States and 181 or 183 ppt in the Smoky Mountains
is again an indication of the regional impact of pollution on the eastern
half of the United States.
The excursions above the 183 level during the six-day study period fall
into two categories. During the first three days of the study, there were
small, gentle excursions. Beginning the night of the 23rd and 24th, and
continuing to the end of the monitoring period, there were more than a dozen
"spikes" with peaks exceeding 230 ppt. These "spikes" did not exceed 2 1/2
hours duration. If one discounts the two F-12 values of 405 and 419 ppt
which occurred during calibration of the nephelometer, the behavior of F-ll
and F-12 was generally similar during the first three days of the experiment.
The correlation coefficient for 83 data points was 0.55. The broad rise
and fall seen for F-12 on the 23rd showed up for F-ll, too. However, during
the second half of the project, F-12 exhibited only a couple of "spikes" of
the sort referred to for F-ll. As a consequence, the 81 data pairs during
this period had a correlation coefficient of only 0.43. The source of the
F-ll during these periods (and of the F-12 on the few occasions when this
component rose along with F-ll) is unknown. The surface winds were generally
0-5 mph and with variable direction during this time, and it seems likely
that the source was close by, perhaps at the sampling site itself, or at the
Elkmont campsite.
30
-------
Carbon Tetrachloride
The behavior of CC14 during the study period was intermediate between
the uniform level of ^0 and the significant excursions of F-ll and F-12.
The CC14 mixing ratio versus time is shown in Figure B-7. For comparison,
the similar data from eastern Washington state is shown in Figure B-8. In
general, the variability is about the same for the two sites, although the
Tennessee trace does exhibit a couple of larger changes than is evident
at the Klemgard site. There were 168 measurements of CC14 in Tennessee
(see the tabulation in Table A-l). The range was 136 ppt to 207 ppt with
the next highest values being 181, 180, 173, 171, 165 and 164 ppt. The
median of 148 ppt is close to the average of 149 ppt (S.D. = 8.7 or 5.8%).
The average only drops to 148 ppt with a standard deviation of 5.9 or 4%
if the six values which lie beyond two standard deviations are excluded.
Table 7 also gave an average 148 ppt (S.E. = 0.8) for the twelve measurements
from 1640-2340 on 9/21/78. This "background" level is lower than the range
of the 154 ppt average (S.E. = 0.4) of the 102 data points in Figure B-8,
possibly due to a calibration offset in the standards used at the two sites.
All of the measured mixing ratios for CC14 which exceeded 160 ppt occurred
during only four periods - with peak values centered at 1140 on 9/22/78,
2140 on 9/24/78, 1340 on 9/25/78 and 1100 on 9/26/78. The third and fourth
excursion periods coincided with the observation that highway maintenance
crews were repainting the centerlines on nearby roads. On Monday, September
25, the crew was working from Sugarlands to Elkmont between about 1300-
1400 EDT. On Tuesday, they continued along the river road west of Elkmont
between about 1000-11.00 EDT. A survey of the cans on the paint truck did
31
-------
not provide information about the ingredients of the containers. The well-
defined peak around midnight on Sunday, September 24, has no such potentially
identified source.
Methyl- Chloroform
By far, the most variable behavior for the halocarbons was observed
for methyl chloroform. The constant change in mixing ratio is seen in
Figure B-9. The values ranged from 99 ppt to 322 ppt. The median of the
165 data points was 186 ppt. The average was 192 ppt (S.D. = 39 or 20%).
The tabulation is given in Table A-l. During the afternoon time period
on 9/21/78 when the "background" levels of Table 7 were calculated for
CC14, CH3CC13 was in the middle of a moderate excursion so that the
average of the seven data points between 1620 and 2020 EOT was 180 ppt
(S.D. = 4.7 or 2.6%), nearly equaling the median value of 186 ppt. The
highest mixing ratios of the major peaks and the times of occurrence are
given in Table 11. The largest excursion(s) and the longest duration of
elevated values occurred on the 23rd.
TABLE 11. TIME OF OCCURRENCE AND MAXIMUM LEVELS ATTAINED FOR MAJOR METHYL
CHLOROFORM PEAKS
Maximum Mixing
Ratio of Peak
(ppt)
208
218
240
322
259
225
228
Time of
Maximum
(EOT)
2140
0740
1340
0140
1120
2020
2100
Date
9/21/78
9/22/78
9/22/78
9/23/78
9/23/78
9/24/78
9/25/78
32
-------
It is difficult to. make an assessment of a "background" level of CH3CC13
at the Tennessee site. The excursions are so numerous and so large that no
quiescent periods existed during the study period. The striking difference
in the behavior of CH3CC13 at the Smoky Mountain site and in rural eastern
Washington state can be seen by comparison of Table B-9 with Table B-10.
The 100 data points from the Klemgard site gave an average of 122 ppt with
a standard error of 0.5. The lowest period at the Tennessee site occurred
in the early morning of 9/22/78. The average of the six data points between
0140 and 0620 EOT was 109 ppt (S.E. = 3.0). These averages do differ, perhaps
partly because of excursions which occurred at the Washington state site.
The next lowest recorded mixing ratios in Tennessee, excluding this one
period, are in the 140+ ppt range.
The time dependent behavior of CH3CC13 does not correlate well with any
of the other halocarbons over the entire monitoring period. For instance,
the correlation coefficient between F-12 and CH3CC13 is 0.02. However, the
largest Cf^CCl3 peak occurred on the 23rd when both F-12 and F-ll showed
a gradual rise and fall and the peak on 9/21/78 occurred for all three of
these compounds. But the "spiking" behavior of F-ll (and occasionally
F-12) during the second half of the project was not exhibited by C^CCls.
The CC14 peak which occurred late on the 24th was mimicked by a CH3CC13
peak at the same time. The other rises in CC14 occurred during periods of
CH3CC13 elevation as well.
There has been a tremendous growth over the last few years in CH3CC13
use in the United States. As a result, the tropospheric burden has increased
on the order of 20% per year in the last two years. This compound is the
only one of the four halocarbons monitored in Tennessee which has an important
33
-------
tropospheric sink (hydroxyl radical attack). We have estimated the lifetime
at 6. 4 years (Cronn and Campbell, 1979). The large changes with time in the
CH3CC13 mixing ratio at the Smoky Mountain site are surprising. The behavior
is not at all that of a remote, rural clean air site like the Klemgard station.
The behavior, in fact, is much like that observed just weeks earlier in
Claremont, California. On that basis, we might expect that a substantial
source of CH3CC13 existed in the vicinity of the Tennessee site. The
variability of direction and low speeds of the wind during the six-day
project period could then account for the pattern observed for
METEOROLOGY AS A DETERMINANT OF HALOCARBON BEHAVIOR
It rained the night of the 21st. This rainfall coincided with the
period of the lowest observed values of F-12, F-ll, CC14 and CH3CC13 during
the project. Table 12 lists the average levels observed during the rain-
storm. Except for this brief period, the next lowest levels observed in
the Smoky Mountains exceeded the average values in the western United States.
Table 13 compares the weekly averages measured at the rural Washington state
site with the typical short-term low values observed in Tennessee which
were previously reported in Table 7. The high pressure systems over the
eastern United States during the September, 1978 field work allowed for the
regional pollution build-up which caused the differences depicted in Table
13 (Husar, 1979). The build-up of regional pollution is seen in the gradual
increase in the lowest levels observed following the rainstorm during the
early morning hours on September 22. F-ll "minima", for example, rose
from 169 ppt to 175-180 ppt early on 9/24/78 and then to 186-190 ppt by the
end of the measurement program. F-12 went from 284 ppt through 295-300 ppt
34
-------
TABLE 12. HALOCARBON LEVELS DURING RAINSTORM
Compound
F-12
F-ll
CC14
CH3CC13
Tennessee
Average
(ppt)
284
169
141
109
S.D. (%)
3.5 (1.2)
3.5 (2.0)
3.0 (2.1)
7.4 (6.8)
Number
7
8
8
6
Time Period
(EOT on 9/22/78)
0340-0740
0340-0820
0340-0820
0140-0620
TABLE 13. AMBIENT "BACKGROUND" MIXING RATIOS FOR WESTERN VERSUS EASTERN
UNITED STATES - SEPTEMBER, 1978
Compound
F-12 (ppt)
F-ll (ppt)
CH3CC13 (ppt)
CC14 (ppt)
N20 (ppb)
Mixing Ratio
Western U.S.
282
11
112
172
6
112
122
5
100
154
4
102
330
4
112
Eastern U.S.
305
4
10
182
3
10
180
5
7
148
3
12
328
4
87
X
S.D.
n
I
S.D.
n
I
S.D.
n
X"
S.D.
n
I
S.D.
n
35
-------
to about 300-305 ppt on 9/26/78. Likewise, CC14 began at 141 ppt and ended
at about 150 ppt. Obviously, the most dramatic example of this phenomenon
is methyl chloroform. From 109 ppt, the lowest burdens observed climbed
through 140 ppt at midnight on the night of September 23 to 166 in the early
morning hours on the 26th.
Superimposed on this regional scale impact were the excursions indica-
tive of transport of the various halocarbons from nearby sources. The
twenty-four-hour airmass trajectory data for Knoxville (Parkhurst, 1979) does
not indicate that the sampling site was directly downwind of either Knoxville
or Gatlinburg for any extended period of time. However, the winds were
generally light during the study period, and quite variable from day to day,
allowing advection of air from nearly all directions at least sometime during
the field work. Because of the light winds, the terrain effects on nearby
source emissions would have been important. But terrain effects were not
included in the airmass trajectories of Parkhurst. Thus the influence of
localized sources would not be expected to be accomodated by the airmass
trajectory data.
It is quite clear from the halocarbon monitoring that except for a
brief few hours during the early morning hours of September 22, the Smoky
Mountain sampling site never experienced clean air. As a result, it might
be difficult to study natural aerosol formation at this site due to the
difficulty of discriminating between naturally-formed aerosols and the an-
thropogenic contribution.
TERPENES AND OTHER GASEOUS HYDROCARBONS
Went (1960) first proposed the suggestion that blue hazes such as those
observed in the Smoky Mountains were due to the oxidation of gaseous plant
36
-------
the magnitude of TNMHC concentration and the time of day (morning or after-
noon) that the sample was collected. However, in general, higher isoprene
concentrations appeared in the afternoon samples; an observation consistent
with the fact that isoprene is a daytime, light-dependent emission (Tingey,
1978). Although the total non-methane hydrocarbon burden was not high,
four of the twelve samples exceeded 70 jag/m3 (105 ppbC). The range was
from 25 to 99 yg/m3 (38 to 148 ppbC). The site could be influenced at
times by the nearby road traffic and the campfires at the Elkmont
campgrounds. Quite noticeable in Table 9 are the very high acetylene
levels. Since other automotive hydrocarbons were not simultaneously
elevated, these high levels of acetylene are attributed to the campfires.
Acetylene has been identified as an emittant of wood fires (Sandberg, et
al., 1975 and Westberg, et al., 1979).
Isoprene and the terpenes appeared to be the dominant olefins during
the study period. Isoprene ranged from only a trace to a high of 11.6 yg/m3
(17.4 ppbC) with an average of about 4. Isoprene ranged from less than
1% to 13% of the TNMHC with an average of about 6%. The biogenic hydrocar-
bons (sum of isoprene and the terpenes) ranged from 6% to 28% of the TNMHC
excluding the one sample where the biogenic hydrocarbons accounted for 51%
of TNMHC. The four terpenes (a-pinene, B-pinene, myrcene, and A3-carene)
identified in these samples have been identified as the major emissions
from various pine species (Zimmerman, 1979). The study site could be com-
pared to a pine association (see Zimmerman, 1978).
The biogenic hydrocarbon levels measured over the course of eight days
were sometimes higher than the levels measured during the two days the EPA
samples were obtained (see Table 6). Thus, the biogenic contribution can be
38
-------
higher than the 6% measured .in the EPA samples. Furthermore, the observation
of low concentrations of gaseous precursors does not in itself prove that
there is then a low contribution to the aerosol burden. This is especially
true for biogenic hydrocarbons because of the very short lifetimes expected
for these compounds due to destruction by reaction with ozone and/or hydroxyl
radicals, and due to the diverse nature of the sources. Tables 15 and 16
(taken from Cronn and Campbell, 1978) present calculated lifetimes for var-
ious terpenes against hydroxyl and ozone attack. Ozone appears to be the
more important oxidizing species by a small factor. The issue of the rela-
tive importance of natural emissions to the aerosol burden cannot be resolved
alone by data such as presented here. But the relative importance of the
biogenic contribution to global inventories of aerosols should not be ignored
at this time.
TABLE 15. LIFETIMES FOR OH ATTACK
a-Pinene
B-Pinene
d-Limonene
A3-Carene
cis-Ocimene
nOH = 5x1 O5
9.5 Hours
8.3
3.7
6.4
1.7
nOH = 2xl06cm"3
2.4 Hours
2.1
0.9
1.6
0.4
39
-------
TABLE 16. LIFETIMES FOR OZONE ATTACK
mr (03) = 50 ppb
a-Pinene 1.5 Hours
B-Pinene 6.2
d-Limonene 0.35
A3-Carene 1.9
cis-Ocimene 0.11
40
-------
REFERENCES
Arnts, R. R., and S. A. Meeks. Biogenic Hydrocarbon Contribution to the
Ambient Air of Selected Areas. U.S. Environmental Protection Agency, EPA/600-
3-80-023, January, 1980.
Cicerone, R. J., J. D. Shetter, D. H. Stedman, T. J. Kelly, and S. C. Liu.
Atmospheric N20: Measurements to Determine Its Sources, Sinks, and Variations.
J. Geophys. Res., 83:3042-3050, 1978.
Cronn, D. R., and M. J. Campbell. Review of Current Knowledge of Atmospheric
Emissions from the Wood Product Industries. Paper presented at the 33rd
Northwest Regional American Chemical Society Meeting, Seattle, WA, June 14-
16, 1978.
Cronn, D. R., and M. J. Campbell. Measurements of Halocarbons at Various
Altitudes and Measurements of Hydroxyl Radicals in Both Hemispheres. Pro-
ceedings of the Conference on Methyl Chloroform and Other Halocarbon Pollu-
tants, Washington, D.C., February 27-28, 1979.
Cronn, D. R., R. A. Rasmussen, and E. Robinson. Measurement of Tropospheric
Halocarbons by Gas Chromatography-Mass Spectrometry. Report on EPA-R0804033-
01, submitted by Washington State University, Pullman, Washington, 69 pp,
1976.
Cronn, D. R., R. A. Rasmussen, and E. Robinson. Measurement of Tropospheric
Halocarbons by Gas Chromatography-Mass Spectrometry. Report on EPA-R0804033-
02, submitted by Washington State University, Pullman, Washington, 79 pp,
1977.
Cronn, D. R., R. A. Rasmussen, E. Robinson, and D. E. Harsch. Halogenated
Compound Identification and Measurement in the Troposphere and Lower Strato-
sphere. J. Geophys. Res., 82:5935-5944, 1977.
Cronn, D. R., and E. Robinson. Determination of Trace Gases in Learjet and
U-2 Whole Air Samples Collected During the Intertropical Convergence Zone
Experiment. In: 1977 Intertropical Convergence Zone Experiment, I. G.
Popoff, W. A. Page, and A. P. Margozzi, eds. NASA Technical Memorandum
78577, Moffett Field, California, pp. 61-106, 1979.
Goldan, P. D., Y. A. Bush, F. C. Fehsenfeld, D. L. Albritton, P. J. Crutzen,
A. L. Schmeltekopf, and E. E. Ferguson. Tropospheric N£0 Mixing Ratio Mea-
surements. J. Geophys. Res., 83:935-939, 1978.
Harsch, D. E. Evaluation of a Versatile Gas Sampling Container Design. Atm.
Environ., in press, 1979.
Husar, R. B. Meteorological Maps for the Time Period of the Smoky Mountain
Study. Report to Environmental Protection Agency, August 9, 11 pp., 1979.
41
-------
Parkhurst, W. J. Airmass Trajectory Data in the Vicinity of the Great Smoky
Mountains National Park, September 19-27, 1978. Tennessee Valley Authority
Report I-AQ-79-9, Muscle Shoals, Alabama, 20 pp, 1979.
Rasmussen, R. A. Inter!aboratory Comparison of Fluorocarbon Measurements.
Atmos. Environ., 12:2505-2508, 1978.
Rasmussen, R. A., and D. Pierotti. Interlaboratory Calibration of
Atmospheric Nitrous Oxide Measurement. Geophys. Res. Lett., 5:353-355,
1978.
Rasmussen, R. A., J. Krasnec, and D. Pierotti. N20 Analysis in the
Atmosphere Via Electron Capture-Gas Chromatography. Geophys. Res.
Lett., 3:615-618, 1976.
Tingey, D. T., H. C. Ratch, M. Manning, L. C. Grothaus, W. F. Burns,
and E. W. Peterson. Monoterpene Emission Rates from Live Oak. EPACERL-
045, Washington, D.C., 1978.
Weiss, R. F., and H. Craig. Production of Atmospheric Nitrous Oxide
by Combustion. Geophys. Res. Lett., 3:751-753, 1976.
Went, F. W. Blue Hazes in the Atmosphere. Nature, 187:641-643, 1960.
Westberg, H., K. Sexton, and D. Flyckt. Ozone Production and Transport in
Slash Burn Plumes. Paper Number 79-6.5 presented at the National Air Pollu-
tion Control Association Meeting, Cincinatti, OH, June, 1979.
Zimmerman, P. R. Testing of Hydrocarbon Emissions from Vegetation Leaf
Litter, and Aquatic Surfaces and Development of a Methodology for Compiling
Biogenic Emission Inventories. Washington State University report to
Environmental Protection Agency (Number 79-13-5), Research Triangle Park, NC,
February, 1979.
Zimmerman, P. R. Determination of Emission Rates of Hydrocarbons from
Indigenous Species of Vegetation in the Tampa/St. Petersburg Florida Area.
Washington State University report to Environmental Protection Agency
(Contract 68-01-4432), Pullman, WA, July, 1978.
42
-------
TABLE A-l.
APPENDIX A
HALOCARBON CONTINUOUS MONITORING DATA
ENTIRE HALOCARBON AND NITROUS OXIDE DATA SET FOR THE SMOKY
MOUNTAIN PROJECT
Date
9/21/78
9/22/78
Time
(EOT)
1220
1300
1340
1420
1500
1540
1620
1700
1740
1820
1900
1940
2020
2100
2140
2220
2300
2340
0020
0100
0140
0220
0300
0340
0420 '
0500
0540
0620
0700
F-12
(ppt)
w
__.
_
304
303
315
315
320
315
311
379
344
389
414
412
350
326
342
296
300*
306*
349*
284*
284*
280*
283*
281*
289*
F-ll
(ppt)
165
190
178
183
184
186
189
180
196
205
201
203
195
197
192
181
179
171*
173*
177*
184*
170*
167*
168*
172*
175*
165*
CH3CC13
(ppt)
127
147
_._
162
171
180
182
178
186
181
183
207
208*
203*
179*
170*
157*
105*
118*
113*
99*
105*
___
116*
145*
CC14 N20
(ppt) (ppb)
139
158
145
.
147
144
147
153
151
151
148
145
150
149
145
148
150
146*
141*
150*
142*
140*
136*
142*
143*
142*
145*
(continued)
43
-------
TABLE'A-1 (continued)
Date
9/22/78
9/23/78
Time
(EOT)
0740
0820
0900
0940
1020
1100
1140
1220
1300
1340
1420
1500
1540
1620
1700
1740
1820
1900
1940
2020
2100
2140
2220
2300
2340
0020
0100
0140
0220
0300
0340
0420
0500
0540
0620
0700
0740
0710
0840
0920
1000
1040
1120
F-12
(ppt)
289*
405*
419*
287*
305*
301*
305
321
313
305
305
301
314
311
305
307
311
311
302
321
315
320
324
324
323
338
337
339
336
340
333
327
327
330
321
340
334
329
329
326
327
F-ll
(ppt)
165*
170*
188*
180*
198*
183*
191
196
203
184
185
187
182
179
182
179
181
186
181
196
192
192
194
192
199
207
209
212
213
220
209
206
203
219
202
200
_
206
208
201
205
201
CH3CC13
(ppt)
218*
216*
124*
151*
153*
162*
161
172
240
214
220
204
199
193
169
170
172
174
183
237
228
228
237
270
294
317
322
308
292
287
274
265
251
250
247
240
_
247
248
255
257
259
CC14 N20
(ppt) (ppb)
141*
137*
149*
148*
148*
162*
148*
151
147
146
141
145
144
142
145
143
142
144
144
147
150
153
153
152
152
150
148
151
151
148
148
148
153
148
151
148
___
151
149
148
148
147
(continued)
44
-------
TABLE A-l (continued)
Date
9/23/78
9/24/78
Time
(EOT)
1200
1240
1320
1400
1440
1520
1600
1640
1720
1800
1840
1920
2000
2040
2120
2200
2240
2320
0000
0040
0120
0200
0240
0320
0400
0440
0520
0600
0640
0700
0740
0820
0900
0940
1020
1100
1140
1220
1300
1340
1420
1500
1540
F-12
(ppt)
324
311
323
___
342
327
314
301
302
295
315
365
318
333
359
342
327
326
312
340
___
___
.
.
.
_
___
428
383
377
486
387
298
305
357
319
324
314
308
311
312
F-ll
(ppt)
201
191
200
___
210
193
194
187
181
188
181
180
181
180
240
191
191
227
181
235
___
.
___
309
176
204
234
256
175
179
183
183
189
183
182
181
206
CH3CC13
(ppt)
254
224
210
___
218
208
201
171
156
160
161
174
155
147
155
177
168
157
150
140
___
___
_
___
172
156
153
159
154
157
168
182
183
186
191
186
185
182
CC14
(ppt)
146
147
143
___
148
146
146
143
143
141
142
142
141
140
139
143
138
139
140
141
___
_
___
___
139
139
138
140
140
141
140
144
144
144
144
144
142
143
N20
(ppb)
325
326
332
328
331
321
330
329
326
321
331
328
327
333
330
331
--_
_
_
___
--_
335
331
330
331
332
327
328
331
330
_
___
(continued)
45
-------
TABLE A-l (continued)
Date
9/24/79
9/25/78
Time
(EOT)
1620
1700
1740
1820
1900
1940
2020
2100
2140
2220
2300
2340
0020
0100
0140
0220
0300
0340
0420
0500
0540
0620
0700
0740
0820
0900
0940
1020
1100
1140
1220
1300
1340
1420
1500
1540
1620
1700
1740
1820
1900
1940
2020
F-12
(Ppt)
306
308
312
307
312
311
321
318
318
356
336
333
318
326
317
325
313
311
310
319
309
300
311
326
307
297
308
295
___
310
300
303
305
302
300
301
301
307
299
320
316
312
F-ll
(ppt)
181
188
187
186
200
229
188
200
234
200
186
187
192
182
181
209
191
182
187
183
187
183
184
247
222
185
198
186
_-_
187
183
186
184
182
182
182
182
181
204
183
182
299
CH3CC13
(ppt)
188
188
196
199
206
214
225
224
222
223
190
192
178
184
182
191
181
180
178
179
172
171
170
187
177
177
177
182
___
197
188
196
201
190
192
187
180
182
182
198
214
224
CC14
(ppt)
143
143
149
148
150
165
173
180
181
161
156
156
150
151
150
152
152
151
153
153
151
150
148
146
146
146
146
147
___
154
150
207
155
151
162
158
153
153
152
152
153
154
N20
(ppb)
« _
324
322
319
326
325
328
324
325
324
323
328
329
325
336
324
329
327
326
328
325
322
___
-
___
___
318
--_
___
328
324
327
327
325
329
330
327
327
327
324
324
324
(continued)
46
-------
TABLE A-l (continued)
Time
Date (EOT)
9/25/79 2100
2140
2220
2300
2340
9/26/78 0020
0100
0140
0220
0300
0340
0420
0500
0540
0620
0700
0740
0820
0900
0940
1020
1100
1140
1220
1300
1340
1420
F-12
(ppt)
304
304
304
304
305
304
307
306
304
306
303
306
303
321
301
323
382
532
347
310
305
299
350
319
321
322
F-ll
(ppt)
273
248
185
185
186
203
184
192
185
228
186
186
202
199
327
217
231
182
293
239
213
202
193
238
205
206
205
CH3CC13
(ppt)
228
221
220
211
206
180
171
170
170
168
171
166
172
179
178
177
188
193
195
199
203
194
195
188
194
190
205
CC14
(ppt)
157
154
157
155
152
153
151
150
150
149
153
152
158
165
156
148
148
150
152
159
171
171
164
155
157
157
160
N20
(ppb)
329
331
330
328
334
329
333
334
331
336
330
329
332
330
329
330
327
332
332
327
328
325
327
326
336
318
341
*Mixing ratio calculated from time-regressed standard values at
beginning and end of time period only.
47
-------
N2Q
oo
8
CONTINUOUS GROUND MONITORING
SMOKY MOUNTRINS NRTIONflL PflRK
£8
CD
X
o.
8
.
o
9/21/78
9/22/78
9/23/78I 9/24/78
9/25/78 I 9/26/78
Figure B-l. Nitrous oxide mixing ratios as a function of time for the Smoky Mountain site.
-------
N20
CONTINUOUS GROUND MONITORING
8 EflSTERN WnSHINGTON STflTE
8T
£8
o
i i
s
x:
i
z:
8
CD
9/21/78 I 9/22/78 I 9/23/78 I 9/24/78 I 9/25/78 I 9/26/78
Figure B-2. Nitrous oxide mixing ratio as a function of time for the eastern Washington state
site.
-------
F-12
en
o
8
8
s
X
'8
c>
*
x*
x
x
**
9/21/78 9/22/78
CONTINUOUS GROUND MONITORING
SMOKY MOUNTRINS NRTIONRL PRRK
9/23/78 9/24/78
XX
9/25/78 9/26/78
Figure D-3. Fluorocarbon 12 mixing ratio as a function of time for the Smoky Mountain site.
-------
F-12
CONTINUOUS GROUND MONITORING
8 EASTERN WRSHINGTON STflTE
ST
s
'8
d
° 9/21/78 | 9/22/78 | 9/23/78 | 9/24/78 | 9/25/78 | 9/26/78
Figure B-4. Fluorocarbon 12 mixing ratio as a function of tine for the eastern Washington
state site.
-------
en
ro
8
m
O
in
(M
o
o
o-d
CD
2:
tI
X
§..
04-
m
s
CD
F-ll
CONTINUOUS GROUND MONITORING
SMOKY MOUNTfllNS NRTIONRL PflRK
V
X X
9/21/78
x x
*C
X
X
XXX
X
x x *< x*
x x
Uf
X
9/22/78 I 9/23/78
9/24/78 | 9/25/79 |9/26/78
Figure B-5. Fluorocarbon 11 mixing ratio as a function of time for the Smoky Mountain site.
-------
F-ll
en
eo
8
-8
'8
CONTINUOUS GROUND MONITORING
EflSTERN WflSHINGTON STflTE
* >$< X *v X * ^X <*X~ X >
ft
9/21/78 | 9/22/78 | 9/23/78 9/24/78
^**x v* V-^
Ay ^ x/^ N*X s
^^ ^ ^r* ,
9/25/78 9/26/78
Figure B-6. Fluorocarbon 11 mixing ratio as a function of time for the eastern Washington
state site.
-------
CCL4
en
8
S
ED
X
'8
8
0
X
X
9/21/78
CONTINUOUS GROUND MONITORING
SMOKY MOUNTRINS NflTIONflL PRRK
*
X
X
9/22/78
9/23/78
9/24/78
9/25/78 9/26/78
Figure B-7. Carbon tetrachloride mixing ratio as a function of time for the Smoky Mountain site.
-------
CCL4
en
en
8
H-8
Q_S
£8
'8
si
8
,
o
CONTINUOUS GROUND MONITORING
EflSTERN NflSHINGTON STflTE
**** *****
9/21/78
9/22/78
9/23/78
9/24/78
9/25/78
9/26/78
Figure B-3. Carbon tetrachloride mixing ratio as a function of time for the eastern Washington
state site.
-------
en
cr>
8
8-
1-8
O-d
58
*
3:8
8
8
8
8
d
X
X
xx
X XX X
X
X
x<
X X
X
CH3CCL3
CONTINUOUS GROUND MONITORING
SMOKY MOUNTfllNS NRTIONflL PflRK
X * X
*
X
X
*x
X
X
x x
9/21/78
9/22/78 I 9/23/78
9/24/78 I 9/25/78 19/26/78
Figure B-9. Methyl chloroform mixing ratio as a function of time for the Smoky Mountain site.
-------
CH3CCL3
en
8
8
H-8
0-8-
o
ii
£8
8
o
o
o
CONTINUOUS GROUND MONITORING
EflSTERN WflSHINGTON STflTE
9/21/78 I 9/22/78
9/23/78
9/24/78
9/25/78
9/26/78
Figure B-10. Methyl chloroform mixing ratio as a function of time for the eastern Washington
state site.
------- |