United States Environmental Monitoring TS AMD 8192
Environmental Protection Systems Laboratory February 1983
Agency PO Box 15027
^ Las Vegas NV 89114
Research and Development
wr
Lake Tahoe
Visibility Study
prepared for
Region IX
V m 1 \\\\w*i
-------
27415
TS AMD 8192
February 1983
LAKE TAHOE VISIBILITY STUDY
by
Marc Pitchford and Daniel Allison
Environmental Monitoring Systems Laboratory
Advanced Monitoring Systems Division
Las Vegas, Nevada 89114
Prepared for Region IX
U.S. ENVIRONMENTAL PROTECTION AGENCY
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
P. 0. Box 15027
Las Vegas, NV 89114
-------
NOTICE
This document has not been peer and administratively reviewed within
and is for internal Agency use and distribution only.
i i i
-------
ABSTRACT
Visibility monitoring and airborne particulate sampling in the Lake Tahoe
Basin were used to document present visual air quality levels and to assess
the relative impacts of major contributing emission source categories. Visi-
bility data were obtained by long path contrast and particle scattering tech-
niques. Particles were sampled in two size ranges at three locations and were
analyzed for mass concentration and elemental composition (elements greater in
atomic weight than Na). Statistical analysis showed fine particle concentr-
ation (particle diameter less than 2.5 ym) to be related to visibility. Mea-
sured elements plus the mass of material assumed to be associated with them
accounted for only 20 percent of the fine particle concentration. The remain-
ing 80 percent was assumed to be nitrates and carbonaceous materials; the
latter associated with wood smoke, terpenes and transport from upwind areas.
A model was developed to apportion all of the fine particle mass to source
categories. The results of this effort were then used to determine an optical
extinction budget.
[This study indicates 70 pe^n?ent of the basin-wide visibility impact and
30 percent of the South Lake Tahoe visibility impact are caused by natural
and long range transported emissions. Residential wood smoke emissions are
responsible for the majority of the remaining impact; at South lake Tahoe auto-
motive emissions are also significant7]
iv
-------
CONTENTS
Abstract iv
Figures vi
Tables vi i
Introduction 1
Monitoring Program 1
Results 5
Teleradiometer measurements 5
Nephelometer measurements 7
Photography 11 '
Particle sampling 11
Interpretive Analysis 19
Preliminary analysis 19
Source characterization 20
Visibility/fine particle relationship 28
Summary and Conclusions 31
References 34
-------
FIGURES
1. Lake Tahoe Basin with visibility study monitoring locations
indicated. Site numbers refer to Table I
Daily averaged visual range (km) as determined by the tele-
radiometer measurements versus time for Stateline Fire Lookout
and King's Beach sites ,
3. Cumulative frequency distributions of extinction coefficient
(km) as determined by teleradiometer measurements for Stateline
Fire Lookout and King's Beach sites for the summer/fall period.
Equivalent visual range scale from equation number 2 of text . . 8
4. Cumulative frequency distribution of extinction coefficient
(km ) as determined by the integrating nephelometer (assumes
extinction coefficient equals scattering coefficient) at South
Lake Tahoe for two six month periods. Equivalent visual range
scale from equation number 2 of text 9
5. Seasonal averaged dirunal variations in extinction coefficient
as determined by the integrating nephelometer (assumes extinc-
tion coefficient equals scattering coefficient) at South Lake
Tahoe. Equivalent visual range scale from equation number 2
of text 10
6. View from Stateline Fire Lookout towards the south end of the
lake, 9:00 a.m., August 4, 1981. Approximately 98 percent of
the time teleradiometer determined visual range is less than
the 320 km corresponding to this photograph 12
7. View from Stateline Fire Lookout towards the south end of the
lake, 9:00 a.m., June 4, 1981. Approximately 50 percent of the
time teleradiometer determined visual range is less than the 160
km corrsponding to this photograph 13
8. View from Stateline Fire Lookout towards the south end of the
lake, 9:00 a.m., August 13, 1981. Approximately 1 percent of
the time teleradiometer determined visual range is less than
the 60 km corresponding to this photograph 14
9. View from Stateline Fire Lookout towards the south end of the
lake, 9:00 a.m., June 21, 1981. Basin-wide visual range from
the teleradiometer is approximately 240 km while South Lake Tahoe
visual range from the nephelometer is approximately 150 km . . . 15
vi
-------
TABLES
Page
I Lake Tahoe Visibility and Particle Sampling Equipment, Sampling
Frequency, and Locations 4
II Sugarpine Point Average and Standard Deviation of Elemental
Composition When-Above Detection Limits for at Least 50 Percent
of Samples (ug/m ) 16
III South Lake Tahoe Average and Standard Deviation of Elemental
Composition When-Above Detection Limits for at Least 50 Percent
of Samples (yg/m ) 17
IV Sierra Ski Ranch Average and Standard Deviation of Elemental
Composition When-Above Detection Limits for at Least 50 Percent
of Samples (yg/m ) 18
V Results of Multiple Linear Regression on Equation Number 6 of
Text for Telephotometer Visibility/Sugarpine Point Particle Data
and Nephelometer Visibility/South Lake Tahoe Particle Data ... 21
VI Equations for Calculation of Fine Particulate Parameters (Note:
Chemical symbols are used for each measured element) 22
VII Sugarpine Point Fine Particle Source Characterization (Data
excludes Napa and Wrights Lake fire periods) 25
VIII South Lake Tahoe Fine Particle Source Characterization (Data
excludes Napa and Wrights Lake fire periods) 26
IX Sierra Ski Ranch Particle Source Characterization (Data excludes
Napa and Wrights Lake fire periods) 27
X Lake Tahoe Basin Particle Extinction Budget 30
XI South Lake Tahoe Particle Extinction Budget 32
vii
-------
LAKE TAHOE VISIBILITY STUDY
INTRODUCTION
In recognition of the potential threat to the unique environment of the
Lake Tahoe Basin by increased urban and commercial developments, Congress
enacted the Tahoe Regional Planning Compact (Public law 95 551). This legis-
lation directed the Tahoe Regional Planning Agency (TRPA) to "establish envi-
ronmental threshold carrying capacities and to adopt and enforce a regional
plan and implementing ordinances which will achieve and maintain such capaci-
ties while providing opportunities for orderly growth and development consis-
tent with such capacities." Atmospheric visibility, which relates to the air-
borne particulate carrying capacity of the atmosphere, is one of many envi-
ronmental qualities for which TRPA is responsible.
The U.S. Environmental Protection Agency designed and conducted a visi-
bility study in the Lake Tahoe Basin to support TRPA in the establishment of
a threshold for visibility. The objectives of the study were to 1) measure
present visibility levels, 2) determine sources of visibility impairment and
3) provide a means to estimate visibility impacts of proposed activities. A
major time constraint was imposed on the study. Results were required within
12 months of the funding of the program. The sparsity of pertinent historical
data dictated that a major portion of the study be directed towards monitoring.
This was followed by an intensive period of interpretive analysis. The fol-
lowing is a description of the visibility monitoring program, a summary of the
techniques employed to interpret the data and the results of the interpretive
analysis.
MONITORING PROGRAM
The monitoring program was designed to accommodate the geography and emis-
sion source configuration thought to be of importance to visibility for the
Tahoe area. The Lake Tahoe Basin is an area of approximately 1300 km located
in the Sierra Nevada and Carson Mountains (see Figure 1). The lake, at about
two km elevation above sea level, covers approximately 40 percent of the basin,
with the remainder being primarily coniferous forest. Mountains, up to one km
above the lake's surface, surround the basin. Tourism is the basin's primary
industry.
Estimates of air pollution emissions from local anthropogenic sources
were available in an emissions inventory prepared by EPA's regional office in
San Francisco. The major local anthropogenic sources are the products of fuel
combustion for vehicles and space heating, and suspended soil from transporta-
tion and construction activities. Urban population in the Basin is primarily
at the south and north ends of the lake, with the south having the larger
population. Major natural sources include suspended soils, forest fires and
terpenes (from the coniferous forest). Long range transport of pollutants
from upwind sources is also expected to impact the basin.
1
-------
Truckee
King's
Beach
Tahoe City
Lake
Tahoe
'Glenbrook
Sugar
•ine Pt.
South Lake'
_^Tahoe~l
Fallen Leaf/(4
Lake I / v-
Q Visibility Monitor |
/\ Particle Sampler 'v
^Teleradiometer Target \
— Teleradiometer Sight Patfi"*
Tahoe Basin
Figure 1. Lake Tahoe Basin with visibility study monitoring locations indi-
cated. Site nunbers refer to Table I
2
-------
The air pollutants of importance to visibility are airborne particulates
and N0?. The contribution of NCL to visibility impairment is important?oply
in plume optics, not in well mixed layers* or in urban-influenced areas .
It was assumed therefore that N0? was not an important factor in Lake Tahoe
visibility degradation.
The field program consisted of visibility and particle monitoring. Table
I indicates the instruments and their deployment. A number of different types
of instruments were employed to characterize various aspects of visibility at
Lake Tahoe. Teleradiometers provide a measure of apparent contrast, Cr, of a
distant target against its background (typically the sky). This along with
the distance to the target, r, and its inherent contrast, C , can be used to
estimate extinction coefficient, bgxt by
"ext = "I/r ,n W U
Visibility is often expressed in terms of visual range, the greatest distance
at which a dark object is visible. Visual range, VR, is related to extinction
coefficient by
VR = 3.912/bext. 2)
Integrating nephelometers make a point measurement of scattering coefficient,
scat' w^1'c'1 1S a Portion of the extinction coefficient.
bext " bscat + babs'
where b , is the absorption coefficient. If the absorption coefficient is
assumed lo be small compared to the scattering coefficient, then b . can
replace b t in equation 2 in order to calculate visual range. A moreficpmplete
discussion of these two measurement techniques is available elsewhere.
Spectacular views of the lake and surrounding mountains are a highly
valued resource at Lake Tahoe. Color slides were selected to provide an
overall qualitative means to document basin-wide visibility. The long path
measurements of the teleradiometer also monitored basin wide visibility.
Though restricted to contrast measurements of a limited number of elements in
a scene, a teleradiometer provides quantitative data required for interpretive
analysis. The two north Lake Tahoe locations for camera and teleradiometer
monitoring were at different elevations (lake level and 240 meters above lake
level), thus allowing a view from within and above low lying haze layers. The
light scattering measurements made by the nephelometer at South Lake Tahoe (at
the Tahoe Y) provided data representative of the higher pollutant concentra-
tions expected in the urban area.
Particle sampling was accomplished with stacked filter units similar to
those described by Cahill et al. They segregate sampled particles into
two size ranges (coarse particles 15 ym to 2.5 ym diameter and fine particles
less than 2.5 ym diameter), collecting each on separate filters. Gravimetric
and particle-induced x-ray emission (PIXE) analysis were performed on each
filter. The former gives the mass concentration while the latter provides
the elemental composition for elements greater in atomic weight than sodium.
Particle sampling sites were selected to provide a variety of necessary
information. Sugar Pine Point State Park was chosen as representative of
3
-------
TABLE I. LAKE TAHOE VISIBILITY MEASUREMENT AND PARTICLE SAMPLING EQUIPMENT, SAMPLING FREQUENCY, AND LOCATIONS
Activity
Instrument
Sampling Frequency
Location (Map # and Name)
Contrast Measure-
ments of Distant
Mountains
Color Si ides
Color Slides
Light Scattering
Coefficient Mea-
surement
Particle Sampling
in Two Size Ranges
MRI 3010
Multi wavelength
Teleradiometer
Olympus OM-2 with 50 mm
lens, manually operated
Two Olympus OM-2 auto-
matic timer operated;
south-looking with 50
mm lens, north-looking
with 135 mm lens.
MRI 1550 Integrating
Nephel ometer
Three times daily
(0900, 1200, & 1500)
Two times daily
(0900 4 1500)
Two times daily
(0900 & 1500)
Continuous, hourly
average data
Targets
Measurement
Locations
From
Toward
From
Toward
1 Deadman Point
2 Gunbarrel Ski Run
3 8455 foot peak
4 Tahoe Mountain
5 Ellis Peak
6 Kings Beach
7 Stateline Fire Lookout
Stacked Filter Unit
compatible with gravi-
metric, elemental analysis
48-hour samples every
two days
6 Kings Beach
7 Stateline Fire Lookout
8 South Lake Tahoe*
5 Ellis Peak - Northwest Shore
9 Sugarpine Point State Park
6 Kings Beach
8 South Lake Tahoe
8 South Lake Tahoe
8 South Lake Tahoe
9 Sugarpine Point State Park
10 Sierra Ski Ranch, Echo Summit
~Also photographed with 135 mm lens.
-------
basin averaged concentrations. Data from this site were expected to relate
well to the teleradiometer measurements. The South Lake Tahoe particle sam-
pler colocated with the nephelometer was necessary both to relate to the
optical data provided by the nephelometer and to provide information on the
composition and concentrations of particles generated in the urban area. A
third site at Sierra Ski Ranch, on Echo Summit, was expected to be relatively
uninfluenced by pollutant sources within Tahoe Basin. This site is ^400
meters above the lake and generally upwind of the Basin though exposed to the
same distant upwind sources as the Basin. Comparison of data from this site
with data from the two basin sites was intended to yield insight into the
nature of particles transported into the Basin. These three locations are
among nine that have been used in a previous particle monitoring program .
Results from the earlier research influenced the choice of sites for the visi-
bility study.
In addition to the information collected directly for this program, daily
maximum and minimum temperature data at Sugarpine Point and daytime hourly
surface observations (temperature, visibility, dewpoint, wind speed and direc-
tion, altimeter setting and sky condition) at South Lake Tahoe airport were
obtained. Forest Service records of wildfire and controlled burns in the
Basin during the study were also obtained.
The field monitoring program was divided into two six month periods. The
first (summer/fal1), running from June to the end of November 1981, was the
intensive monitoring period during which all of the activities indicated in
Table I were performed. This six month period of sampling was designed to
form the basis for interpretive analysis, the results of which were required
by the spring of 1982. A second six-month period (winter/spring) of less in-
tensive monitoring immediately followed the first. The objective of this mon-
itoring was to assess the degree to which the first period represented a full
year. Equipment operated during the second period consisted of the automatic
camera system and particle sampler at Sugarpine Point plus the nephelometer at
South Lake Tahoe.
Teleradiometer measurements and data analysis, photography, and particle
sampler operation were conducted under a cooperative agreement with the John
Muir Institute (JMI). Particle sample gravimetric and elemental analysis was
conducted under a cooperative agreement with Crocker Nuclear Laboratory,
University of California-Davis (UCD). The nephelometer operation and data
processing was conducted by the California Air Resources Board.
Each of the monitoring groups prepared and followed standard operating
and quality control procedures. Additionally JMI and UCD participated in an
external quality assurance program involving system and laboratory analytical
audits performed by Rockwell International and field audits performed by
Lockheed Engineering and Management Services Company; both under EPA contract.
RESULTS
Teleradiometer Measurements
Figure 2 shows the time plots of the daily-average visual range as cal-
culated from teleradiometer contrast measurements at Stateline Fire Lookout
and Kings Beach. Daily mean values from the two observation sites are highly
5
-------
40O-I
Stateline Fire Lookout
300-
200-
King's Beach
300-
200-
100-
15
June
30
15
July
30
15
August
30
15
September
30
16
October
30
15
November
30
Figure 2. Daily averaged visual range (km) as determined by the teleradiometer measurements versus time
for Stateline Fire Lookout and King's Beach sites
-------
correlated. The Stateline Fire Lookout site, which is higher in elevation
than the Kings Beach site, has a systematically higher visual range. The low
visual range values from June 19 to June 29 and from August 6 to August 14
were associated with impacts from the Napa and Wrights Lake forest fires. Due
to the sporadic and largely uncontrolled nature of forest fires as pollution
sources, data during these two periods were not included in subsequent inter-
pretive analysis. Inspection of the target-specific teleradiometer data in-
dicates that the data from targets number 1 and number 5 showed greater dis-
agreement with the target-averaged values than data from the other targets.
These differences were attributed to the sensitivity of teleradiometer measure-
ments to inherent contrast errors for short path length measurements (targets
1 and 5 are 15 and 24 km from the observation points). They may also have
been due in part to a spatially inhomogeneous distribution of visibility-
reducing pollutants since sight paths for targets 1 and 5 cross only the north-
ern half of the Basin.
Extinction coefficient values derived from contrast measurements of tar-
get numbers 2, 3, and 4 were averaged to create the basinwide visibility data
base used in subsequent interpretive analysis. Figure 3 shows the cumulative
frequency distributions of these values for the Stateline Fire Lookout and
King's Beach sites. Data from the King's Beach observation site were selected
for analyses combining teleradiometer and particle data since its sight paths
are lower in elevation than corresponding sight paths for Stateline Fire
Lookout and thus likely to be more highly related to ground-level measured
particle concentrations.
Nephelometer Measurements
Cumulative frequency distributions of summer/fal1 and winter/spring visi-
bility data from the nephelometer at South Lake Tahoe are shown in Figure 4.
Data during the periods influenced by the two major forest fires were removed.
Scattering coefficient data, measured at 500 nm wavelength, were adjusted to
550 nm to correspond more closely to human eye and teleradiometer peak response
characteristics . The summer/fall period had somewhat poorer average visibil-
ity than the winter/spring period. The seasonally averaged diurnal behavior
of scattering coefficient at South Lake Tahoe is shown in Figure 5. Winter
and spring have large diurnal fluctuations while fall has a less dramatic vari-
ation. Summer has no significant diurnal variation in scattering coefficient.
Trapping of the pollutants below a nocturnal radiation inversion is apparently
responsible for the increasing scattering coefficient in the late afternoon
and evening. However, the decrease of the values occurs too early in the
morning to correspond the breakup of the inversion. This suggests some system-
atic local circulation which disperses the pollutants. Summer scattering coef-
ficient values seem uninfluenced by either mechanism.
The nephelometer-measured visibility at South Lake Tahoe is significantly
lower than the teleradiometer-measured Tahoe Basin visibility as seen in the
cumulative frequency distributions (Figures 3 and 4). To determine whether
this was due in any great part to instrument bias, an examination was conduc-
ted of data from several periods of time during which some degree of spatial
uniformity in the concentration of visibility-reducing pollutants might be
expected. June 13 to June 22 was a period of very good visibility. It is
expected that the meteorological conditions during such an episode of high
visibility would tend to discourage spatially inhomogeneous pollutant dis-
tributions. The mean nephelometer scattering coefficient and teleradiometer
extinction coefficients were the same for this clean period (0.08 km ). By
7
-------
0.50-,
0.40
0.30-
- 0.20-
e
S
u
C
S
o
c
o
o
c
V
X
0.10-
0.09-
0.08-
0.07-
0 06-
£ 0.05-
0.04-
0.03-
0.02-
0.01-
-Stateline Fire Lookout
-King's Beach
i-10
-25
—50
-75
E
Jt
«
01
c
s
-------
0.50^
0.40-
0.30-
- 0 20-
E
e
S
o
£
c
o
u
c
o
K
UJ
0.10-
0.09-
0.08-
0.07-
0.06-
0.05-
0.04-
0.03-
0.02-
0.01-
Summer/Fall
— ——Winter/Spring
r10
-25
-50
-75
-100
-150
-200
-300
9
S>
c
a
c
"5
3
"i 1 1 i—i—i—i i i—i 1 1 r
2 5 10 20 30 40 60 60 70 80 90 95 98 99
Percent
Figure 4. Cumulative frequency distributions of extinction coefficient (km )
as determined by the integrating nephelometer (assumes extinction
coefficient equals scattering coefficient) at South Lake Tahoe for
two six month periods. Equivalent visual range scale from equation
number 2 of text
9
-------
0 14
0 13-
0 12-
0 11-
_ 0 10-
E
i 0.09-
s 0 08-
*
•
c 007H
o
s
E 0 06H
0 05-
0 04-
0 03-
0 02-
0 01-
Fall
Winter
Spring
Summer
•—30
-40
-BO ®
c
(D
E
O
3
•)
>
1-75
-100
-150
-200
-300
"I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1-
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Hour
Figure 5. Seasonal averaged diurnal variations in extinction coefficient as determined by the integrating
nephelometer (assumes extinction coefficient equals scattering coefficient) at South Lake Tahoe.
Equivalent visual range scale from equation number 2 of text
-------
assuming that the impact of the major forest fires was uniform over the Basin
and that it was intense enough to overwhelm the impacts of local sources, a
comparison on poor air quality days was possible. The average scattering coef-
ficient value from the nephelometer for August 12 to August 14 was 0.078 km
while the extinction coefficient from the teleradiometer was 0.053 km" . These
values are close by comparison to differences in the cumulative frequency dis-
tributions for the two instruments on the poor visibility end of the scale
(compare Figures 3 and 4). By comparison, October 21 was selected by the
appearance in photographs of a visible layer of pollutants over the southern
end of the Basin suggesting a spatially inhomogeneous pollutant distribution.
On that date the average scattering coefficient at South Lake Tahoe was
0.423 km and the basin-averaged extinction coefficient was 0.039. Thus, in
at least a semi-quantitative sense, the two methods seem to agree (and dis-
agree) as expected. Diurnal visibility variations may be another factor con-
tributing to the different mean values since the nephelometer provides a 24
hour-a-day measurement while the teleradiometer has discrete 9 a.m., noon, and
3 p.m. measurements. As seen in Figure 5, except in summer, the South Lake
Tahoe visibility during these hours is considerably better than the daily
average. A subset of the nephelometer data was compared to the teleradiometer
data to investigate this factor. Using only paired data (hourly averaged
nephelometer data for hours corresponding to each valid teleradiometer value),
the ratio of the mean nephelometer derived extinction coefficient to the mean
teleradiometer derived extinction coefficient is 2.0 overall, 1.7 for summer,
and 3.0 for fal1.
Photography
The photographic documentation of visibility at Lake Tahoe resulted in
nearly 3000 color slides. Though they have not been used in a quantitative
sense, they have been a valuable aid in the interpretation of other measure-
ments. They have also been used to communicate visibility information to
those unfamiliar with more quantitative measures of visibility. The importance
of photography in a study such as this cannot be overemphasized. Policy
makers have the opportunity to "see" a variety of visibility conditions, and
know, by associating them with quantitative measurements, the frequency with
which they occur. Figures 6, 7, and 8 are photographs representative of un-
usually good, average and unusually poor basin-wide visibility. An example of
good basin-wide visibility coupled with a perceptable haze layer over the
south end of the lake is shown in Figure 9.
Particle Sampling
Means and standard deviations of particle data are presented in Tables II,
III, and IV for Sugarpine Point, South Lake Tahoe, and Sierra Ski Ranch sites
respectively. At Sugarpine Point, the values for the two six-month monitoring
programs can be compared. The coarse and fine particle mass concentrations
are greater during the intensive (summer/fall) period than the following six-
month period. This is consistent with the lower visibility measured by the
nephelometer in summer/fall compared to winter/spring. The largest measured
elemental contributors to the coarse mass concentration are soil-related ele-
ments (i.e., Al, Si, K, Ca, and Fe). For the fine size range, sulfur and soil-
related elements are important contributors. In both size ranges, the major
portion of the mass must be contributed by light elements (lower in atomic
weight than sodium) which are not measured by PIXE analysis. The light elements
constitute a greater portion of the fine size particles than the coarse. Fine
11
-------
Figure 6. View from Stateline Fire Lookout towards the south end of the lake, 9:00 a.m., August 4, 1981.
Approximately 98 percent of the time teleradiometer determined visual range is less than the
320 km corresponding to this photograph
-------
Figure 7. View from Stateline Fire Lookout towards the south end of the lake, 9:00 a.m., June 4, 1981.
Approximately 50 percent of the time teleradiometer determined visual range is less than the
160 km corresponding to this photograph
-------
Figure 8. View from State! ine Fire Lookout towards the south end of the lake, 9:00 a.m., August 13, 1981.
Approximately 1 percent of the time teleradiometer determined visual range is less than the 60 km
corresponding to this photograph
-------
Figure 9. View from Stateline Fire Lookout towards the south end of the lake, 9:00 a.m., June 21, 1981.
Basin-wide visual range from the teleradiometer is approximately 240 km while South Lake Tahoe
visual range from the nephelometer is approximately 150 km
-------
TABLE II. SUGARPINE POINT AVERAGE AND STANDARD DEVIATION OF ELEMENTAL COM-
POSITION WHEN ABOVE DETECTION LIMITS FOR AT LEAST 50% OF SAMPLES
(ug/m )
Summer/Fal1
Coarse Fine
Winter/Spring
Coarse Fine
A1
Si
S
CI
K
Ca
Ti
Mn
Fe
Zn
Br
Pb
Mass
0.76+0.53
1.26+0.74
0.15+0.10
0.20+0.12
0.32+0.26
10.9+5.8
0.06+0.05
0.11+0.09
0.22+0.14
0.06+0.03
0.03+0.01
0.004+0.002
0.04+0.03
0.002+0.001
0.004+0.003
0.02+0.01
7.1+2.6
0.38+0.47
0.78+0.89
0.08+0.04
0.07+0.07
0.13+0.10
0.22+0.25
0.03+0.01
0.26+0.26
0.06+0.07
0.14+0.15
0.18+0.16
0.08+0.13
0.04+0.04
0.01+0.01
0.01+0.01
0.04+0.05
8.0+7.0
0.03+0.02
5.2+3.7
16
-------
TABLE III. SOUTH LAKE TAHOE AVERAGE AND STANDARD DEVIATION OF ELEMENTAL
COMPOSITION WHEN ABOVE DETECTION LIMITS FOR AT LEAST 50% OF
SAMPLES (ug/m )
Summer/Fal 1
Coarse Fine
A1
1.22+0.58
0.05+0.04
Si
2.48+1.15
0.12+0.09
S
0.06+0.03
0.13+0.11
CI
0.13+0.33
-
K
0.26+0.13
0.05+0.03
Ca
0.38+0.19
0.03+0.01
Ti
0.06+0.04
0.004+0.002
Mn
-
0.001+0.001
Fe
0.57+0.29
0.04+0.03
Cu
-
0.001+0.004
Zn
0.01+0.01
0.003+0.002
Br
0.03+0.01
0.03+0.01
Pb
0.08+0.04
0.10+0.04
Mass
17.9+7.9
7.5+3.2
17
-------
TABLE IV. SIERRA SKI RANCH AVERAGE AND STANDARD DEVIATION OF ELEMENTAL COM-
POSITION WHEN ABOVE DETECTION LIMITS FOR AT LEAST 50% OF SAMPLES
(ug/m )
Summer/Fal 1
Coarse Fine
A1
1.09+0.88
0.09+0.08
Si
1.86+1.47
0.20+0.16
S
0.08+0.05
0.27+0.20
CI
0.23+0.54
-
K
0.27+0.28
0.07+0.04
Ca
0.56+0.85
0.06+0.10
Ti
-
0.006+0.004
Mn
-
0.003+0.004
Fe
0.49+0.47
0.07+0.06
Cu
-
0.001+0.003
Zn
-
0.002+0.002
Pb
-
0.004+0.003
Mass
20.1+14.5
7.6+3.6
18
-------
particle mass concentrations are similar at the three sites during the first
six months, though coarse mass concentrations are much lower at Sugarpine Point
than at the other sites.
INTERPRETIVE ANALYSIS
This section describes the methods and results of interpretive analysis
of data collected during the intensive monitoring period from June through
November of 1981. Data taken during the two periods of major forest fire
influence were not used. The overall objective of this analysis was to relate
measured visibility to particulate pollution sources. Basically, this was
accomplished by determining the relative contribution of sources to the par-
ticle concentration, then applying multiple linear regression techniques to
the source contribution and visibility data to relate them.
Multiple linear regressions performed with a standard computer routine
provide the basis for this work. In general , multiple linear regression
of a dependent variable Y on independent variables X,, ..., X determines
coefficients a., ..., a which minimize the mean square error. The linear
model considered here is
n
Y = a + z a.X. + e 4)
0 i=l 1 1
where the e denotes the error. The error is considered to have a mean of
zero. The variance a and the errors corresponding to the various Y's are
assumed to be independent. The level of significance of the coefficients
can be determined using a t test, while an F test is used to determine the
overall significance of the linear model. The proportion of total variation
in Y accounted for by the regression model is given by the coefficient of
determination, R .
r2 _ Sum of Squares Regression
Sum of Squares Total '
A more.detailed discussion of multiple linear regression is available else-
where.
Preliminary Analysis
First, an assessment of the relative importance of coarse and fine par-
ticles to visibility was conducted. This was accomplished using a model of
the form.
bp,e*t * AcMc + Vf 6>
where bp gxt is the particle extinction coefficient, Mc and are the coarse
and fine mass concentrations and A and Af are the regression coefficients
determined in the analysis. The bp ext vSlues are determined by averaging the
b . values corresponding to the particle sampling periods and subtracting the
appropriate Rayleigh scattering coefficient (scattering by gas molecules in an
unpolluted atmosphere). The intercept, a , was forced to be zero because par-
ticle extinction will be zero when particTe concentration is zero.
19
-------
Table V shows the results of this model for teleradiometer data paired
with particle data from Sugarpine Point samples and nephelometer data paired
with particle data from the South Lake Tahoe sampler. The R values indicate
that 84 and 88 percent of the variation in particle extinction (teleradiometer
and nephelometer respectively) is explained by the particle mass concentration.
In Jboth cases, the fine particle coefficient is significant while the coarse
particle coefficient is not.
A dependence of visibility.pn fine particle concentration has been demon-
strated by other investigators. However, testing of the coarse particle
influence on visibility is required before it can be dismissed. Since fine
particles alone demonstrate a significant relationship to visibility, the
coarse particle data will not be included in any subsequent analysis.
Source Characterization
In order to assess the causes of visibility impairment in the Tahoe Basin,
it is necessary to identify the sources of fine particles. One approach in-
volves using compositional characteristics.of sources to identify their con-
tribution to the ambient particle loading. This technique was used to spe-
cify "Soil," "Sulfate," and "Automotive" source categories. The equations
used to calculate the contribution by these sources are specified in Table VI.
Also shown in this table is a category known as "Other" which contains mat-
erials from a variety of minor (in some cases unknown) sources. For each cat-
egory, some measured elemental concentration (or concentrations) is adjusted
based on known or assumed compositional characteristics of sources. In the
case of "Soil" the measured concentrations of a half dozen elements are multi-
plied by numbers which adjust for the additional mass of oxygen assumed to be
associated with them in soil minerals. Since fine potassium (K) is thought to
be from both soil and vegetative burning, iron (Fe) concentration times 0.48
(the ratio of coarse K to Fe at Lake Tahoe) is used for the contribution to
soil by K.
The "Sulfate" value is similarly determined by adjusting the sulfur con-
centration to account for the added oxygen mass. Lead is used as a tracer
for the exhaust products of automobiles. In this case, the multiplicative
factor accounts primarily for hydrocarbons associated with auto exhaust.1RThe
factor of 12 for "Auto" is based on information from Trijonis and Davis.
The category labeled "Other" is also determined by adjusting measured elemental
concentrations to reflect the mass of the associated unmeasured elements.
The same type of approach was tested on another source category, "vege-
tative burning", but was not used. Other investigators have used non-soil -
related fine K as a tracer for smoke from vegetative burning. An average
K to Fe ratio calculated from the coarse particle data is assumed to represent
the relative proportions of those elements in the soil. This ratio multiplied
by fine Fe represents the fine K from soil and the difference between it and
the total fine K is assumed to be related to vegetative burning. Application
of this technique to samples taken during the impact of the two major forest
fires indicated that the excess fine K (total fine K minus soil related fine
K) to smoke concentration was not constant. The impact, in terns of enhanced
fine particle concentration, of the Napa fire was accompanied by a significant
increase in excess K while the even greater impact from the Wrights Lake fire
had only a minor increase in excess K above background levels. There are two
notable differences between the two fires. The Napa fire was 200 km from
Tahoe and was fueled by grass, shrubs and hardwood forest while the Wrights
20
-------
TABLE V. RESULTS OF MULTIPLE LINEAR REGRESSION ON EQUATION NUMBER 6 OF TEXT FOR TELEPHOTOMETER VISIBILITY/
SUGARPINE POINT PARTICLE DATA AND NEPHELOMETER VISIBILITY/SOUTH LAKE TAHOE PARTICLE DATA
r ......
Ac
t-test
significance
Af
t-test
signi ficance
R2
n
Teleradiometer
-0.00015
0.430
0.00236
0.000
0.84
57
Nephelometer
0.00048
0.443
0.00448
0.039
0.88
19
-------
TABLE VI. EQUATIONS FOR CALCULATION OF FINE PARTICULATE PARAMETERS (Note:
Chemical symbols are used for each measured element)
1. ^ Soi1
Soil = Mg x 1.66 + A1 x 1.67 + Si x 2.14 + Ca x 1.44 + Mn x 1.44 + Fe x 1.40
+ (Fe x 0.48) x 1.21
2. Sulfate
Sulfate = S x 3.0
3. Auto
Auto = Pb x 12
4. Other
Other = P x 2.29 + V x 1.78 + Cr x 1.46 + Ni x 1.27 + Cu x 1.25 + Zn x 1.25
+ CI x 1.65 + Se
22
-------
Lake fire was adjacent to the Tahoe Basin and primarily consumed coniferous
forest. A special study was performed, in which samples containing smoke from
conifer wood was compared to hardwood smoke and grass smoke samples. The
ratio of K to total mass concentration for the conifer wood smoke was much
smaller and more variable than the other two. Therefore, without information
concerning the types and mix of fuel involved, excess K could not be used in
a quantitative way as a tracer.
At this point in the analysis, all of the heavy elements (larger in atomic
number than 11, sodium) and light elements which are assumed to be associated
with them are accounted for as "Sulfate," "Auto," "Soil" or "Other." However,
only about 20% of the mass of the Lake Tahoe fine particle samples is explained
by these sources. The 80% unexplained is composed of light elements. It is
most likely that the major portion of this mass is carbonaceous in nature,
with a good share of that being organic hydrocarbons. Nitrates may also be a
significant contributor to the unexplained fraction. Some portion of the un-
explained mass may be water bound to the particles. This is not thought to
be a major component.
The sources contributing to this unexplained category can be classified
as either of local origin or the result of transport. Nitrates and hydrocar-
bons can be transported into the Tahoe Basin in much the same way as sulfates
from upwind sources in California. As with sulfates, nitrates are assumed to
result from transport from distant sources. Thus, the local sources primarily
provide hydrocarbons. Based on calculated emissions for the basin, the major
contributors to hydrocarbons are burning wood for residential heating and aero-
sols resulting from the terpene emissions by coniferous trees. On the basis
of their estimated emissions, these two local sources dominate all the other
hydrocarbon sources in the basin, including diesel, which has about one half
the expected emissions of "Auto". Another class of short term hydrocarbon
sources is wild fires and prescribed burning. Except for impacts from occa-
sional major wild fires, wood burning for residential heating represents the
overwhelming fraction of wood consumed in the Tahoe Basin.
A model was developed to apportion the unexplained fraction of the fine
mass concentration to three classes of sources: the transported light elements,
the terpenes, and the residential wood smoke. The transported light element
concentration is assumed to be proportional to the sulfate concentration,
since it is also assumed to result from transport. A proportionality constant
of two was derived from typical measured values for the ratio of nitrates +
total hydrocarbon to sulfates for the Sacramento Valley. The remainder of
the unexplained mass is assumed to be terpenes plus residential wood smoke.
The production of smoke from residential burning of wood is assumed to be pro-
portional to degree-heating hours. This is an extension of the empirical rel
tionship between fuel demand for residential heating and degree heating days.
The assumptions were that the proportion of wood to other fuels is independent
of temperature or season, that wood smoke emissions are proportional to wood
consumed and that degree-heating hours would more faithfully represent fuel
demand in an area with frequent large diurnal temperature fluctuations than
degree heating days. The hourly temperature values used to calculate degree-
heating hours were derived from a sinusoidal fit to the measured daily high
and low temperatures, with the period of the sine function controlled by the
time between sunr^e and sunset. Terpene emission rates increase exponentially
with temperature. Using the hourly temperature values, the relative emis-
sions of terpenes can be calculated for any period of time.
23
-------
The following equations relate the various components of the model.
U = TLE + NTU
TLE = ? *sn
NTU =
7)
8)
9)
where, U, TLE, and NTU are the unexplained, transported light elements and
non-transported unexplained concentration. SO. is sulfate concentration.
"D" and "T" are the degree heating hour and relative terpene missions respec-
tively. Wood smoke concentration is the product "aD" while terpene concentra-
tion is "bT." By making the assumption that "a" and "b" are constants, they
can be determined by multiple linear regression analysis on equation number 9.
In this way the gross variations of NTU with temperature can be accounted for
(NTU generally increases with increasing temperature at Sugarpine Point and
Sierra Ski Ranch and decreases at South Lake Tahoe). However, ambient concen-
trations are not directly proportional to emissions. The use of constant "a"
and "b," determined by regression analysis, results in over or under predicting
the NTU concentration for a typical day in spite of a good average prediction.
To avoid this situation, a further adjustment is made to the constant values
determined from multiple linear regression. For each sample period these
values are multiplied by the ratio of the observed to the predicted concentra-
tion of the non-transport unexplained mass, (aD+bT^' adJustment tenm
could be interpreted physically as a meteorological factor. The term is
greater than one when the meteorology promotes stagnation resulting in higher
than average concentrations of pollutants from local sources. When the term
is less than one the meteorology retards the concentration of pollutants from
local sources (i.e., increases their dispersion, dilution or deposition).
The model was applied to non-wildfire related particle data from each
of the three Tahoe particle sampling sites. The degree of success of these
applications of the regression portion of the model are demonstrated by the
high level of significance for the two regression coefficients (less than
0.05$ chance of the null hypothesis being true for all except "a" at Sierra
Ski Ranch with 14%) and the high multiple R value, 0.81 at Sugarpine, 0.73
at South Lake Tahoe and 0.68 at Sierra Ski Ranch. The physical significance
of the adjustment term is supported by photographs which show rain during many
of the periods where the term is considerably less than one. Apparently the
washout by the rain prevented the higher levels of terpene and wood smoke
predicted on the basis of temperature alone.
Tables VII, VIII and IX present the results of the source characterization
assessment techniques described above for the three particle monitoring sites.
The "Sulfate" and "Transported Light Element" source categories were combined
and labeled "Long Range Transport." One of the more striking results is the
difference between the "Residential Wood Smoke" estimate at the three loca-
tions. As one might expect, the highly developed South Lake Tahoe urban area
has higher smoke concentrations from residential heating than at the more re-
mote basin site at Sugarpine. The "Residential Wood Smoke" concentrations for
Sierra Ski Ranch located at Echo Summit are lower than either of the two basin
sites. The "Long Range Transport" is greatest at Sierra Ski Ranch on Echo
Summit perhaps as a result of its exposure to transport from the west. The
difference in transport values at the two basin sites might result from en-
hanced vertical mixing at the south end of the basin. This might also explain
why the mean "Terpene" concentrations which are about the same at Sugarpine
and Sierra Ski Ranch are about a factor of two higher than at South Lake Tahoe.
24
-------
TABLE VII. SUGARPINE POINT FINE PARTICLE SOURCE CHARACTERIZATION (Data
excludes Napa and Wrights Lake Fire Periods)
Source
Overal1
Summer
Fall
yg/m
%
ug/m
%
ug/m3
%
Soil
0.46
6.4
0.67
7.8
0.32
5.1
Auto
0.19
2.7
0.12
1.3
0.24
3.8
Long Range
Transport
2.0
28
2.0
23
2.0
32
Residential
Wood Smoke
1.2
17
0.73
8.4
1.6
25
Terpene
3.3
46
5.1
59
2.1
34
Other
0.02
0.3
0.02
0.2
0.02
0.3
Total
7.15
100
8.65
100
6.20
100
25
-------
TABLE VIII. SOUTH LAKE TAHOE FINE PARTICLE SOURCE CHARACTERIZATION (Data
excludes Napa and Wrights Lake Fire periods)
Overal1 Summer Fal 1
v 3 3 3
Source ug/m % yg/m % vg/m %
Soil
0.46
6.1
0.67
9.3
0.35
4.5
Auto
1.2
15
0.76
11
1.4
18
Long Range
Transport
1.2
16
1.5
21
1.0
13
Residential
Wood Smoke
3.0
40
1.6
22
3.7
49
Terpene
1.7
23
2.7
38
1.2
15
Other
0.03
0.4
0.02
0.2
0.04
0.5
Total
7.51
100
7.18
100
7.67
100
26
-------
TABLE IX. SIERRA SKI RANCH SOURCE CHARACTERIZATION (Data excludes Napa arid
wrights Lake Fire periods)
Sources
Overal1
pg/m3 %
Summer
pg/m3 %
Fall
vg/m3
%
Soi 1
0.80
11
1.2
13
0.49
7.5
Auto
0.17
2.2
0.13
1.5
0.21
3.2
Long Range Transport
2.4
32
2.5
28
2.4
36
Residential Wood Smoke
0.70
9.2
0.46
5.2
0.90
14
Terpene
3.5
46
4.6
52
2.5
39
Other
0.02
0.3
0.02
0.2
0.02
0.3
Total
7.61
100
8.89
100
6.51
100
27
-------
An alternate explanation for low terpene at South Lake Tahoe may be related to
the expected lower ozone values in the more urbanized South Lake Tahoe area
caused by higher NO concentrations. Ozone is highlv.reactive with terpene
causing the gaseous terpene to form fine particles. This would result in a
lower ratio of particle to gaseous terpenes at South Lake Tahoe than the other
locations in the area. The "Auto" concentrations are nearly the same at
Sugfcrpine and Sierra Ski Ranch and as expected are much smaller than at South
Lake Tahoe. Mean "Soil" concentrations are about the same for the two basin
sites, though nearly half the values for Sierra Ski Ranch. This may have
resulted from the somewhat less ground cover and ski area maintenance during
the summer at the latter location.
Visibility/Fine Particle Relationship
The optical extinction efficiency of particles from distinct source cat-
egories are not necessarily the same. The size distribution, shape, density
and index of refraction characteristics of particles emitted by a source will
determine its optical affect per unit mass concentration . The product of
the extinction efficiency for a collection of particles and its mass concen-
tration is the particle extinction coefficient associated with those particles.
The particle extinction coefficients associated with particles from each of
the sources which impacts an air parcel can be summed to obtain the total par-
ticle extinction for that parcel. Lack of detailed physical information on
particles associated with each source makes it impractical to calculate the
corresponding extinction efficiencies. However, it is possible to estimate
them using multiple linear regression.
The multiple linear regression model employed has particle extinction,
derived by visibility measurement, as the dependent parameter and the mass
concentration of particles associated with the various source categories
as independent parameters. The coefficients determined by the regression
analysis are the estimates of the extinction efficiencies for the particles
associated with various sources. A satisfactory model is obtained by this
method when 1) a large fraction of the variation in the dependent variable
is explained (i.e., a high R value), 2) the regression coefficients are
statistically significant and 3) it is physically meaningful. These condi-
tions are often not met upon the initial application of regression analysis.
If this is the case, a process of trial and error is employed in which in-
dependent parameters are eliminated or combined and re-evaluated in the mul-
tiple linear regression model. A simple example of this was the elimination
of coarse particles from the visibility analysis because the coefficient rela-
ting it to particle extinction was statistically indistinguishable from zero.
Data from the Kings Beach teleradiometer location for targets number 2
through 4 and particle data from the Sugarpine Point site were used to develop
a regression model for basin-wide visibility. The teleradiometer-derived
particle extinction values corresponding to each of the particle sampling
periods were averaged. The mean particle extinction was the dependent vari-
able; the independent variables were the particle emission source categories.
A series of multiple linear regressions were run on the data trying different
combinations of source categories for independent variables in an attempt
to find a satisfactory model. The most useful set of source classifications
tested as independent variables are "Long Range Transport," "Wood Smoke" and
sum of "Soil", "Auto", "Terpene" and "Other". The last of these is referred
28
-------
to as "Non-Transport-Non-Smoke." The following equation results from the
multiple linear regression analysis performed on these parameters.
b a . = 0.00451 LRT + 0.00374 RWS + 0.00074 NTNS
p ,ext
The tp ext, LRT, RWS and NTNS are the particle extinction, "Long Range
Transport," "Residential Wood Smoke" and "Non-Transport-Non-Smoke" parameters.
The R value for this model is 0.84. All the coefficients are significant at
the 90% level or better and the F-ratio indicates a highly significant
regression.
Table X shows the results of employing this model to determining the
particle extinction budget using the mean particle values shown in Table VII.
Comparing the percentage values in the two tables reveals several interesting
results. Though "Terpenes" represent the highest source contribution they
contribute less to the extinction than "Long Range Transport" and less than
"Residential Wood Smoke" except in summer. This is due to the enhanced cap-
acity of transport and smoke-related particles for impacting visibility as
indicated by their regressions coefficients which are five to six times greater
than the coefficients for the rest of the particles at Tahoe. The disparity
in optical properties may be the result of the size distribution of the particles
in question. The smoke and transport-related particles are probably in the
0.5 to 1.0 ym diameter size range which is highly effective for light scatter-
ing. The terpenes are most likely smaller (less than 0.5 ym diameter) and
therefore less efficient at scattering light. Fine soil-related particles are
too large for efficient scattering. While the auto-related particles may be
in the proper size range, apparently their low concentrations prevented them
from surviving the earlier multiple linear regression attempts as a separate
independent parameter. Another factor which may share the responsibility
for some particle source categories having higher efficiencies than others
is their spatial distributions, especially in the vertical. If the concen-
tration of particles associated with a source fell off rapidly with height,
then the 200 m average height difference between the teleradiometer sight
paths and the particle sampler would result in a lower "effective" optical
extinction efficiency than the true extinction efficiency. Near surface
emissions for "Soil", "Terpene" and "Auto" would suggest that their concen-
trations decrease with height and thus result in a lower effective than true
extinction efficiency.
A similar analysis was performed on nephelometer and particle data
from the South Lake Tahoe site. Nephelometer derived particle extinction
values corresponding to each of the particle sampling periods were averaged.
No satisfactory model was found which could encompass the six month period.
Two seasonally dependent models were required to adequately explain the
visibility, a summer model based on "Long Range Transport" and a fall model
based on "Residential Wood Smoke." The following equations resulted from the
analysi s.
bp ext = 0.00963 LRT + 0.00509 NLRT
b« = 0.02031 RWS + 0.00619 NRWS
Pf»ext
29
-------
TABLE X. LAKE TAHOE BASIN PARTICLE EXTINCTION BUDGET
Overal1 Summer Fal 1
Source km"1 % km"1 % km-1 %
Soil
0.0003
2.0
0.0005
3.1
0.0002
1.4
Auto
0.0001
0.9
0.0001
0.5
0.0002
1.0
Long Range
Transport
0.0089
54
0.0090
56
0.0089
53
Residential
Wood Smoke
0.0046
28
0.0027
17
0.0058
35
Terpene
0.0024
15
0.0038
24
0.0016
9.3
Other
0.0000
0.1
0.0000
0.1
0.0000
0.1
Total
0.0165
100
0.0161
100
0.0167
100
30
-------
t,le bp ext' LRT' f,LRT rePresent the sumner particle extinction, "Long Range
Transport," and "Non-Long Range Transport" parameters, while b , RWS, NRWS
P* ,ext
represent the fall particle extinction "Residential Wood Smoke" and "Non-Resi-
dential Wood Smoke" parameters. The multiple R values for the summer and fall
regressions are 0.87 and 0.85 respectively.
Table XI shows the results of employinq these models to determine the
particle extinction budqet usinq the mean particle values shown in Table VIII.
As a comparison of the two tables shows, the wood smoke is the major con-
tributor overall (averaqe of sumner and fall) to both the source and the
extinction budget, however the contribution is rouqhly 3 times qreater in
the fall than in the summer. The lonq ranqe transport and terpene parameters
are approximately equal in both summer and fall. The values of terpene and
transport are lower in the fall than in the summer.
SUMMARY AND CONCLUSIONS
Results of this study indicate generally good visibility at Lake Tahoe
with a summer/fall median visual ranqe for the basin greater than 150 km.
Somewhat poorer visibility in the urbanized South Lake Tahoe area with median
visual ranqe for the sane period of about 70 km, demonstrates an effect on
visibility from local pollutant sources. Basin wide visibility was measured only
durinq the sumner/fall intensive monitoring season. The degree to which this
six month period is representative of annual basin wide visibility remains an
area of speculation. The nepheloneter measurements at South Lake Tahoe indi-
cate better visibility in the winter/sprinq period than in summer/fall (median
visual ranqes of about 90 km and 70 km respectively). Assuming the sane trend
for the basin wide visibility would lead to the conclusion that annual basin
wide median visual range is greater than that reported for the summer/fall
monitoring period.
Multiple linear regression analysis indicates that visibility is siqmf-
lcantly related to fine particle concentration but is not related to coarse
particle concentration. Automotive emission, soil, and sulfates account for
approximately 20 percent of the mass of fine particles. The remaimnq 80 per-
cent is composed of elements lighter in atomic weiqht than sodium. These were
assumed to be composed of nitrates from out of basin sources, and carbonaceous
material resultinq from residential wood burmnq and terpene emissions by con-
iferous forest. A model was developed to estimate the relative contribution
of each of these materials for each particle sampling period. Except at South
Lake Tahoe, terpenes represented nearly half the fine particle mass, long range
transport (nitrates + sulfates + hydrocarbons) represented 30 percent and wood
smoke 10 to 20 percent. In the South Lake Tahoe urban area, residential wood
smoke, terpenes, and long ranqe transport accounted for approximately 40, 25,
and 15 percent of fine particle mass concentration respectively.
Since fine particles from different sources do not necessarily have the
same optical extinction efficiency, the relationships between visibility and
source categories were determned using multiple linear regression. For basin
wide visibility the most important sources of impact are long range transport,
residential wood smoke and terpenes, accounting for approximately 50, 30, and
31
-------
TABLE XI. SOUTH LAKE TAHOE PARTICLE EXTINCTION BUDGET
Overall Summer Fall
Source km"* % km"* % km"* %
Soil
0.0027
4.0
0.0034
7.8
0.0021
2.0
Auto
0.0062
8.8
0.0039
9.0
0.0084
8.4
Long Range
Transport
0.0103
14
0.0142
33
0.0064
6.4
Residential
Wood Smoke
0.0418
58
0.0079
18
0.0757
76
Terpene
0.0105
15
0.0138
32
0.0072
7.1
Other
0.0001
0.2
0.0001
0.2
0.0002
0.1
Total
0.0716
100
0.0433
100
0.1000
100
.32
-------
15 percent of the visibility impact respectively. In South Lake Tahoe resid-
ential wood smoke accounts for nearly 60 percent of the impact while long
range transport and terpenes account for about 15 percent each and automotive
accounts for nearly 10 percent.
v The implications of these findings are that approximately 70 percent of
the basin-wide visibility impact and 30 percent of the South Lake Tahoe visibility
impact are caused by sources beyond the reasonable control of Tahoe Regional
Planning Agency (i.e., natural and long range transport emissions). Of the con-
trollable sources, wood smoke for residential heating affords the greatest
degree of potential visibility improvement. This is especially true at South
Lake Tahoe in the colder seasons. Though automotive sources and to some degree
wind blown soil sources are controllable, their share of the extinction budget
is quite small and therefore the potential visibility improvement by their
control is al so smal 1.
Further studies of visibility and particles are warranted at Lake Tahoe.
Characterization by analytical techniques of the light element fraction of the
fine particles would be an important check on the estimation technique employed
in this work. In such work, it's expected that distinguishing between wood
smoke and terpene carbonaceous materials would not be triveal since smoke from
coniferous wood contains a large quantity of terpene hydrocarbons . Particle
sampling and visibility monitoring should be conducted in a routine way for at
least two years so that yearly variations can be investigated. This would
also allow a much better evaluation of the relationships between visibility
and particle source categories on a seasonal basis.
33
-------
REFERENCES
1. A. P. Waggoner, R. E. Weiss, N. C. Ahlquist, D. S. Covert, S. Will, R. J.
Charlson. "Optical characteristics of atmospheric aerosols," Atmos.
Environ. 15:1891(1981).
2. 6. T. Wolff, M. A. Ferman, N. A. Kelley, D. P. Stroup, M. S. Ruthkosky.
"The relationships between the chemical composition of fine particles
and visibility in Detroit metropolitan area," JAPCA 32:1216(1982).
3. T. G. Dzubay, R. K. Stevens, C. W. Lewis, D. H. Hern, W. J. Courtney,
J. W. Tesch, M. A. Mason. "Visibility and aerosol composition in
Houston, Texas," Environ. Sci. Technol. 16:514(1982).
4. R. J. Charlson. "Atmospheric visibility related to aerosol mass concen-
tration," Environ. Sci. Technol. 3:913(1969).
5. P. J. Groblincki, G. T. Wolff, R. J. Countess. "Visibility-reducing
species in the Denver 'Brown Cloud'-I. Relationships between extinc-
tion and chemical composition" Atmos. Environ. 15:2473(1981).
6. Environmental Protection Agency (EPA), "Protecting Visibility: An EPA
Report to Congress," EPA-450/5-79-008 (1979).
7. Environmental Protection Agency (EPA), "Interim Guidance for Visibility
Monitoring," EPA-450/2-80-082 (1980).
8. W. Malm. "Considerations in the measurement of visibility," JAPCA
29, 1042 (1979).
9. W. C. Malm, M. Pitchford, A. Pitchford. "Site specific factors influenc-
ing the visual range calculated from teleradiometer measurements,"
Atmos. Environ. 16:2323(1982).
10. W. E. K. Middleton. 1952. Vision through the atmosphere. University of
Toronto Press, Toronto, Canada.
11. T. A. Cahill, L. L. Ashbaugh, J. B. Barone, R. A. Elred, P. J. Feeney,
R. G. Flocchini, C. Gooddart, D. J. Shadoan, G. W. Wolfe, "Analysis
of respirable fractions in atmospheric particulates via sequential
filtration," JAPCA 27:675 (1977).
12. T. A. Cahill, R. A. Eldred, J. B. Barone, L. L. Ashbaugh, "Ambient Aerosol
Sampling with Stacked Filter Units," Report No. FHWA-RD-78-178,
Feb. 1979. Federal Highway Administration Report Available from the
NTIS, Springfield, VA 22161.
34
-------
13. J. B. Barone, L. L. Ashbaugh, R. A. El red, T. A. Cahill, "Further invest-
igation of air quality in the Lake Tahoe Basin," a final report to
California Air Research Board, contract number A6-219-30, March 27,
1979.
14. v R. J. Charlson, N. C. Ahlquist, H. Selvidge, P. B. MacCready Jr. "Moni-
toring of atmospheric aerosol parameters with the integrating nephel-
ometer," JAPCA 19:927 (1969).
15. N. Draper, H. Smith, Applied Regression Analysis, John Wiley Interscience,
New York, NY. 1966.
16. A. P. Waggoner, R. E. Weiss, "Comparison of fine particle mass concentra-
tion and light scattering extinction in ambient aerosol," Atmos.
Environ. 14:623 (1980).
17. M. Pitchford, "The relationship of regional visibility to coarse and fine
particle concentration in the Southwest," JAPCA 32:814 (1982).
18. J. B. Barone, T. A. Cahill, R. E. Eldred, R. G. Flocchini, D. J. Shadoan,
T. M. Dietz, "A multivariate statistical analysis of visibility
degradation at four California cities," Atmos. Environ. 12:2213
(1978).
19. J. Trijonis, M. Davis. "Development and Application of Methods for
Estimating Inhalable and Fine Particle Concentrations from Routine
Hi Vol Data," Final Report to California Air Resources Board,
contract number A0-076-32, December 1981.
20. C. E. Lyons, I. Tombach, R. A. Eldred, F. P. Terraglio, J. E. Core,
"Relating particulate matter sources and impacts in the Willamette
Valley during field and slash burning" presented at the 72nd Annual
APCA Meeting, June 24-29.
21. California Air Resources Board, "California Air Quality Data," Volume
XI, numbers 3 and 4; Volume XII, numbers 2 and 3.
22. D. B. Turner, "Diurnal and day to day variations of fuel usage for space
heating in St. Louis, MO." Atmos. Environ. 2:339 (1968).
23. D. T. Tingey, M. Manning, L. C. Grothaus, W. F. Burns. "Influence of
light and temperature on monoterpene emission rates from slash
pine," Plant Physiol. 65:797 (1980).
24. "Ozone and Other Photochemical Oxidants," Vol. I National Research
Council, National Academy of Sciences, Washington, D.C. 1976.
25. D. S. Ensor, M. J. Pilat. "Calculation of smoke plume opacity from par-
ticulate air pollution properties," JAPCA 21:496 (1971).
26. R. A. Rasmussen, Oregon Graduate Center. Private Communication, December
1982.
35
------- |