Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
CA/DOH/AIHL/SP-29·
·i .• ,:
\. ,.
l ; \ J.
' ( ;,
lt I i }
VISIBILITY REDUCTION AS RELATED TO AEROSOL CONSTITUENTS
Interagency Agreement No. ARB A1-081-32
Final Report
October 1983
Prepared by
B. R. Appel, Y. Tokiwa, J. Hau, E. L. Kothny, E. Hahn and J. J. Wesolowski
Air and Industrial Hygiene Laboratory California State Depar~ment of Heal th., Servi~es
2151 ·Berkeley Way , · ·Berkeley Way
.. Ber~~ley, ·· California 94704
Board"\,
iaion. ,·-:·· . . . i:r;- ~~~,?,~~~es
The statements and conclusions in this re-port are those of the Contractor
and not necessarily those of the State Air Resources Board. The mention of
commercial products, their source or their use in connection with material
re-ported herein is not to be construed as either an actual or implied
endorsement of such products.
- 2 -
ABSTRACT
A laboratory and field study was performed to assess the contribution to
visibility reduction of both light-scattering and absorJ)tion by air
pollutant particles and gases. Gaseous precursors to important visibility
reducing aerosol species were measured. Emphasis was placed on minimizing
sampling artifacts for nitrate and sulfate since ~revious visibility studies
were generally subject to substantial errors from these sources. Optical
techniques for measuring the particle absorption coefficient and elemental
carbon were evaluated. The aerosol species measured were fine and coarse
particulate mass, sulfate, nitrate, and elemental carbon, plus organic
carbon and ammonium ion. The gases measured were nitric acid, NH 3, so 2,
N02, and 03. Sampling was done at San Jose, Riverside and downtown
Los Angeles.
At all sites light scattering by sulfate, nitrate and elemental carbon
particles contributed more than half of the light extinction. Light
absorption by ~articles, due almost ex c 1 us i ve ly to e 1 emen ta 1 carbon,
contributed 10 to 20% of the extinction. The light-scattering efficiency of
fine particulate nitrate appeared to be higher than that of sulfate, in
contrast to the findings of most prior studies.
- 3 -
ACKNOWLEDGEMENT
We wish to acknowledge the assistance of Vincent Ling who performed many of
the carbon determinations and aided in data analysis. Meyer Haik also
performed carbon determinations. SuzAnne Twiss supervised the multiple
regression analysis of the data. Dr. Allen Waggoner, University of
Washington, provided the modified integrating nephelometers together with
detailed instructions and advice on their calibration and use. Dr. Ray
Weiss, University of Washington, provided valuable comments and aid in
setting up and com-paring results for the integrating plate method. Dr.
Stephen Cadle, General Motors Research Laboratory, provided analyses
permitting interlaboratory comparison of our procedures for elemental carbon
measurement. Drs. T. Novakov, H. Rosen, A. Hansen and L. Gundel, University
of California, Lawrence Berkeley Laboratory, provided information on the
design of their laser transmission method, samples for interlaboratory
comparison, and useful discussions. Dr. John Trijonis provided many
valuable comments on the final report.
Sampling sitea .were furnished by Professor Kenneth MacKay, San Jose State
University, Dr. Arthur Winer, University of California, Riverside, and Mr.
Frank Baumann, Southern California Branch Laboratory Section, Laboratory
Services Branch, California Department of Health Services.
Finally, we acknowledge Dr. Douglas Lawson and Dr. John Holmes, Air
Resources Board Research Division, for their helpfulness and sup-port
throughout this program.
This report was submitted in fullfillment· of Interagency Agreement No. A1-
081-32, Visibility Reduction as Related to Aerosol Constituents by the
California De-partment of Heal th Services under the sponsorship of the
California Air Resources Board. Work was completed as of November 10, 1983.
- 4 -
TABLE OF CONTENTS
Abstract Acknowledgements List of Figures List of Tables
I. Summary and Conclusions 12
II. Introduction
A. Visibility and its Relation to Aerosol Constituents............. 17
B. Limitations of Prior Studies.................................... 20
C. Objectives of the Present Study................................. 24
D. Strategy........................................................ 24
III. Measurement of the Particle Absorption Coefficient (b )ap
A. The Integrating Plate Method (IPM).............................. 26
B. A Laser Transmission Method (LTM).............................. 29
c. Comparison of B ap Determinations by the IPM and LTM............. 30
IV. Determination of Elemental Carbon (C)e
A. Introduction • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 32
B. The Laser Transmission Method for Ce ••·•••·····•···••··•·•·•·•··• 32
C. Intermethod and Interlaboratory Comparisons..................... 33
D. Comparison of C by Optical Methods with Particle Absorption Coifficient Results •••••••••••••••••••••••••••••••••• 35
v. Measurement of the Scattering Coefficient for Liquid Water
A. Introduction • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 37
B. Method Evaluation and Correlation with Aerosol Constituents..... TI
VI. Assessment of Sampling Artifacts
A. Artifact ~articulate Sulfate on Glass Fiber Filters............. 39
- 5 -
B. Artifact Particulate Nitrate on Glass Fiber Filters............ 39
c. Loss of Particles in the Denuder of the Particulate Nitrate (PN) Sampler. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 40
D. Loss of Non-Volatile Carbon in the Denuder of the ?articulate Carbon Sam-pler. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 41
E. Retention of Carbonaceous Material on Hi-Vol Filter Samples..... 41
VII. Atmospheric Results -- Part I.
A. Tabulation of Results........................................... 42
B. Summary of Visibility Results................................... 42
c. Diurnal Variation of the Dominant Aerosol Species and Ozone..... 43
D. Size Distribution for Aerosol Constituents...................... 43
E. Calculated vs. Observed Ammonium. Ion Levels..................... 45
F. True Particulate Carbon (PC) and Total Low Volatility Carbon (LVC).................................................. 46
G. Atmospheric Nitric Acid......................................... 46
VIII. Atmospheric Results -- Part II. Statistical Analyses
A. Correlation between Pollutant Concentrations.................... 48
B. Correlation between Single Pollutant Concentrations and Visibility Parameters......................................... 48
C. Multiple Regression Analyses between Particle Scattering Coefficients and Pollutant Concentrations •••••••••••• 49
D. Average Contributions of Aerosol Species to the Scattering and Total Extinction Coefficients. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 51
E. Comparison of Present Results with those for General Motor's Denver Visibility Study....................................... 52
F. Comparison of the Present Results to those Reported by Trijonis et al( 48) •••••••••••••••••••••••••••·••••••••••••••••• 53
- 6 -
IX. References
Appendices
A. The Impact of Increased Diesel Emissions on Visibility......... 62
B. Operation of the Integrating Nephelometers..................... 65
C. Evaluation of Methods for Absorption Coefficient Measurement... 66
D. Calibration of Optical Methods for Elemental Carbon............ 79
E. Measurement of Ambient Temperature and Relative Humidity....... 85
F. Experimental Difficulties with the True Particulate Carbon Measurements.......................................... 86
- 7 -
FIGURES
Number
1. Light Scattering by Aerosols as a Function of Particle Diameter Com~uted for Unit Density Spherical Particles of Refractive Index 1.5.......... 19a
2. Schematic of Integrating Plate Method (IPM)............................. 26a
Comparison of b Measurements at Two Particle Loadings ••••••••••••••••• 28a ap
Comparison of the Integrating Plate Method and Laser Transmission Techniques for Absorption Coefficient Measurement ••••••••••••••••••••• 29a
5. Schematic Diagram of Laser Transmission Method, Viewed From Above ••••••• 30a
6. Comparison of b Measurements by the IPM and LTM Methods ••••••••••••••• 30b ap
Elemental Carbon by the LTM vs. Reflectance Method•••••••••••••••••••••• 34a
8. Relation between Absorption Coefficient and C for Denver Winter Aerosol 35a e
9. Elemental Carbon by LTM vs. Absorption Coefficients..................... 35b
10. Elemental Carbon by the RM vs. Absorption Coefficients.................. 35c
11. Mean Elemental Carbon by LTM and RM vs. Absorption Coefficients......... 35d
12. Artifact Sulfate on Glass Fiber Filters, 4-hour Hi-Vol Samples.......... 39a
13. Nitric Acid Retention on Glass Fiber Hi-Vol Filters ••••••••••••••••••••• 40a
14. Nitric Acid Retention of Glass Fiber Hi-Vol filters ••••••••••••••••••••• 40b
15. Nitric Acid Retention of Glass Fiber Hi-Vol Filters ••••••••••••••••••••• 40c
16. Nitric Acid Retention of Glass Fiber Hi-Vol Filters ••••••••••••••••••••• 40d
17. Loss of Non-Volatile 'Particles in the Denuder for Nitrate Sampling •••••• 40e
18. Loss of ton-Volatile Particles in the Denuder of the Particulate Carbon Sampler. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 41 a
19. Observed vs. Calculated 16-hour Total Carbon............................ 41b
20. Mass Comparisons 1982 ARB Visibility Study.............................. 43a
21. Comparison of Fine and Total Particulate Sulfate........................ 43b
22. Comparison of Fine and Total Particulate Elemental Carbon............... 43c
23. Fine Elemental Carbon vs. Total Carbon in 4-hour Samples................ 43d
- 8 -
Number Page
24. Fine Particle Mass vs. Fine Particle Elemental Carbon••••••••••••••••••••• 43e
25. Total Particle Mass. vs. Fine Particle Elemental Carbon•••••••••••••••••••• 43f
26. Comparison of Filter Collected Carbon with and without a Diffusion Denuder for Carbonaceous Vapors ••••••••••••••••••••••••••••••••••••••••••••••••• 43g
'Z7. B Values San Jose••••••••••••••••••••••••••••••••••••••••••••••••••••••••• 44a
28. B Values Riverside •••••••••••••••••••••••••••••••••••••••••••••••••••••••• 45a
29. B values Los Angeles •••••••••••••••••••••••••••••••••••••••••••••••••••••• 45b
30. Diurnal Variations in Mean Pollutant Concentrations (San Jose) •••••••••••• 45c
31. Diurnal Variations in Mean Pollutant Concentrations (Riverside) ••••••••••• 45d
32. Diurnal Variations in Mean Pollutant Concentrations (Los Angeles) ••••••••• 46a
Scatter Diagram of Fine Particulate Mass Against the Dry Particle Scattering Coefficient ••••••••••••••••••••••••••••••••••••••••••••••••••
34. Comparison of Calculated and Observed b' ext•••···•··•··•··•···········••·· 51a
C-1. Linearity of LTM Detector Response •••••••••••••••••••••••••••••••••••••••• 74a
C-2. Comparison of AIHL and Lawrence Berkeley Laboratory LTM Instruments Analyzing the Same Atmospheric Particulate-loaded Millipore Filters ••••• 76a
D-1. Size distribution for Aged Carbon Particles from a Butane Flame ••••••••••• 80a
D-2. Calibration of LTM with Aged Carbon Particles ••••••••••••••••••••••••••••• 80b
D-3. Calibration of LTM with LBL Black Carbon Samples •••••••••••••••••••••••••• 81a
D-4. Light Absorption Efficiency for Graphitic Carbon•••••••••••••••••••••••••• 81b
D-5. Calibration of Reflectance Meter with Aged Carbon Particle Samples •••••••• 82a
D-6. Calibration of Reflectance Meter with Aged Carbon Particle and LBL Black Carbon Samples •••••••••••••••••••••••••••••••••••••••••••••••••••••• ~ ••• 82b
E-1. Temperature Comparisons 1982 ARB Visibility Study••••••••••••••••••••••••• 85a
E-2. RH Comparisons 1982 ARB Visibility Study•••••••••••••••••••••••••••••••••• 85b
- 9 -
TABLES
Number
Summary of Extinction Coefficients per Unit Mass Obtained in Various Regression Studies •••••••••••••••••••••••••• 18a
2 Contribution of the Chemical Species to the Extinction Coefficient, for Denver••••••••••.•••••••••••..•.••••.•• 21a
3 Description of Samplers Employed for Visibility Study •••• 25a
4 .Analytical Strategy for Visibility Study ••••••••••••••••• 25c
5 Calibrations of Optical Methods for Elemental Carbon ••••• 33a
6 Comparison of Total Carbon and Elemental Carbon (C) Values ( JJg/m3) ••••••••••••••••••••••••••••••••• ~ ••••••• 33b
7 Comparison of Elemental Carbon Measurements by General Motors' Combustion and AIHL Optical Methods ••••••••••• 34b
8 Pearson Correlation Coefficients Between b , Pollutant and Meteorological Parameters •••••••••• ~~ •••••••••••••• Tia
9 Multiple Regression Analysis Between b w and Aerosol 8Constituents ••••••••••••••••••••••••••••••••••••••••••• 38a
10 Summary of Atmospheric Results ••••••••••••••••••••••••••• 42a
11 Explanation of Headings Used in Table 10 ••••••••••••••••• 42c
12
13
Mean Concentrations of Aerosol Constituents and Meteorological Parameters ·········••••···••··•·····•··•
Summary of Optical Measurements at San Jose, Riverside and Los Angeles ••••••••••••••••••••••••••••••
42d
42e
14 Pearson Correlation Coefficients Between Aerosol Constituents Ozone and Meterological Parameters 48a
15
16
Correlation Between Fine Aerosol Constituents (or Aerosol Precursors) and b ••••••••••••••••••••••••• sp
Pearson Correlation Coefficients Between Visibility Parameters, Individual Pollutants and Meterological Parameters •••••••••••••••••••••••••••••••
48b
48c
- 10 -
Number
17
18
Multiple Regression Analysis Between Aerosol Constituents and b ••••••••••••••••••••••••••••••••••• sp
Average Percent Contributions of Chemical Species to the Particle Scattering Coefficient, b and to the Extinction Coefficient, b' t •••• ~Y............... . ex
50a.
52a
C-1
C-2
Precision of Parallel Sampling for b Measurement ••••••• ap
Effect on Transmittance of Overlaying Loaded Nuclepore Filter with a Microsco~e Slide •••••••••••••••••••••••••
68a
69a
C-3 Effect of Neutral Density Filter on t T of Loaded Filters ·•·••••··•••••·•··•···••·······•·······•• 70a
C-4 Effect of Particle Orientation on% T •••••••••••••••••••• 71a
C-5 Interlaboratory Comparison of IPM (% T) •••••••··••••••••• 72a
C-6 Batch-to-Batch Variations in Mean Transmittance Values for 0.4 um Pore Size Nuclepore Filters ••••••••••••••••• 73a
C-7 Influence of Particle Orientation with the IPM and LTM •••••••••••••••••••••••••••••••••••••••••••••••• 76b
C-8 Influence of Wavelength and the Discrepancy Between the IPM and LTM Methods •••••••••••••••••••••••••••••••• 78a
D-1 Size Distribution for Carbon Particles ••••••••••••••••••• 79a
- 11 -
I.· SID.i!MARY AND CONCLUSIONS
A laboratory and field study was performed to assess the contribution
to visibility reduction at both light-scattering and absorption by air
pollutant -particles and gases. Gaseous precursors to important
visibility-reducing aerosol species were also measured. Em-phasis was
placed on minimizing sampling artifacts for pollutant particles since
preceding visibility studies were generally subject to substantial
errors from this source. 0-ptical techniques to measure the particle
absorption coefficient and elemental carbon were evaluated.
The aerosol species measured were fine and coarse particulate sulfate,
nitrate, elemental carbon and mass. In addition, total pa rt i culate
o rg a ni c carbon and ammonium ion were determined. The gases measured
were nitric acid, NH3, S0 2 , N0 2 and o3• Artifact sulfate formation was
minimized by using Teflon filters. Particulate nitrate and nitric acid
were measured using the denuder difference technique. Particulate
organic carbon was sampled with two conventional hi-vol samplers with
glass fib.er filters, providing both 4-hour and 16 to 24 hr samples. In
addition, a- diffusion denuder plus a quartz filter and a fluidized
Al 2o3 bed were used in an effort to determine particulate carbon whi 1 e
minimizing both positive and negative sampling errors. However,
ex-perimental difficulties prevented the development of useful data from
this sampler.
The particle absorption coefficient was measured by the integrating
plate method using two samplers operating at different particle
loading rates. 'fillemental carbon was measured optically by both a
reflectance method and by a method which measured the absorption of a
He-Ne laser beam. Two integrating nephelometers operated side-by-side
to measure light-scattering by particles with and without heating the
incoming air. The difference in results between these instruments was
used to approximate light-scattering due to liquid water.
- 12 -
Sampling was done for a total of 109, 4-hour periods, generally between
0800 and 2400 hours, July 10 - August 8, 1982, at San Jose, Riverside
and downtown Los Angeles. The mean visual range encountered, as
estimated from the calculated extinction coefficients and the
Koschmeider equation, were 43 Km at San Jose, 15 Km at Riverside and 13
Km at downtown Los Angeles. The mean four-hour average ozone
concentrations ranged from 0.02 to 0.(17 ppm. Thus relatively light
photochemical smog was experienced.
Light scattering by dry particles was the largest contributor to the
total extinction (b' t) at all sites, representing more than half of ex
the extinction coefficient. Light absorption by particles, due almost
exclusively to elemental carbon, was relatively unimportant (.s_ 12% of
b' t) in the South Coast Air 'Basin where high concentrations of ex
sulfate and nitrate were present. At San Jose, however, absorption
averaged nearly 22% of b' t· The apparent contribution of scatteringex
from liquid water was relatively high at Riverside and Los Angeles but
small at San Jose, paralleling the concentrations of the hygroscopic
aerosol species, sulfate and nitrate. Light absorption· by gases, due
only to N0 2 , was small at all sites.
Considering both light scattering and light absorption, and using
multiple regression analysis, fine sulfate and fine nitrate (and
associated water), fine elemental carbon, and nitrogen dioxide
contributed on average, 30%, 36%, 20%, and 8%, respectively, to the
total extinction (excluding Rayleigh scattering).
Sulfate, nitrate, elemental carbon, and total carbon were predominately
in the fine ( < 2. 5 µm) fraction averaging 92%, 60%, 80%, and 62%
respectively, measured by comparing results with and without a cyclone
preceding the sampler. The percentage of fine total carbon can be
compared to the range 56 to 66% previously obtained at three sites
using a dichotomous sampler (DS). We previously suggested that the
proportion of coarse particle carbon with a DS ma.y be erroneously high
- l3 -
because of the low face velocitv for the coarse narticle filter
(34,40). The nresent data fail to su-pnort this.
Multinle regression analyses between the drv particle scattering
coefficient and aerosol s~ecies concentrations vielded the
relationshin:
where F = fine
C • coarse
Tlms the scattering efficiency per unit mass of nitrate apl)ears to be
greater than that for fine sulfate, in contrast to most -previous
studies. The scattering efficiency for fine elemental carbon (C )e
was
similar to that for fine sulfate. The results for coarse sulfate are
surprising, iml)lying a scattering efficiency twice that for fine
sulfate. We believe that some material whose concentration is much
higher but pro-portional to that of coarse sulfate is the actual
scattering source. Sea salt chloride, for example, is about seven
times the weight concentration of sea salt sulfate. A significant
amount of sea salt would be ex1'ected in the coarse fraction. Re~lacing
the coarse sulfate concentration with a material seven times more
abundant would decrease the scattering efficiencv per unit mass by an 4 1
equal factor of seven (i.e. to< 0.02 x 10- m- /~g/m3). The remaining
aerosol species, coarse nitrate, coarse elemental carbon, and organic
carbon, did not exhibit a significant scattering efficiency in this
evaluation.
Multi~le regression analyses between the measured scattering
coefficient for liquid water and the concentrations of the hy'grosco-pic
aerosol s-pecies S0 4= and N03-, using various functions of relative
humidity, showed only moderate correlation (r i 0.81). This may be
- l4 -
indicative of lack of speciation data for sulfate since, for example,
H zSO 4 and ( 'f\'.JH 4 ) 2SO 4 differ substantially in hygroscopici ty. It may
also indicate error in measuring the scattering coefficient for liquid
water, since species other than liquid water may be volatilized on
entering a heated nephelometer inlet (e.g. organic comuounds, 'tifH 4N03).
Artifact particulate nitrate formation on ~lass fiber filters (prefired
Schleicher and Schuell, 1qs1 EPA Grade) was assessed for both 4-hour
and 16-hour atmospheric samples. With the short-term samples,
essentially all particulate nitrate and nitric acid was retained,
supporting the use of glass fiber as total inor~anic nitrate (TI'tll)
collectors. With the long-term sam'!)les, however, onlv at low nitrate
levels such as found at San Jose ( TIN i '32 iag/cm2) was all TIN
retained; at Riverside and Los Angeles with 't'IN values up to 147
µg/cm 2 , from 24 to 42% of the TIN ~enetrated the filters, presumably as
HNO 3.
Expected HN03 concentrations can be calculated from the concentration
of NH 3, temperature, and corresponding dissociation constant for
In general, no correlation between the observed and expected
HN03 levels was found. Similarily, the correlation between NH 3 or HN0 3
and the particle scattering coefficient remained relatively low.
The principal conclusions from this study are:
1. The light-scattering efficiency of fine particle nitrate appears to
exceed that of sulfate in contrast to most previous studies. Since
both positive and negative sampling artifacts were minimized we
believe these results to be correct.
2. Light absor-ption by elemental carbon plays a significant role in
visibility reduction in California, especially under conditions in
which the concentrations of fine sulfate and nitrate are relatively
low.
- 15 -
3. Particle--phase water contributed substantially to visibility
reduction at sites with relatively high levels of sulfate and
nitrate, species which can absorb moisture.
Recommendations for Further Studies
1., The significance of true particle phase organic com-pounds in visibility
reduction. Although organic carbon, based on hi-vol filter sampling, was
not a statistically significant contributor to visibility reduction,
sampling errors may have masked such a contribution.
2. The a c curacy of elemental carbon measurement by optical methods. While
the elemental carbon measurements reported are consistent with other
o-ptical measurements, calibration techniques for elemental carbon
measurements by optical methods require more work before their adoption
for routine monitoring of elemental carbon can be recommended.
3. The relative significance of aerosol constituents in visibility reduction
using the same.techniques at other times of the year.
4. Additional statistical analysis of results from the present study to
a) assess the significance of the aerosol and gaseous pollutants to
visibility reduction by site, b) assess the hypothesis that coarse sulfate
is functioning as a surrogate for sea salt, and c) assess the impact of
co-linearity of variable.
- 16 -
II. INTRODUCTION
A. Visibility and Its Relation to Aerosol Constituents
Visual range ( v ) is defined as the greatest distance at which an m
observer can distinguish a contrast between an object and its
background. At this distance the contrast of the object has reached
its threshold value, which is usually taken to be 2~. With this
threshold value, the simple Koschmeider relation for a perfectly
black object holds (1):
3.912 V :a -------
m bext
The atmospheric extinction coefficient, bext' may be written as
follows:
b =- b + b + b + bext ap ag sp sg
where: b = absorption by particlesap
b ag
= absorption by gases (essentially only N0 2)
b • scattering by particlesSl)
b = Rayleigh scattering by gases= 0.12 x 10_1io._ 1 sg
For a perfectly black object the visual range depends only on the iz
total extinction coefficient, b t' and is independent of the ratio ex of scattering to absorption (2).
Much of the emphasis in atmospheric visibility studies has been to
establish the chemical spa cies cont ri bu ting substantially to
- 17 -
visibility reduction (3-8). Typically b tis approximated by b , ex s~
as measured by the integra.ting nephelometer or as approximated from
human observer data. Indeed, b will often represent more than 2/3Sp of the total extinction coefficient. Multiple regression analyses
are performed with equations of the form:
= -where R = Remaining aerosol = TSP - SO 4 - NO 3
RH a relative humidity
Relative humidity de-pendence may, alternatively, be incorporated
into the sulfate and nitrate terms. In the cases where data on
carbonaceous material are available, additional terms [e.g.
b (benzene soluble organics)] a.re added.5
Table 1, taken from a recent survey by Trijonis,(3) summarizes the
scattering coefficients per unit mass for a series of visibility
studies. Princi-pal conclusions that ca.n be reached from these
studies ~re:
1. Sulfates, per unit mass, are the dominant contributor to
visibility reduction due to light scattering.
2. Nitrates, -per unit mass, can be an important contributor to
visibility reduction due to scattering. However, in locations
not dominated by photochemical smog, their contribution will be
relatively small because of the low nitrate levels.
The contribution of relative humidity as well as organics is
illustrated with results from a study by Hidy et al.: (4)
- 18 -
TABLE 1
LOCATION
SUMMARY OF EXTINCTION COEFFICIENTS PER UNIT MASS OBTAINED IN VARIOUS REGRESSION STUDIES.
(From Trijonis et al. Reference 3)
Locations in California.
Other Studies SOUTH COAST AIR BAS IN (White and Roberts 1977)Various Los Angeles Basin Sitest
(Cass 1979)Downtown Los Angeles
(Grosjean et al. 1976) Eastern Los Angeles Basin
(Leaderer and Stolwijk 1979)Los Angeles Int. Airport
The Present Study SOUTH COAST AIR BASIN Burbank
Long Beach
Ontario
t San Bernardino
SAN FRANCISCO BAY AREA AIR BASIN Oakland
San Jose
OTHER COASTAL LOCATIONS Paso Robles
San Diego
SAN JOAQUrn VALLEY AIR BASIN Bakersfield
Fresno
SACRAMENTO VALLEY AIR 8J\SIN Red Bluff
Sacramento
*Values marked by an asterisk are
EXTINCTION COEFFIC!EtlTS PER U~IIT M.il.SS (104 m)-l/(µg/m3)
Sulfates rlitrates Remainder of TSP
.01. .as. .015,.
.06 .04 .020
.11. NS. .008
.09 .OS NS*
.21 .04 ~,s
.16 .03 NS
•17. ~•s• · NS .., .15 NS .008
.16. .04* .011.,..
.11 • NS .013
•16,. .1s • r~s. .17 .11 NS
.12,. NS,. .019..,
.06 .OS .019
.12 .. NS,. .014.
.08 NS .014
NS* .06.., NS* NS .04 NS
.08.,, Dos* .008.,,
.,06 .as .006
.19. .04.,,, NS,.
.15 .OS NS
NS* .09. NS* NS .07 NS
NS.., 014* ~lS.,. M$ .07 HS
~s. .03* .009.,,, NS .04 .008
•16* .as• .014* .15 .03 .014
based on the nonlinear RH regression model with insertion of average RH. Va1ues not so marked ~re based on the J;near RH model.
tBased on nephelometry data rather than airport visibility data.
- lBa-
TABLE 1 SUMMARY OF EXTINCTION COEFFICIENTS PER UNIT MASS OBTAINED IN VARIOUS REGRESSION STUDIES (Continued).
(From Trijonis et al. Reference 3)
Locations in the ~ortheast and Rocky~ountain Southwest.
EXTINCTION COEFF!C IENTS PER U1'1IT MASS (104 m)-l/(~g/rn3)
LOCATION Sulfates ~Ii tra tes Remainder of TSP
NORTliEAST
(Trijonis and Yuan, 1978b) Chicago .04..,
.03 NS,. NS
NS* NS
Newark NS.., NS,. .026* .06 NS .014
Cleveland .08 NS* NS* .07 tlS NS
Lexington .06. .06
NS* NS
NS* .019
Charlotte. .11... NS* NS* .11 t4S NS
Columbus .12... .09. NS* .13 .06 NS.
(Leaderer and Stolwijk 1919) Ne•,., Yorkt .07 .OS NS
New York .10 NS NS
New Haven .16 NS- NS
St. Louis .08 NS NS
ROCKY MOUNTAIN SOUTHWEST
(Trijonis 1979)
Phoenix County Data NASN Data
.04
.03 .OS .OJ
NS NS
Salt Lake City .04* .04
•13,. .10
.004...
.004
*1/alues marked by an asterisk are based on the nonlinear RH regression Model ~ith insertion of average RH. Values not so marked are based on the linear RH model.
tBased on nephelometry d~ta rather than airport visib11ity data.
- 18b-
+ 0.025(total carbon) + 0.025(R)
RHwhereµ a 1 - 100
Ra remaining aerosol a TSP - S04 a
- N03- - total carbon
Some studies have incorporated relative humidity-dependent
sulfate terms as well (3,7).
The dominance of sulfate in contributing to light sea ttering
and the influence of relative humidity on light scattering can
be interpreted in relation to 'Figure 1. Growth caused by
pickup of water would be expected to enhance light scattering
for sulfate particles initially smaller than 0.5 µmin
diameter, since scattering efficiency exhibits a sharp maximum
at 0.5 lJDl• Sulfates in various urban areas around the United
States have been reported to have mass median diameters (mmd)
ranging from 0.1 to 0.6 lJDl (9). Nitrate mass median diameters
are much more varied. In California's South Coast Basin, 24-
hour mean values from 0.3 to 1.6 µm have been reported ( 9),
probably reflecting the proportions of fine mf4N0 3 and coarse
NaN03. Assuming a size distribution for NH~03 similar to that
for particulate sulfate, increasing relative humidity would
increase· the scattering efficiency for this material as well.
The limited data on carbonaceous particules suggest mean
diameters of several tenths of a micrometer. For example, mmd
values of about 0. 3 µm were found for both organic and
elemental carbon in Denver (38). This together with their low
hygroscopicity may explain their lesser scattering efficiency.
- 19 -
0 1 I I I I I J I I I I I I • . 03 .05 .07 0.1 0.2 0.3 0.5 o.7 1.0 2.0 3.0 5.0 7.0 10.0
Cl 0
·r-f
~ S... -f-l .10Cl (1) C) Cl 0
CJ en ,..... .08 ~ CV)
~ ~ bl}
~ 3-c:: '-.... .06 p r--f
IH ,-... (1) a ~ ~
0 ...., bD ,_,. .04"° c:: ""-../'11 •r-f
H (1) -f-l
~ t) .02 r/l
-f-l ..c:: bD
·r-f ~
Particle Diameter (pm)
Figure lo LIGHT SCATTERING BY AEROSOLS AS A FUNCTION OF PARTICLE DIAMETER. COMPUTED FOR UNIT DENSITY SPHERICAL PARTICLES OF REFRACTIVE INDEX lo5 (White and Roberts 1977)
B. Limitations of Prior Studies
1., Most -prior studies have neglected consideration of light
absorption by aerosol constituents.
f
f
2.. Studies based on visual range measurements obtained by a
human observer yield daytime visibility at one or two times
per day. Attempts to correlate these results with 24-hour
average -particulate sampling results (e.g. Ref. 7) are
subject to substantial error since -particulate loadings
will generally be higher during daytime hours.
(
3. Aerosol samples have ty-pically been collected on filters
subject to sampling errors. These include artifact
particulate sulfate and nitrate em-ploying glass fiber
filters, artifact -particulate nitrate for those using
cellulose ester filters, loss of -particulate nitrate
("negative artifact") from inert filters (e.g. Teflon), and
positive and negative artifacts with organics common to all
studies using filter sampling.
4. The aerosol species
limited in number.
measured were generally relatively
a. Artifact Particulate Sulfate (16,17,18)
Artifact particulate sulfate results from -partial
retention of S02 on filters and can yield, with 24-hour
atmospheric samples, errors in the range 3 to 6 ug /m 3 •
To illustrate the potential effect of this error, the
NASN annual geometric mean sulfate value for down town
Los Angeles in 1974 was s.4 ug/m 3 • A -positive error of
4 ug/m 3 in this value would renresent about a 50t
error. The coefficient for the contribution of sulfate
- 20 -
to b would be expected to be too low by an equalSp
percentage. Perhaps more significant, the variability
introduced into sulfate measurements can decrease the
correlation with visibility parameters.
The importance of submicron aerosols in visibility
reduction by light scattering leads to the desirability
of size-segregated particle sampling in visibility
aerosol chemistry correlation studies.
Recent visibility studies have considered more than
particle scattering contributions. In urban areas the
particle absorption coefficient, due primarily to
graphitic carbon, can contribute substantially to b t ex
(6,12-14). For example Groblicki et al. (Table 2)
reported a mean contribution to b t of 31% due to ex
light absorption by particulate carbon for 41 days of
sampling in Denver ( 6). The contribution of light
absorption by particles to the total extinction
coefficients in California urban areas has only
recently been reported (58).
In Denver the dominant source of atmospheric graphitic
carbon was concluded to be from the combustion of wood.
However, in California urban areas, mobile source
emissions, especially from diesel-powered trucks, are
the predominant source of graphitic carbon. The
increasing use of diesel-powered passenger cars may
serve to increase this source of light extinction.
General Motors has estimated that by the year 2000, 25%
of all passenger vehicles will be diesel-powered (15).
A simple calculation (Appendix A) permits estimation of
a > 12% decrease in mean visual range in downtown Los
Angeles by the year 2000, relative to the 1974-1 976 ! i I ~ /1
- 21 -
11
TABLE 2 CONTRIBUTION OF THE CHEMICAL SPECIES TO THE EXTINCTION COEFFICIENT, FOR DENVER (NOVEMBER - DECEMBER 1978)a
Fine Particulate Species Mean% Contribution
20.2(NHt+) 2S0 t+ and associated H20
NHt+NO 3 and associated H2O 17.2
Organic C 12.5
Elemental C (scattering)
Elemental C (absorption)
Elemental C (total) 37 .7b
Remainder fine particulate 6.6
-2.:1.. TOTAL 1oo.
a. Data of P. J. Groblicki et al., Reference 6.
b. Includes fine and coarse particulate C • e
- 21a -
mean (7. 5 miles), due only to
in gra-phitic carbon, assuming
emissions.
the associated increase
no controls on such
If used for short-term sampling, glass fiber filters
would vield substantially greater artifact· sulfate,
expressed in µg/m3.
b. Nitrate Sampling Errors
(
Artifact -particulate nitrate results from retention of
gaseous HN03. Under atmos-pheric conditions,
and with 4-hour samples, essentially all HNO 3 is
retained by glass fiber filters resulting in errors
which can re-present 50% of the observed nitrate ( 21 ) •
As a result, the corresponding nitrate coefficient in
regression equations with visibility parameters may be
too low by a factor of about two. Similarly, in short
term sampling with cellulose aceta.te filters, 12%
retention for HNO 3 was observed (20) leading. to a
corresponding negative error in the nitrate coefficient
in correlations with visibility.* Studies employing
inert filters (e.g. Teflon) vield nitrate
concentrations which are lower limits to the true
particulate nitrate values because of com-plete
penetration of HN0 3 , and the possibility of nitrate
loss by volatilization and reaction with strong acid
particles (e.g. H2S0 4 ) and gases (e.g. l-ICl) (21-2'3).
In such cases the nitrate coefficients in the
visibility re~ression equation would be ex-pected to be
*As with sulfate,
affected.
the coefficient for the remaining aerosol term is also
- 22 -
too high by an amount which depends on meteorological
variables and concentrations of airborne acids.
In addition to altering coefficients, the variability
in the re lationshi-p between true particle nitrate and
that measured by simple filter methods can severely
decrease correlation coefficients with visibility
parameters.
c. Organic Particle Sampling Errors (24-34, 63)
Of the major classes of aerosol constituents least is
known about the errors in sampling organics. Many of
the organic materials retained on a filter exist in the
gas phase as well. The proportion in each phase
in the atmos-phere probably varies with atmos-pheric
conditions (e . .g. temperature and R.H.), the
concentration of the pollutants. The contribution of
organics to visibi li ty reduction should reflect the
fraction of the organics in the particle -phase at any
moment. Following collection on a filter, volatile
organics may be lost depending on the degree of
saturation of the air stream with each organic
compound. As an opposing source of error, gas -phase
organics may be retained by sorption on previously
collected particulate constituents and on the f i 1 t er
medium. As a result, neither the ma~nitude nor
direction of the ·error can be estimated without
specific measurements.
As with sulfate and nitrate, the variability introduced
in f i 1 t er sampling of organic -particles will probably
decrease or eliminate correlations with visibility
parameters.
- 23 -
c. Objectives of the Present Study
1. To assess the contribution to visibility reduction of light
scattering and absorption by specific pollutant -particles and
gases, measuring the aerosol s-pecies with minimum sampling
artifacts.
2. To measure atmospheric visibility by determining the total light
extinction coefficient due to light scattering and absorption in
the San Francisco Bay Area, and Los Angeles Basin.
3. To measure gaseous -precursors to important visibility-reducing
aerosols species to aid in data interpretation and assessment of
air quality.
4. To -provide further field trials of techniques to measure aerosol
species with minimum errors.
D. Strategy
This study included both a laboratory phase and a field sam-pling and
sam-ple analysis -phase. The former focused on the setup, calibration
and evaluation of techniques to measure b and elemental carbon ap (Ce). Atmospheric sampling was performed at San Jose, Riverside and
downtown Los Angeles for nine, 16-hour days at each site, during the
summer of 1982. Table 3 summarizes the sampling scheme. Fine and
total sulfate samples were collected minimizing artifacts using
Teflon filters. True fine particle nitrate (FPN) was measured with a
cyclone -plus acid gas denuder and total inorganic nitrate collector
(21,36). Total fine nitrate (-particle plus gas phase), FN, was
measured the same way as for FPN but without the denuder. Nitric
acid was then obtained by difference, FN-FPN.
True particulate carbon was sampled with a denuder, intended to
remove the gas -phase com-ponent of relatively non-volatile organic
- 24 -
com-pounds, followed by a particle filter and a fluidized bed of
Al20 3• A second unit without the denuder was intended to measure the
total of -r;,article plus gas phase com-ponents of such relatively non
volatile carbonaceous materials (40). Two hi-vol sam-plers were
operated to provide both short and long-term samples of carbonaceous
material. They were also analyzed for sulfate and nitrate for
measurement of sampling artifacts.
Samplers 9 and 10 (Table 3) collected fine particulate matter on 25
mm and 47 mm Nuclepore filters, provided samples for the integrating
plate method (41) differing by a factor of about five in filter
loading -per unit area. Sampler 11 provided a sample for measuring
b and C with the laser transmission method (42). Two integratingap e nephelometers were used, one operating at ambient tem-perature and the
second with the inlet air heated above ambient. The difference
provided an u-pper limit estimate of b , the scattering coefficient SW'
due to liquid water. The limitations of this approach are discussed
in Section V. Appendix 13 details the calibration and operation of
. these ne-phelometers. The· analytical strategy, totaling 2·,826
determinations, is shown in Table 4.
- 25 -
TABLE~ DESCRIPTION OF SAMPLERS mMPLOYED FOR VISIBILITY STUDY
Sam:e,ler No. Sampler Sam~lin__g Rate (1PM) Filter Size Time (Hr) S~ecies Measured
2 Cyclo~, acid-gas denuder, c Teflon and NaCl/W41 filter
20 47mm 4 True fine particle N0 3-,S0 4
1d
3a
Cyclone, teflon and NaCl/W41
Teflon and NaCl/W41 filter
filter 20
20
47mm
47mm
4
4
Fine particle N03 - + HN03, S0 4-,
1 -TSP, total particle N0 3 + HN0 3,
fine mass
True total S0 4
4 Organics denuder, quartz filter, Al203 fluidized bed
9.5 37mm 12 True particulate carbon, Ce
5 Quartz filtere, fluidized bed
Al203 9.5 37mm 12 Particulate+ low v.p. gaseous carbon, Ce
I\) VI p,
6
7
Quartz filterf, 2 oxalic acid impregnated filtersg
Hi-vol (prefired glass fiber)h
20
40 (cfm)
47mm
20 X 25cm
4
4
NH 4 +
,
-SO 4 ,
NH3
-NO 3 , C ,O
Ce, iCt
8
9a
10d
11
Hi-vol (prefired glass fiber)
Cyclone plus Nuclepore filterj
Cyclone plus Nuclepore filter
Cyclone,kquartze filter
Nephelometer, (MRI 1560)
Nephelometer, heated (MRI 1560)
Dew point and Ambient Temperature
40 (cfm)
8
8
27 .5
28 (cfm)
5 (cfm)
20 X 25cm
25mm
47mm
47mm
16-24
4
4
4
continuous
continuous
continuous
so 4=, N03-, C , C , Ct o e
b (integrating plate method)ap
b (integrating plate method)ap
b (laser transmission method), C ap e
b at ambient temperaturesp
b for dry particlessp
Relative humidity and temperature
Hygrothermograph continuous Relative humidity and temperature
TABLE 3 DESCRIPTION OF SAMPL"ffiRS EMPLOYED FOR VISIBILITY STUDY (Cont'd)
Filter Sam-eler No. ~amp_ler SamElJ.n~ Rate (1PM) Size Time (Hr) Sp_ecies Measured
Dasibi Model 1003AH continuous 03 0 3 Analyzer
Teco Model 14B/E continuous NO,N02
Teco Model 43 continuous S02
:samplers 3 and 9 sampled from the same Teflon-lined AIHL cyclone, with combined flow rate 28 Lpm (50% cut-point for unit density spheres, 2.2 µm). 2 lJll pore size Ghia Zefluor filters.
I\) ~NaCl-impregnated Whatman 41 (cellulose) filter. ~ Samplers 1 and 10 sampled from the same Teflon-lined cyclone, with combined flow rate 28 Lpm (50% cut-point _for unit density spheres, 2.2 µrn).
;Prefired Pallflex 2500 QAO quartz filters. Prefired Pallflex 2500 QAST quartz filters.
~xalic acid impregnated glass fiber (prefired Schleicher and Schuell 1981 EPA Grade). ~-Prefired Schleicher and Schuell 1981 EPA Grade glass fiber filters. ~C = elemental C, C = organic C, Ct= total C J e . · oko.4 lJll pore size.
1An all Teflon cyclone with total flow rate 28 Lpm. The collection effieciency of this cyclone for 3.5 J.111 particles was about 80%. Sampled through a 15 cm long, 3.2 cm I.D. polycarbonate tube on to a filter in a Nuclepore open face filter holder.
,-._
-- --C
Sampler No.
I\) \.n C)
2
2
3
3
4
4
5
5
6
6
1
8
9
10
11
Medium
Teflon
NaCl/\rl41
Teflon
NaCl/\rl41
Teflon
NaCl/\rl41
QAO Quartz
Al203
QAO Quartz
Al203
QAST Quartz
Oxalic/glass fiber
Baked glass fiber
Baked glass fiber
Nuclepore
Nuclepore
2500 QAO
Totals:
Maas
108
108
216
Table 4
= S04
-108
108
108
108
18
450
ANALYTICAL
N03 -
108
108
108
108
108
108
108
18
774
STRATEGY FOR VISIBILITY STUDY
b and C + eapLTMNlI4
108
108
108
216 180
bap IPM -
108
108
216
ct
36
36
36
36
216
36
396
Reflectance
e
36
36
108
180
III~ MEASUREMENT OF THE PARTICLE ABSORPTION COEFFICIENT (b )ap
A. The Integrating Plate Method (IPM)
The measurement of b is most commonly done with samples collected ap on an appropriate filter medium using the integrating plate method
( IPM) developed by Lin et al. at the University of Washington
(41,43). The IPM methods is fast, simple, inexpensive, reported to
be relatively insensitive to the geometry of the measuring device,
and has received rather wide use (44,45). In the IPM, a collimated·
1 i g ht be am ( A • 500 nm) is directed onto a filter, normal to the
filter surface, with the particulate matter facing the beam (Figure
~). A piece of opal glass beneath the filter diffuses the
transmitted and forward-scattered light. The resulting isotropic
light flux emitted from the opposite side of' the opal glass is
measured with a photodetector. the transmittance of a loaded
filter, T, is mesured relative to that of a blank filter (set equal
to 100%). The purpose of the neutral density filter will be
discussed later.
Particulate matter on the filter both absorbs and scatters light,
the latter principally in the forward direction. The integrating
plate (i.e. the opal glass) is intended to permit measurement of
the sum of transmitted and scattered light. If effective, the
decrease in transmittance is related only to the absorbance of the
particulate matter. The abso?'J)tion coefficient is then measured
by:
T(%) I -L•b = - = e apToo I
0
Where: I= light intensity measured with loaded filter
I= light intensity measured with blank filter 0
- 26 -
INTEGRATING PLATEFILTER PARTICLE~ I
..ILLUMINATION ..LIGHT .. ►
NEUTRAL DENS ITV .FILTER / SCATTERED LIGHT
Figure 20 Schematic of Integrating Plate Method (IPM).
- 26a -
L ~ the ratio of the air volume sam~led to the
particulate deposit area on the filter
ln I /I0Thus: b ~ in the units (distance)-1
ap L To yield accurate absorption coefficients, it is necessary that the
r incident beam strikes each particle only once, and that light
absorbing particles not be shielded by other particles.
Accordingly, particle loadings< 20 µg/cm2 have been recommended.
Use of a filter with minimal absorbance also should increase the
precision of measuring b • The Lin procedure uses Nucleporeap filters, which provide low absorbance, a refractive index similar
to that of atmospheric -particles, and minimal penetration of
particles into the filter so that multiple absorption is minimized.
(_
The Lin -procedure is subject to error due to multiple reflection
between the up-per surface of the o-pal glass and the Nuclepore
filter. This multiple reflection provides an o~portunity for
multiple absorpt;i.on of light by the particulate matter. · To
minimize this effect, a neutral density (ND) optical filter was
inserted between the Nucle~ore filter and opal glass (43). The ND
filter attenuates the multiply-reflected light.
In support of this technique, Lin reported (41) that a 'Nuclepore
filter loade~ with non-absorbing NaCl aerosol exhibited a
transmittance of 99%, indicating that the light scattered from
these particles still reached the detector. Since no neutral
density filter was employed with this experiment, and the particles
faced away from the light source, this value may have limited
relevance to the IPM as currently done.
A more recent evaluation (45) with a non-absorbing aerosol
(NH4) 2S04, as well as mixtures of absorbing and non-absorbing
materials, also done without an ND filter, revealed a small
- 27 -
apparent absorption for (NH4) 2S04, about 1% of that for an equal
mass of carbon (soot).
For the present study our objectives we re ( 1 ) to eva 1 ua t e the
sensitivity of the technique to wavelength and geometry, (2) to
evaluate and im-prove, if necessary, the precision of the method,
(3) to evaluate factors influencing the accuracy of the method
(e.g. the use of neutral density filter), and (4) to compare
results with those obtained by University of Washington personnel
with the same samples. Appendix C details these evaluations.
The precision of the IPM method with loaded filters was found to be
0.3 to 1.4t coefficient of variation (c.v.). In parallel
atmospheric sampling, the c.v. between samplers was i 1.7'1,. A
comparison with the University of Washington yielded % transmittance (T) values for loaded Nuclepore filters which agreed
within 1 to 2% T.
The accuracy and precision of the method varied with loading. Very
lightly loaded samples gave reduced precision. However, sam-ples
with particle loadings > ca. 20 JJg/cm2 no longer approximate the
requirements for the IPM method. Accordingly, in the present
study, -parallel atmospheric samples were collected with 25 mm and
47 mm filters at the same flow rate (8 Lpm) to obtain particle
loadings per unit area differing by about a factor of five. It was
intended that at least one of the sample pairs would provide an
appropriate sample. In this sampling, the loadings on the 47 mm
filter samples were i 16 µg/cm2 , and, except at San Jose, most of
the 25 mm samples exceeded 20 µg/cm2 loadings.
Figure 3 compares b determinations for the 47 and 25 mm filter ap samples. Samples yielding b values < 0.04 have been excluded. ap The data show good correlation and a ratio of mean values, 47 mm:
2 5 mm filters, of 1.04. Nevertheless, the least squares slol)e and
intercept differ significantly from 1.0 and zero, respectively (p =
O. 99). We consider the b values< 0.1 to be more reliablyap
- 28 -
Figure 3. Comparison of b Measurements at Two Particle Loadingsap
1.6 0
0 1.4
-I 1. 2E ..... I
ts}__. 1.0
~
L QJ
I _µi r-t.,
•r-1 0. 8
/ /
/ /.,,,,,I 4- 0 ·/
E E 0. 6
0 0 . // / / /~ ~ / //..._, Y = -0.046 + 1.21X
/
/// .. ~ ~c;
/ / ~<:,UY
./ / 'i,~~t-,.'v
/
/ /
/ /
~ /
/ /
0. 4 8-_!)
(fj0. 2 0
0.0L,Q?00°0
0
0
0
/~
0. 0
·/00~ / .
~,,,,6 0 0 r - 0.956 0
1 I0. 2 I I I I I0. 4 0.6 0. 8 1. 0 1. 2 1. 4
bap (25 mm filter) 10--1m-1
measured by the 25 mm filter. Values > 0.6 are considered to be
more reliably measured with the 47 mm filter.
B. A Laser Transmission Method (LTM)
Rosen et al., at the I.a.wrence Berekely Laboratory, have devised a
version of the IPM which measured the transmission of 1 i g ht from a
He-Ne laser (632.a nm) passing through atmospheric particulate-loaded
filters (42). In this system the integrating plate is eliminated and
the filter, itself, is used to provide uniform intensity to all
forward-scattered light. In addition, a large lens immediately
behind the sample filter is used to focus the light onto the
detector, increasing the sensitivity of the measurement and possibly
decreasing errors due to light scattering. The decreased light
intensity reaching a detector relative to that with a blank filter,
is assumed to be due to light absorption by the particles.
The reported advantages of the LTM compared to the IPM, are its
applicability to more heavily loaded samples, and the suitability of
filter media other than Nuclepore, including quartz fiber. The
latter would also permit combustion analysis for total, organic and
elemental carbon. Accordingly, this technique was included in the
present study.
Figure 4 compares University of Washington IPM and LBL LTM results
for co-collected samples (46). The LTM method was used with both
Millipore cellulose ester membrane fil tars and prefired Pall flex
quartz fil tars. The LTM method showed reasonable agreement for the
two filter types but differed by, on average, a factor of 2. 5 from
the IPM. R. Weiss (47) believes that the factor of 2.5 higher
results by LTM is due to multiple absorption because particles
penetrate substantially beneath the cellulose ester and quartz filter
surfaces. Within the filter matrix, each particle could intercept
the incident light more than once as the light is scattered by
surfaces. Since the IPM has been reported to measure b with an ap
- 29 -
,{
f I,
----
X
X
.xf~
X
~ NUCLEPORE ·1. vs
__....,-I NUCLEPORE vsE ,o-4
MILLI PORE
__...., uJ a: 0 a.. IJJ 10-5
•_J ~ I u ::> z X ••x
x•
•
QUARTZ .._...,
C.
bo 10-6
:e: X 0--
Figure 4 Comparison of the Integrating Plate Method and Laser Transmission Techniques for AbsorptionCoefficient Measurement (x =Millipore,•= Quartz)
- 29a -
error of ~ 15%, (43) the b measurements by the LTM would requireap
substantial correction. However the LTM was also compared to a
photoacoustic method and showed almost perfect agreement (45). When
compared to a similar photoacoustic method, the IPM gave higher
results by a factor of 1.3. Given these inconsistencies, the factor
of two uncertainty ·in absorption coefficient measurements estima tad
elsewhere (45, page 28) seems prudent.
Figure 5 is a diagram of an LTM apparatus as constructed at AIHL.
The unit differs from the LBL apparatus principally in the detector,
a photodiode/op amp in place of a photomultiplier tube. Our initial
evaluation of this technique considered: (1) Linearity of the
detector response with change in light intensity, (2) precision, (3)
effect of filter medium, (4) effect of particulate orientation
( toward or away from the light source), (5) comparison with the IPM
for b measurement, (6) the influence of wavelength, and (7)ap
comparison of LTM results by AIHL and LBL.
As detailed in Appendix C, linearity was excellent throughout the
range relevant to loaded quartz filter samples. The precision of b ap
measurements was 2 and 4-4% (c.v.) for Millipore and quartz filters,
respectively. Analyses of the same filter samples by AIHL and LBL
showed high correlation (r = 0.9999).
c. Comparison of b Determinations by The IPM and LTM a
Figure 6 compares results by the two methods for the atmospheric
samples collected at the three sampling sites in the current field
study. For this comparison, IPM results with 25 mm and 47 mm
parallel fil tar samples were averaged, except where one or both b _ 4 _ l ap
values was < 0.04 X 10 m • For such low absorption coefficients,
the higher face velocity results (i.e., those for 25 mm filters) were
considered mo re accurate. The results are highly correlatad. The
ratio of mean values LTM:IPM is 2.72. Since the preponderance of
current data suggests the IPM to be more accurate for the measurement
- 30 -
f.Sample
D I D He-Ne Laser Lens
A Lens
B
Detector
Digital Voltmeter
~PVC Light-Tight Box Slide mechanism to insert and withdraw sample supported in rigid frame
Laser: 632.8nm, O.Smw, Spectraphysics Model 155, random polarizationG
Lens A: 12.4mm diameter, 14.3mm focal length, plane-convex lens to expand beam to a 1.2cm diameter disc on the filter sample.
Lens B: 60mm diameter, 39mm focal length, aspheric lens, Rolyn optics.
Detector: EG & G Model HUV-1 OOOB silicon photovoltaic detector/operational amplifier combination, positioned at focal point of lens B.
DVM: Fluke Model 8000A digital multimeter.
Base: 1/2" flat steel plate optical bench permitting use of clamps with magnetic bases.
Enclosure: 1/2" PVC sheet.
Figure5. Schematic Diagram of Laser Transmission Method, Viewed From Above.
(not to scale)
- 30a -
Figure 6. Comparison of b Measurements by the IPM and LTM Methods ap
- . .--------•••-'•·- csu,-•-·•-•-••~-~- • -
1~5
0
I um ... I s
1 ts;)
I ...... .75 t ..tr . I ,,...,.
X n.. '-' ~
.50t ...0
.25
BAPIPILTMSYNBOLa. LUETYPE11
~!t)c
a--5. 2116 •.al. 5138 bee•.«« -.-=S.0176 -,..-e. 4528 · r-=0.9Z17
0L· L.•---- =------L-.~---~-- L----..-~~.l--."·-·---•-......~ 0 .50 1.80 1,50 ~ ~50 aae J_ ~c-~.
b•(LTM) 10-• m-1
-•-~ ...:::::=:;;£.:..z;;;;:::__.. ~l l~P ........a.._. WI ~....._, L.l.....~ ..::..:::J,i,.a.::...-::::.....i.-- rm--
of b , the LTM was judged potentially suitable only for elemental ap
carbon measurement.
f
(
(
- 31 -
IV. DETERMINATION OF ELEMENTAL CARBON (C)e
A. Introduction
Elemental or black carbon is the major contributor to the particle
absorption coefficient, b However, discrimination between ap organic and elemental carbon (C ) is difficult. A selective
e combustion method (SCM) and a reflectance method (RM) were
previously used for this purpose (34). The former gave results
which were believed to be in error by ca. 20%. The RM yielded
results which, with 24-hr hi-vol filter samples, were too low by a
factor of two. While more accurate, the SCM is slow and relatively
difficult. Accordingly, an effort was made to improve the
calibration of the RM. In addition, the laser transmission method
(LTM) was calibrated for use in measuring C. e
As detailed in Appendix D, for the current study the LTM as well as
the RM were calibrated with butane flame-derived soot of known
particle size distribut1.on prior to collection. In addition,
calibration employed samples supplied by LBL including diesel soot,
atmospheric particulates and vehicle tunnel air particulate samples.
The LBL samples were analyzed for elemental carbon by a thermal
analysis technique correcting for decomposition (60). The RM
calibration appeared to be relatively insenai tive to the source of
the carbon standard whereas the LTM calibration showed great
variation.
B. The Ia.ser Transmission Method for C (42).e
LBL has used the LTM technique to measure black or elemental carbon.
c = [c ] + [c ]total black organic
Thus by measuring total C and black C, organic C can be obtained by
difference (correcting for carbonate C if necessary). If black
- 32 -
carbon is assumed to be proportional to the absorbance of a loaded
filter:
1 =- - [Attn]
K
Where: Attn=- Attenuation=- -100 ln (I/I) 0
K =- a constant
The attenuation per unit mass of Ctotal' or specific
attenuation, a, is defined by:
K[cblack]
[ctotal]
Employing laboratory-generated and ambient air elemental carbon
samples, the LBL group determined a mean value for K of 20 2 .
cm /µg when deposited on a Millipore filter. Thus:
Attn=--20
c. Intermethod and Interlaboratory Comparisons
Calibrations results are summarized in Table 5. For ~ase in
iden tif'ica tion, the calibrations were coded as shown. Elemental
carbon values obtained with these calibration for ten fine
particulate atmospheric samples are compared to the corresponding
total carbon values in Table 6. Samples identified with the letters
J - Q were collected in Riverside and the remaining, in downtown Los
Angeles.
Comparing ratios of mean values, elemental carbon calculated from
LTM-1 (i.e. the LTM using the aged butane flame carbon) yielded
- 33 -
~
TABLE 5
CALIBRATIONS OF OPl'ICAL METHODS FOR ELEMENTAL CARBON
Code for Code for Carbon Reflectance for RM Laser Transmissionb LTM Source Method (RM)a Calibration Method (LTM) Calibration
Butane :Flame a= 0.0309, b = 1.68 a = 5. 59, b = 6 • 88 LTM-1 ( r = 0. 957) (r = O. 992)
LBL Carbon Samples a= 0.0366, b = 1.62 a= 11.1, b = 23.1 LTM-2 ( r = 0. 99) (r = 0.964)
LBL Samples and a= 0.0337, b = 1.65 RM-4 w Butane Flame (r = 0. 974 l,J
PJ
a. Slope, inte2ept and.correlation coefficient for the equation logy= log a+ blog x, where: y = (100-R) /200R and x = elemental or black carbon, µg/cm2 • R =%Reflectance.
b. Slope, intercept and correlation coefficient for the equation -100 ln (I/I)= a+ b (c ), where I and I are light intensity though loaded and clean filters, respect~vely, and C eis elemental or b~ack carbon, µg/cm2. 8
TABLE 6
COMPARISON OF TOTAL CARBON AND ELEMENTAL CARBON (C)e
VALUES (µg/m3) a
(
f
Sample ID
J11 59Q 11
M1176Q011
01183Q011
01185Q011
P1186Q011
Q1192Q011
T1202Q011
U1209Q011
W1214Q011
X1221Q011
Total C
25.15
18.03
13.89
12.0
16. 50
13.ao
17.02
4.46
36.16
13. 54
C e (LTM-1)
28.55
6.62
12. 37
11 • 51
22.88
9.24
25-6
7. rs,
59.34
10.03
C (LTM-2)e
5.gr
1 • 61
2.73
2. 56
4.,76
2.12
5.30
1.79
11 .83
3.99
C (RM-4)e
5.41
1.40
2.44
2.. 12
4.35
1.83
3.98
1. 01
11.00
3. 01
bMean C
e
5-64 ± .33
1. 51 ± • 15
2. 59 ± • 21
2.34 ± • 31
4. 56 ± .29
1.~ ± • 21
4.64 ± • 93
1. 40 ± • 55
11 • 42 ± .59
3. 50. ± .69
Ratio of Means : 1. 00 1. 13 0.25 o. 21 0.23
a.
b.
The terms LTM-1 , LTM-2 and RM-4 are defined
Mean of resul ta by LTM-2 and RM-4 methods.
in Table 5.
- 33b -
values which averaged 13% higher than the total carbon. Thus this
calibration was rejected as inappropriate. The remaining techniques
yielded C values averaging 21 to 25% of the total C. This compares e
to an average of 20% reported elsewhere by similar optical
techniques (59). The results by the laser method (LTM-2) were
higher by, on average, 19%, compared to those by reflectance.
Figure 7 is a scatter diagram of Ce determinations by LTM-2 against
those by RM-4 for about 100 fine particulate samples from the
current program. Symbols A,B,C,D indicate 1,2,3,4 data points,
respectively, at the same locations. The results are highly
correlated, show a negligible intercept, and average 15.5% higher by
the LTM.
Eight respirable particle samples which, by our optical methods,
ranged in C loading from 1 to 8 µg/ cm 2 , were analyzed by the e
General Motors Research Laboratory. In the GM technique (49) filter
samples are dropped into a furnace at 950• in helium. The evolved
pyrolysis products are subsequently converted to CO 2 and analyzed by
a non-dispersive infra-red method to yield "apparent organic
carbon". Oxygen is then introduced into the furnace section
containing the sample and the resulting CO 2 is used to measure
"apparent elemental carbon". The method is believed to be subject
to error from charring during analysis which would yield elevated CE
values.
Table 7 lists the AIHL and GM results. The AIHL values show
C obtained by the LTM, RM and mean values by those techniques. In e
addition to C results, GM's values for organic carbon are also e
given. The optical methods were highly correlated with the GM
values (r ~ 0.95) but yielded results which were about 25% lower.
Given the uncertainty in defining organic and elemental carbon in
atmospheric samples, none of these methods can as yet be considered
to be more nearly "correct".
-----------
---------
Figure 7. Elemental Carbon by the LTM vs. Reflectance Method
/ A11 +
__ / A
10 + /
-/ A
/9 +
/ I
- I - -- ------- ---- - -- - ---- ---- ----- ------ --/a +
I --- - --- - ------ ---------'--------- - --- - -- -- ---- -I - - ---· ~~,L_
7 + ~ -- - -- - -a->'#'~ ----- --- . RM-4 / _,,.i,~7-- -- -- ----
C (µg/m1e
6 +
I w I•-... IU 5 +
---------------
__
-- / _-
/A-- --✓- --- A
- L ------- / - B
/_A------7 ·-= A
_____ _x_ = r =
-0.075 + p~~'!~~'--------4 + 0.986
- 3 + __ _
I' - -- - -2... - --
1 I -~AA:1 + ~E'3~
I("_· BA
~
_Q_ ~ -+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+---------+--· o 1 2 3 4 5 6 7 e 9 10 11 12
LTM-2 C (µg/m 3)e
TABLE 7
COMPARISON OF ELEMENTAL CARBON MEASUREMENTS BY GENERAL MOTORS' COMBUSTION AND A.DIL OPrICAL METHODS (l,lg/m2 )
C C e 0
Sample LTMa RMb
J1161Q011 2. 92 2.49
L1169Q011 1. 07 0.94
R1194Q011 2.67 1.95
R1196Q011 1. 20 0.82
V1210Q011 s.35 6.55
W1217Q011 3.32 2 .. 63
X1218Q011 8.01 7 .16
X121 9Q011 5.05 3.96
Blank 0 0
Regression F4uations: Mean C =- -0.141 e
LM C == -0 .. 36 e
RM C = -0.375 e
a. Laser transmission method. b. Reflectance method. c. Mean results by the LTM and RM.
C cMean e
2 .. 70 ± o.o;
1. 01 ± 0.09
2 .. 31 ± o. 51
1. 01 ± 0.27
7. 45 ± 1. 27
2.gr ± 0.49
7. 59 ± o. 60
4. 51 ± 0.77
0
+ 0.748 (GM C)e
+ 0. 845 (GM C )e
+ 0.702 (GM C)e
GM
;. 9 ± 0
2.4 ± 0.1
3. 5 ± o. 1
2.3 ± 0.3
11. 0 ± 0.4
;.2 ± 0.1
8. 6 ± 0.1
7 .1 ± o. 3
< 0.1
r = 0. 956
r=0.969
r = 0. 947
GM
12.4 ± 0.4
9.1 ± 0.3
9. 5 ± 0.5
7.6 ± 0.4
18. 1 ± 0.8
8.7 ± 0.4
15. 2 ± 0.8
11.s ± 1 • 5
1.8 ± o.a
- 34b -
D. Comparison of C by Optical Methods with Particle Absorptione
Coefficient Results
Since C is the nearly exclusive source of light absorption in e
suspended atmospheric particles, and since most of the C is in the e
fine particle fraction, a linear correlation is expected between
fine particle concentration C and the light absorption coefficient a
(b ) • For example, Figure 8 shows General Motors' results for ap Winter Denver aerosol (6). The mean slope was 11.8 m2/g (0.118
_4 x:10 m2/µg) and correlation coefficient, 0.93. The bap values for
{ the Denver study were obtained by the integrating plate method
( IPM), (but without a neutral density fil tar as employed in the ... present study) and Ce, by General Motors' controlled combustion
technique ( 49). The latter yielded results equal, within
( experimental variability, to the interlaboratory mean values for C a
and organic carbon C by combustion methods with 24-hr hi-vol 0
samples, in the GM Round Robin comparison.
( Figures 9 - 11 are scatter diagrams for about 100 atmospheric
samples from the present study plotting, C (LTM-2), C (RM-4) and a a
mean fine particle C a
values against means b ap values by the IPM.
The intercepts are negligible in all cases, and the mean slopes are
( 11 • 1 , 12. 9 and 12. 0 m2/ g, res pa c ti v e 1 y • Thus the r a s u 1 t s a r a
similar to those by GM in Denver, implying reasonable agreement
between our optical methods and GM's selective combustion technique.
These values for atmospheric C may be compared to the absorption( e
coefficient per unit mass ·for soot. Heitzenberg reports a value of
*Better accuracy is expected for the IPM with N.D. filter. The GM b values( ap
are inferred to be 5-10% too high (50).
- 35 -
0
I -....
•J II.
•-.. 0.• -Iii
.,e o c.S
·-l ..,o.. •-C!-
•1ft d
I
••
•
• X
• •
•
11 11 a ,p 1D •PLne EL...,,iaL Carbon 1J9I•
Figure 8. Relation between Absorption Coefficient and C for eDenver Winter Aerosol (Ref. 6)
- 35a -
----- -- --------
---
Figure 99 Elemental Carbon -by LTM vs. Absorption Coefficients
___ 1 .. 4 _+ _______ ------ --- ----------- ---------- ---------- -----------~·- ------ - --------•-·--··--·--A
1. 3 + . _ ------- ----------------
_ 1. 2 -t- -- --- - --· -1
A .l.l_-1-__________________________
I I
_l. Q __-t____ ---------------- ·------- -- ------------------
-- 0. 9 --+
I bap _Q.8_:t- ________________ ----------- ------- - -------·- -------------- ----- ____________________,__________________
---------------------
----------
4 1 (10 m-) 0.7 - -------------------------- _____,., ---- --- -
------- -------~AA _A_________ ------
A A
. A - . A ____ A_B ___A_____________ ------
A A
A
__ o. 6 + ________________ _
_ O•. !:L-t-_ _____________ y = -0 o 053_7 +__ 0 o lllx_ r = 0.925
__ 0.4_t___ ------------------------A
_o. 3_+ A
A ~ACC0.2 + _ AACC AA
BBB A A BA A A _.7n AB
0. 1 + B ,,A'd'AA - A . . . I A A 1AA A AAA l A~.(c AA
0 .. 0 -+~~-~-B_~__ A_ __ ----~---------_ -----~ ______________ - +-- _______________ +-------_-- ______ ~- _-------- -----:;_-------·-;-···- -------------:;:- _____________ ~- _________ -+- _____ ... __ +. I w U'l o 1 2 3 4 s 6 7 a 9 10 11 12 t:1
LTM-2 C . (µg/m 3 )e
- ------- - ------ ---
,=
b ap
_4 _l (10 m )
w
_,........,_ ,,-,, ,....,,""" r"- - ..."..
Figure 10 Elemental Carbon by the RM vs. Absorption Coefficients
1. 4 + : A I
-1 ~::i-- ± ---- ------------------------ --------
--1. 2 + --------
1.1- +_ ____________________ AI ----·---- -- -- -- - ----------
LO+
0.9 +
o.a + I. I
__ Q. 7. -t. -- - ------·--. ----- ----------~-------------- -
I A
--v. 6- ---------------- -----;JC- ----------- ·----- - . - ---··· -- -- ---- ------
1 _Q_!L~ --------·--------- .. -Y =-o .037__ ,t __Q!..~ ~~-
A ,;/'A A A r = 0.917
C. 4 +__ -- -·--·------------A ----- -------A ;/-A,______ A _____________ ---
A _Q •. 3 ± ___ ------ __________A___L__A__ ----
1 BA~ A A A AA1,,AAAB AA
0. 2 + --------- --------- ------~ACD____A_______________ ------~- _____________
AC~ A BA A A
0. l + AAy("B :·A____ A
I ~BA BAA
a. a !L~~;::..~ _ .. ---+-- _ -+ _ _ _ _ ·+ _ ___ :. _ _ :.:. _ __ + _ _ -+-- .. . __ + --------+ ____ 0 lJl
0 1 2 3 4 5 6 7 8 9 10 11
RM-4 C (µg/m 3) e
Figure 11. Mean Elemental Carbon by LTM and RM vs. Absorption Coefficients
1. 4 + --- ---------- --·-------~------- --·----- -----A
..L--1,_±.__________
L-2-~----
A -1~1-±-
1 I
-1. 0 -t.______ -- ----- --- ---------
: a
0_9_
b .0.8 + --------- ·-------- - ---- ---- ---ap
_4 -1 : ( 10 m ) Q_7 .+____ A
0. 6 +__ -----
_Q_ 5 + -- --- -- -- .. ---------------- - --------· AA A--1('~-------- ----------
y =-0.0483 + 0.120x0. 4 +. .. - - --- A - ·-- - -- ··- -· . A -·-------- --\0" r = 0.924
/2 A A _Q.3 + ----- - ----·-------- ----
B A ~
A A C.,,J:_A A A A 0.2 + . --~---EC __ .A.- ---------·- -·-· .. ----------------------
,,,,KAA B AA A Atl.A A0. 1+ pJB~A A.
I A Bt A BA I _),AB AA
w 0.0 ++~-~AAA:--~-----+--------+---------+----- __ -~-- ------+-~ ----+--- -----♦---------♦ ---------+---------+-----U1 p., 0 1 2 3 4 5 b 7 8 9 10
Mean C (µg/m 3)e
11
These values for atmospheric C may be compared to the absorptione
coefficient per unit mass for soot. Heitzenberg reports a value of
9.7 m2 /g ( 13) but notes that in mixtures with non-absorbing
material, the value increases. Clarke and Waggoner ( 50) report a
mean value of 10. 1 ± 2 .1 m2 /g ( n = 4) for soot collected on
Nuclepore filters.
Previous studies showed measurable absorption by coarse as well as
fine particle C. Employing mean C values and multiple regression·· e _ e analysis for 104, four-hour periods yields the equation (in units
- L+ 110 m- ):
b - - 0.04a ± 0.15 + 0.116 ± o.ooa (c )f. + o. 02 ± 0.029 (c)ap e ine e coarse r • 0.925.
Thus the absorption efficiency for coarse C , per unit mass e
concentrations, is not significantly different from zero.
Furthermore, the addition of coarse C does not significantlye
improve the correlation with b (r = 0.924, Figure 11). Hqwever,ap
as will be shown in Section VIII A, fine and coarse elemental carbon
show a relatively strong linear correlation ( r = 0.81). As a
result, the standard deviations for the coefficients in the multiple
regression with b are increased. Thus coarse elemental carbon mayap
have a relatively small but real contribution to light absorption.
- 36 -
V. MEASUREMENT OF THE SCATTERING COEFFICIENT FOR LIQUID WATER
A. Introduction
Water is a significant constituent of suspended particulate matter
and contributes to light extinction by light scattering. To
estimate the contribution of such scattering, two integrating
nephelometers were operated side-by-side (Appendix B) with the
incoming sample of one unit heated 16 + 3° C above ambient,
following the recommendation of A. Waggoner, to dehydrate the
aero so 1 • The s cat t e ring co e ff i c i en t f o r watar, b , can beSW
calculated as the difference (b ) bi t- (b )h t d. In spam en sp ea e principle, this technique is subject to error since other aerosol
constituents, including organic compounds, NH4No 3 and halide salts,
might vo lat i 1 i z e as we 11. However, it should provide an upper
limit for the scattering coefficient for particle-bound water.
B. Method Evaluation and Correlation with Aerosol Constitutents
The water content of atmospheric aerosols should vacy with relative
humidity as well as the aerosol composition, since both sulfate and
nitrate aerosols are hygroscopic. Table 8 lists Pearson
correlation coefficients between b , pollutant and meteorologicalSW
parameters as well as visibility parameters as measured in the
current study. The highest correlations (0.6 - 0.7) were found
between b , fine sulfate, ammonium, fine mass, and b , the SW ag
absorption coefficient f_or NO 2 • If volatilization of organic
compounds was an important contributor to b , as measured by the SW
difference method, then a higher correlation between b and C SW 0
would be expected compared to the 0.32 observed.
To assess the combined influence of RH and aerosol constitutents
b data were fit to equations of the form: SW
- 37 -
Table 8
PEARSON CORRELATION COEFFICIENTS BETWEEN b ,SW
POLLUTANT AND METEOROLOGICAL PARAME'rERSa
Parameterb r
= FS04 0.74
= CS04 o. 54
,f
FN03 - 0.48
CN03 - 0.4;
HN03 0.. 29
NH4 + 0.. 7;
0.05CNH3
FC 0.27 e
cc o. ;1e
C 0.. 32 0
F Mass 0.64
03 0.19
RH 0.41
T - 0.06
b 0.67 ag
b o. 58 sp
b o.. 18 ap
a .. b .. c.
n = 107 - 109 F = fine , C = coarse Not significantly different from zero
- 37a -
b SW
= a + + ( 1 )
and
(2)
Where x and y are constants between 0 and 2, andµ=% RH/100.
The results shown in Table 9 indicate a general improvement in
correlation relative to linear regression with two parameters.
However employing equation 1, the addition of a relative humidity
factor shows only a small improvement in the correlation with
little change for x • 0.5 to 1 .o. Similarly, using the other
equations shown in table 9, there is relatively little change for x 2
or y =- 0.67 to 2.0. The range in the variance, r, with equations
including humidity dependence was only 0.6; to 0.66. Based on the 2
criterion of maximizing r, clearly the choice of a specific model
from the six listed is not crucial. For simplicity we have
employed equation 9-2 with y =- 1.0, although most recent studies
have used equation (1).
2 2 The correlation obtained here (r =- 0.66) compares to a value r =
2o.a7 observed in Denver with equation 9-1, x = 1 and r = 0.17 with
the same equation in the eastern U.S. (51). Poor correlation was
suggested to be due to the lack of speciation of sulfates since
H2so 4, for example, is more hygroscopic than (NH 4) 2 S0 4•
- 38 -
Table 9
MULTIPLE REGRESSION ANALYSIS BETW~EN b AND AEROSOL CONSTITUENTSSW
E9.uationa
X _J_ a b b C r r2
9-1 o.o - 0.029 0.046 00023 0.76 0.58 0.5 - 0.039 0.031 0.019 0.79 0.63 0.7 - 0.034 0.026 00018 0.80 0.64 1.0 - 0.015 0.019 0.015 0.81 0.65
9-2 0.67 - 0.031 0.069 0.047 0.81 0.66 1.0 - 0.012 0.081 0.062 0.81 0.66 2.0 0.083 0.125 0.115 0.78 0.61
9-3 0.5 - 0.039 0.044 0.040 0.78 0.60 1.0 - 0.03a 0.042 0.063 0.79 0.62 2.0 - 0.011 0.040 0 .. 125 0.79 0.63
9-4 0.5 - 0.039 0.044 0.019 0.77 0.,60 1.0 - 0.04 0.042 0.015 0.78 0.61 2.0 - 0.017 0.040 0.0073 0.79 0.62
9-5 0.5 - 0.0;4 0.066 0.024 0.79 0.62 1.o - 0.030 0.090 0.026 0.80 0.65 2.0 - 0.010 0.152 0.029 0.81 0.66
9-6 0.5 - 0.037 0.033 0.024 -0.79 0.62 1.o - 0.032 0.021 0.026 0.80 0.64 2.0 0.0037 0.0076 0.031 0.79 0.62
+ c(N03)aEquation 9- 1 : b = a + b~S0 4 ' SW
(1 - µ)x ( 1 - µ) X
Equation 9- 2: b =a+ b (S0 4 ) µY + c (NO 3) µYSW
Equation 9-3: b =a+ b (S04) + c (N03)µYSW
a + b (S04) + c (N03)::::zEquation 9- 4: b
SW ( 1 µ)x
a + b (S0 4 ) µY + c (NO 3)Equation 9- 5: b = SW
a + bS0 4Equation 9- 6: b = + c (N03)SW
(1 - µ)x
bNone are significantly different from zero at~= 0.95e
- 38a -
VI. ASSESSMENT OF SAMPLING ARTIFACTS
A. Artifact Particulate Sulfate on Glass Fiber Filters
Glass fiber bi-vol filters (Schleicher and Schuell, 1981 EPA Grade.
fired at 450°C) were used to collect samples for analysis of
carbonaceous materials. However, these were also analyzed for
sulfate and nitrate to assess what difference in sampling artifacts
resulted from the firing at 450°C relative to the filters as
received. In addition, glass fiber filters were assessed for HN0 3
retention with varying doses.
Figure 12 is a scatter diagram of blank-corrected sulfate results on
4-hour hi-vol glass fiber filter samples against those on Teflon
filters obtained in an open face 47 mm filter holder (i.e., without
deliberate particle size segregation). Riverside and Los Angeles
data were fit with a single line, and the San Jose data, with a
second line of markedly higher slope. Results on glass fiber are
significantly higher. The average difference (glass fiber)f
(Teflon), was 4.9 µg/m3 • Thus, although the sulfate artifact is
smaller than expected for short terms samples (52) the positive
artifact clearly persists.
Bo Artifact Particulate Nitrate on Glass Fiber Filters
Previous studies with 4-hour hi-vol sampling have indicated that at
least for short term sampling, glass fiber filters serve as total
inorganic nitrate (TIN) collectors, retaining both particulate and
gaseous nitric acid (21). The present study repeated this
comparison at San Jose, Riverside and Los Angeles and, in addition,
evaluated 16 to 24-hour bi-vol filters (prefired S & S glass fiber
filters) as TIN collectors. For this purpose, N0 3- on glass fiber
filters was compared to the sum of the nitrate retained by Teflon
filter and a NaCl-impregnated filter sampling in tandem.
- 39 -
Figure 12. Artifact Sulfate ·on Glass fiber Filters, 4 Hr. Hi-Vol
r·· ---- -----,,.,/
40~. / /_," ✓i + SAN JOSE
0. .,/
/
/
l O RIVERSIDE / /,.
,.., ✓
35~ l
o LOS f.NGELES / ," i
(fl .,/' / E ! _.,/ /
'~ ,/ ,.._ ,
// ~...,.,,m 30 i-' :) //o ,..~v./
" ~"Y.,◊ / .ci..V'
,,........ ◊
;~--✓ -
0 ~-0~) /'
0 0:: 2s l-UJ 1 ~6 0 ~g,,
cP / /'m I _,'6w 1---j I 0/ /~ LL ., /
20 L. ,p·/ /.,.0 U)
¢◊◊ /. 0 ./(f) <t:: !
_J I $ ~~~ / ◊ /9'bo /~ 15~
'-"" 0~~800 /o·~C tf ,/
+ ~~i /10 I - ✓~-. 0 ◊ /0""""' (f) ! ,~-- /
l_,,~"" /', i· .,r +./~ ,5 t-;;t_.,, /
: ~ /
i *+ ~;1:i ~
✓
0 L --- -- .. j__________J___ ·- _--.....L .. -- .. ____L . ··---·--- .L ~,. _. -1... --- ·-· -- 1.______L 0 5 10 15 20 25 30 35 40
so_.2- (TEFLON) ug/m3
Samples
SAN JOSE SYMBOL=+ LINETYPE= · · · · · · · · · · · · y=a+b*xt1=36 a=l.9337 e
11;:0. 9219
b=l. 7590 s 11=0. 3902 s.,..=1. 6798 r=0. 6116
RIVERSIDE &LOS ANGELES SYMBOL=o LINETYPE= -----------y=a+b*xn~7~ a=6.5235 s.=3. 4016 b=0.9326 s.,=0. 0297 e-,.x=l. 6570 r=0.9673
Figure 13 compares 4-hour results for the three sites. The results
are highly correlated, with a ratio of means (glass fiber N0 3-)/TIN
of O. 997. Thus these data also support glass fiber filters as TIN
samplers.
The 16 to 24-hour hi-vol filter nitrate data are compared to the
corresponding TIN values calculated from 4-hour samples in Figures
14-16. Only at San Jose, which exhibited low nitrate levels (TIN i 32 1,1g/cm2) did the glass fiber fil tars continue to approximate TIN
samplers (ratio of mean, glass fiber/TIN• 1.03). At Riverside and
Los Angeles, an average of 24 to 42% of the nitrate (presumably
HN03) penetrated the filters. The difference in results with
location probably results from the much higher nitrate dose in the
Los Angeles basin(~ 147 µg/cm2 TIN).
c. Loss of Particles in the Denuder of the Particulate Nitrate (PM)
Sampler
Previous studies have suggested that coarse particle nitrate may be
1 ost in transit through the diffusion denuder employed to remove
HN03 from particulate nitrate leading to negative errors in
measuring PN. To minimize such errors the present study measured
fine particle nitrate, excluding coarse nitrate and other aerosols
from the denuder with a cyclone. Loss of < 2. 5 µm nitrate was not
expected to be significant. However, if it occurred, it would
produce a· negative error in the fine particulate nitrate
measurements.
To measure the loss of fine particle constituents in the denuder, it
is necessary to compare concentrations of a stable species collected
on a filter with and without a preceding denuder. Sulfate collected
on Teflon filters was used for this purpose. Figure 17 compares
results for sampler 3 (with denuder) to those for sampler 1 (no
denuder), and indicates very high correlation of results and no
significant loss of sulfate in the denuder. Although sulfate
- 40 -
r~
E " "'-m J
'-"
Cl)
0 z a:::I~w
PJ Q.
CD .,.._..I
LL
(J) (J)
< _J ~
FIGURE13 Nitric Acid Retention on Glass Fiber Hi-Vol Filters, (4 Hr.
60 r------ss r / 0
s+ V' / 0
+ s. J.. (),
◊ RV45~ 0 LA. i ) ,0. / 0
40l -✓I .A_ 0
35
030
2l I
""-.20
15' l
r-ml
I
r-
slit◊ LJ0L...- I I I I 1----L-___ L.___ I -- l l l __ L__ 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
TOTAL N03 (u9/m3)
Samples)
NITRATE SYMBOL=·LlNETYPEr: y=a+b*xn=104 a=t.3808 e.=0. 5152 b=0.9349 sa.=0. 0190 s,_,.=3. 0401 r=0.9796
,t
l
r- Y = 1.23 + 0.855 X
~ r = 0.939
+I-
..
-~
~
1. +
...
...
+
Figure 14. Nitric Acid Retention on Glass Fiber Hi-Vol Filters (San Jose, 16-24 hr Samples}
~
!llr-.. I
10 1
I
~ e 9!-
I
i' tn ..._,:)
8 i-I I
C "" I
z 7~ l
0:: w co 6~ t--t I LL
5~(..f) I (Jj
< --1 4l..... L? l
l I~--L..
~ (\J
I 2~cc __.
lr I
I...0L_ I
0 1
i 1 I
!
I I I I ....L.. t I I I -L..-t~...-1.......1-._.L__.J__ 1 g2 3 4 5 6 7 8 10 11
16-24 HR MEAN TOTr,L N03 (ug/m3)
- 40b -
I
12
t:;' E
' O') :) ~
{
0 "' z a:= LU CD ....... LL.
CJ1 en < _J 0
0::: ::r:: ..:t N I
(0 ......
Figure 15. Nitric Acid Retention on Glass Fiber Hi-Vol Filters (Riverside, 16-24 hr Samples)
30 I -----·-------·-•-····-·--··-··---·--1
I
29·I ' I r ir ;
' 281r i
27·t
r 25~
I
i 25~
;
l 24t-
I i
2sL l t ' 22L. I
2'i ' I
j I I
·r I I20L.. ! I
!1aL j
Y = 3.47 + 0.646 X r = 0.881
!
I ! \
! i
18 ~- .L ___ ____J_J t! ..L.,.___.,__,_.____ .J~-L I 1-.L-l----.il..--._l L._J___J__ J I
20 21 22 23 24 2s 2s 21 2a 29 a0 s1 32 33 34 ss 36 37 38
16-24 HR MEAN TOTAL N0 3 (u9/m~
- 40c -
~ E
' en ::)
'-'
('l't
0 z ~ w CD ........ LL (/) (/) <C: __,' t:)
er: :r: ~ ('\J
I co .-t
Figure 16. Nitric Acid Retention on Glass Fiber Hi-Vol Filters (Los Angeles, 16-24 hr Samples)
I 30 ·---, ~
fI Ii·"9
i t Y = 9.87 + 0.27 X ~ l i· I r = 0.909 t-
ii l
! I""
ol 25L
i
.. ~
I i
l
' l r • I r I I
020 r-1 o. I
r I
0 I
i..
L I I
i i +-l
15 r-1 I
0l I 1. I I 1 LJ.....1.-l- I I I I I I I LJ-a I i I -.. I I I I 20 25 30 35 ' J 40 45 ' 50 55
16-24 HR MEAN TOTAL N03 (ug/m3)
- 40d -
~
Figure 17. Loss of Non-Volatile Particles in the Denuder for Nitrate Sampling
(W)
E
' m J
~ I (Y)
i> ~
n:: I LL.I
___j
(L ::E < (f) '-../
...,. 0 (ID
40 ~ ·--··-· ~
3BL L
361-.....
34 ,-. ~
32~ t
30r +-
28 L 26;_
24l ;
22~ I
1-
20 ~ r
18 r-....
16~
14~ +-
12 f~
10~ l-
8 ~ r--
6 ~ ,- i-
4 ;_
+ s. J. ◊ RV o L.A.
0
0
0
,-..
! j
0 ~:.LL_LLLL LLJiLI.J..J ...LLJ.-1...L1._l_.1....L.J. l l I 1-1. J.i.J_Lt __ l._LLJ 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40
S04 (SAMPLER 01) ug/m3
SULFATE SYMBOL==· LINETYPE== -y==a+b*xn=108 a=-0.2214 s.==0. 1302 b==0.9910 s1.::;•t 0139 s,..•==0. 9030 r==0.9897
particles are expected to be somewhat smaller than nitrate (9) and,
therefore, not the ideal surrogate to estimate N0 3- loss, these
results do not support a significant error in particulate nitrate
measurement due to particle loss.
D. Loss of Non-Volatile Carbon in the Denuder of the Particulate Carbon
Sampler
To assess the extent of loss of non-volatile particles in the Al 2o3-
coated denuder of the true particulate carbon sampler, the quartz
fiber fil tars from samplers 4 and 5 were analyzed and comp a red for
elemental carbon by the reflectance method (RM-4), discussed
previously. Figure 18 is a scatter diagram of the Ce results, on
the two samplers. On average the loss of C was negligible.8
E. Retention of Carbonaceous Material on Hi-Vol Filter Samples
The retention of carbonaceous material on filters is the net result
of adsorption of gaseous organic compounds on filters a-nd on
previously collected particles, and volatilization from the filter
of initially particle-phase materials. Some insight into the
relative significance of such errors is obtained by comparing
results for a single long-term filter sample with a series of short
term samples. Figure 19 compares observed total carbon values for
samples collected for 16 hours against 16-hour average values
calculated from four, successive 4-hour samples. Calculated values
are consistently higher, suggesting that adsorption of gaseous
species on relatively fresh filter surfaces and/or volatilization
from filters during prolonged sampling were important. Adsorption
of gaseous material on previously collected particles, which would
yield 16-hour sample results higher than those calculated from the
short-term samples, appears to be less significant.
- 41 -
Figure 18. Loss of Non-Volatile Particles in the Denuder of the Particulate Carbon Sampler.
1:t CW')-s ....... = .5
LO a: w ,_J a.({ :E <( rJ) m
CJ
t
8.00 ---------------------------
7.00
6.00
5.00
4.00
3.00
2.00
1.00
y = 0.0874 + 0.999x
r = 0.9985
0.00-------...._i.....i,_.._______......._.__.,__.__...___...__..____._-'-_._...._.,__,,--i.,__.__..._..i....., 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
C8 SAMPLER 4 (µg/m3)
- 41a -
>-0 => t-
0 ~ (J)
..Q,._, ta
>-c...>
r-1 t-ta +J 0 E-t 1--t ,._, ::, _j 0 :I: I t-i
\.0 CDr-1
"O Q) t--1 +J ta (J)
r-1 ::, CJ t--t
r-1 a > Cl]
::>
"O CD
Q)
::> a:::,._,
<QJ U]
..Q0
i-1°' N a, ,._, CD ::, O'I 0)
·r-1 ~ ~
·-\
\ -, •\ \
·.\
\ \
+◊o
un N
0 N
+
tl1 ,-◄
+ M,;:JO r-1 \.0 \.0 O"I
00
II II
~ ,._,
0 r-1
; i l
..... 0 LN
Il i ~
in
,......., ('Y')
~ O'I ;1....._,,,
a tJ.J >a:: w (I) CD a
<EUI/.6rt) 1J t3noH gt 031 v1nJ1VJ
VII. ATMOSPHERIC RESULTS, PART-- I
A. Tabulation of Results
Aerosol collection was done for 109, 4-hour periods. Table 10
compiles results for the more significant parameters measured
directly or indirectly for aerosol and gaseous pollutants. Except
as noted all chemical s-pecies are listed in µg/m 3 • Table 11
explains the meaning of the headings used. Table 12 provides mean
concentrations for the chemical sp~cies and meterological
parameters. The downtown Los Angeles site shows highest average
levels for sulfate and carbonaceous material. Riverside exhibited
the highest particulate nitrate.
B. Summary of Visibility Results
Table 13 lists mean and maximum values measured for visibility
parameters at the three sampling sites. Mean values are for 30 to
36, 4-hour periods at each site.
Light scattering by dry paticles (i.e., b as measured by an . sp integrating ne~helometer heated 16 ~ 3°C above ambient temperature)
was the largest contributor to the total extinction at all sites,
representing somewhat more than half of the extinction coefficient,
b'ext· Light absorption by particles, which can be related ~lmost
exclusively to elemental carbon (42), was relatively small at sites
in the Los Angeles basin where high concentrations of efficient
light sea tt e ring species were present. However, at San Jose
absor~tion averaged nearly one-fourth of the b' t. Conversely,ex
the apparent contribution of water was relatively high at Riverside
and Los Angeles and small at San Jose. This is consistent with the
relatively high sulfate and nitrate concentrations in the L.A.
basin, species which can absorb and retain water. The light
absorption calculated for N0 2 , b , remained relativelyag
insignificant at all sites. The daily changes in the values of
- 42 -
~ ri-~
Table 10. Summary of Atmospheric Results
TIME FINE COARSE FIN£ COARSE HN03 NH4 NH3 FINE COARS~ TOTAL FINE TOTAL R BAG 03 RH TEMP BSP i3SW BAP
504 504 N03 t-!03 CE C£ co MASS MASS FINE E-4/M PPB '¼ C E-4/M 710:i.571019 1. 96 1.48 2. 52 ·-0. 83 3.98 0.34 2.04 0.61 0.2'1 7.61 44. 78 50.28 29.04 43. 00 24.89 0.01 7101971023 1. 59 0. 40 1. 36 0.20 0.64 0.0 5.41 0.76 0.20 4.69 22.93 28.69 13. 16 65.80 18.33 0. C-4 7102371103 0.62 0. 41 0.84 0.09 0.65 0.0 5. 74 0. 72 0.02 0.98 16. 53 19.54 13.01 77. 50 15. 72 0.05 7110371107 1. 13 0. 71 1. 20 0. 18- 0. 67 0.0 4.41 1.08 -o. 57 -0.51 11.94 30.98 · 8. 43 76. 80 15. 17 0.05 7110771111 2. 62 0.21 2. 46 0.74 0.29 0.40 8.47 0. 36 0. 53 4.81 19.87 38.90 8.01 51. 50 20. 83 0. 06
7111171115 1. 27 0. 06 1. 65 0.0 3. 65 0.66 0. 14 6.80 14. 17 25.00 2.61 35.80 27. 50 0.22 0.07 7111571119 l. 45 0. 19 L75 : 0. 21 - l. 58. 0. 0 - 2. 18 - 0. 47 -0. 20 7. 79 13.47 23.03 -0.03 40.50 24.44 0.21 0.02 7111971123 l. 56 0.60 1. 77 0. 70 0.62 0.0 2. 19 0. 71 0.03 5. 18 17.06 32. 42 5. 92 0. 12 57. 00 20. 56 0. 25 0.06 7112371203 1. 39 0. 90 1. 84 0.47 0.22 0.0 3. 76 1. 81 18.99 40.21 10. 74 0.09 67. 00 18. 15 0.27 0. 05 o. 21 7120871212 1. 86 . 0. 81 8.32 4. 88 _0. 74 -- 1. 17.. 5.43 3.89 .0.27 16. 77 58 33 .109. 25 . 25. 31 0. 05 --·- ---- -- 23. 05 0. 92 0.07 0. 68 7121271216 2.38 1. 15 12. 81 8.43 6. 19 1. 30 7.82 5.36 0. ~6 22.91 49.64 109.37 7.80 0.02 28. 5:; 1. 15-0. 24 l. 13 7121671220 2. 16 0. 43 5. 90 2. 58 8.05 0.27 2.44 2.28 0. 72 14.25 29. 59 55.02 4.81 0.05 30. 15 0. 56-0. 05 0.69 7122071224 l. 84 1. 18 1.92 1. 99 0.67. 3.02 1. 41 0. 1~ 5.95 18.42 54.87 5.35 0.06 49. 50 22.35 0.40-0.0l 0.09 7130871312 2. 83 0. 98 10.87 1. 61 1. 04 1. 83 9.62 3. 94 0. 52 17.20 37.6i 96.04 2.97 0.29 45. 00 23. 15 0. 93 0. 11 o. 4-~ 7131271316 1. 74 0. 36 3. 98 0.47 2.61 0. 17 3.74 1. 84 0.25 12.22 21. 76 45.83 0.81 0.07 28.80 29.65 0.41-0.05 0.36 7131671320 1. 09 0.41 3.25 0.39 0. 76 0. 10 7. 63 1. 21 0.34 9.70 19.96 43. 13 - 4.01 0.07 34.80 28. 55 0.27-0. 01 0.22 7132071324 0.83 0. 43 1. 36 0. 42 0.36 0. 0 4.43 1. 22 0. lo 5. 57 7. 51 29. 50 -2.63 o.os 8.0 48.80 22. 11 0.22 0.0 O.C,5
7140871412 1. 54 0. 40 4.26 2. 33 0. 10 0.0 4. 78 2. 15 o. 4n 12. 17 22.01 73.35 1. 15 0.09 1::.0 44.80 22.11 0. 58-0. 10 0.14 71-11271416 l. 36 0.45 3. 79 4. 26 _ 9. 38 0.0 _ 4. 85 2.45 0.4:L 18. 15 22. 13 62. 74 -5. 12 0. 12 57. 5 30. 50 30.28 0. 61 0.02 0. 16 7141671420 1. 03 0.37 4.85 1. 86 2.25 0.0 3.06 1. 66 0.24 11.98 23.06 44.30 4. 07 0. 11 35.0 30. 80 31. 11 0. 32-0. 02 0. 12 71.:;2071424 1. 58 0.89 ~ 91 !:J ::IA 0 85 0.0 4.20 2.26 1. 11 19. 66 19.50 145. 48 -11.80 0. 14 8. 0 42.00 24. 72 0. 44-0. 01 o. 19 7150871512 1. 90 0. 75 3. 89 5. 45 1. 97 0. 51 3.66 3.20 0.32 10.44 28.00 82.61 5. 39 0. 11 20.0 43. 50 23. 78 0. 57-0.06 0.22
ii:::. 7151271516 2. 09 l. 07 5.03 9. 80 8.64 0. 71 3. 19 2.67 0. 41- 18. 07 28. 71 79. 76 -2.98 0. 14 37. 5 32.00 31.67 0. 77-0.08 0.21 t\J 7151671520 1. 47 0. 73 2. 73 l. 89 1. 36 0. 17 11. 37 1. 33 0. 10 9J-J9 22. 51 49. 13 4. 10 0. 12 30.0 33. 80 28. 33 0. 48 0. 0 0. 08 Ill 7152071524 1. 20 1. 31 2.06 1. 57 0.67 0. 10 11. 13. 1.86 0. 1 i 9. 76 22.09 74.90 4.69 0. 13 30. 0 37. 50 21. 83 0. 37 0. 11 0. 14
7160871612 2. 51 0.33 3. 10 0. 47 3. 24 0. 10 2.35 i. 49 0.44 7. 47 36.83 62.35 17. 53 0.05 20.0 51. ~o 19. 44 o. 49 o. 12 o. 06 7161271616 1. 75 0. 53 4. 51 2. 61 0. 51 0.25 3. 10 1. 42 0:38 11. 50 28.27 58.65 5.83 20. 0 33. 50 26. 28 0. 47 0. 08 0. 18 71c167l620 l. 30 0.80 1. 90 1. 12 0.87 0. 10 4.05 0.92 0.24 6.49 16.46 43.96 3. 76 0.07 20.0 40.00 23.22 0.35 0. 11 0.09 7162C71624 1. 50 0. 76 1. 26 0.67 1. 01 0.0 2. 52 0.83 0. 18 3. 47 14. 76 33. 41 6.26 0.09 20.0 57. 50 17. 39 0. 29 0. 19 0.07 7170871712 1. 75 0. 53 1. 53 1. 40 1. 17 0.0 1. 70 0. 73 -0. Oc-! 6.03 20. 46 31.08 a. 02 0.02 20. 0 53.80 19. 17 0.27 0. 13 0. 06 7171271716 1. 5:5 . 0. 48 2. 10 4. 16 4. 17 ·- 0. 47 2.46 0.97 0.25 13.29 17.04 37.01 -3.80 20.0 33.80 27. 11 0.38 0. 10 0. 08 7171671720 0. 83 0.20 1. 21 l. 24 1. 66 0.0 1. 85 0.67 0. 17 8.45 10. 39 25. 58 -2.68 20.0 41. 50 25. 56 0. 19 0. 07 0. 04 7172071724 1. 06 0. 24 1. 14 0. 49 0.86 0.0 3.38 0.82 0. 19 2.89 10. 41 28. 11 3. 45 0.07 25. 0 59. 30 19. 61 0. 20 0. 15 0. OS 7180871812 2. 77 0. 57 5.~5 4. 39 1. 46 1. 41 2. 67 .. 0. 84 0.07 8.99 13.03 57.65 -5.35 0. 05 17. 5 64.00 18. 33 0.94 0. 44 0. Ol 7181271816 2.88 0.63 5.28 ~.68 4.39 1. 04 3.93 0.89 0. t~} 11. 80 19.00 37.87 -1. 25 0. 05 20.0 40.00 26. 56 0.95-0.05 O.Ol 7181671820 1. 42 0.39 1. 54 2. 36 2. 29 0.09 1. 90 0.69 0. 1::1 9. 91 15.00 26. 74 -0.45 47. 5 44.80 25. 17 0. 24 0. 10 0.01 7182071824 1. 47 0.27 1. 15 0.89 0. 72 0. 10 2.05 0.95 -0.2:i. 5.27 11. 53 18.26 1.23 0.08 20.0 57. 50 20. 44 0.22 0. 16 O.Ol 7220872212 9.32 0. 53 37. 83 18. 70 0.67 9. 57 11. 56 5.64 1.36 27. 76 73.93 175. 64 11.20 38.30 28. 06 0.34 7221272216 6. 58 1. 11 19. 32 12. 20 9. 84 5. 96 43 68 2.62 0. 95 26. 15 45.24 123. 73 -0.65 207. 5 25. 00 36. 28 2. 34-0. 11 0. 18 7221672220 5.46 0.81 6. 77 8.99 7.09 4.39 21.81 20.08 42. 14 91.47 3. 13 0.09 105.0 25. 50 34.89 1. 15-0.06 0. 18 7222072224 4. 45 1. 13 6.00 8.98 2.83 4.30 22.20 3. 76 -1. l'l 17.04 30. 51 110. 36 -6. 10 0. 13 40.0 37.00 28.89 0. 90 0. 17 7230872312 7. 81 1. 71 23. 09 11. 59 ~o. 40 12 43 20. 55 4.36 1.0tl 20.57 60.66 173. 74 1. 65 6 7. 5 42. 50 27. 11 1.84 1.61 0.46 7231272316 6. 88 1. 12 22.92-17.34 15. 56 5. 43 44. 08 2.37 0. 7U 20.66 47.02 109.98 6. 75 o. 16 137. s 31.00 33.e9 2.04 o. 14 o. 2J 7231672320 7. 06 0. 91 7. 14 5. 74 10. 23 4. 19 29. 78 2.09 0. 5c;, 18.20 38. 44 96. 77 2.02 0. 17 92. 5 30.30 34. 17 1.25 0.23 0.23 7232072324 6. 88 L 61 s. 42 9. 53 0. 99 4.67 20.03 2.27 0. ;j,:~ 17. 13 39.03 114.49 -0.55 25.0 42.80 28. 50 1. 11 0.34 0.23 7240072404 5. 75 2.03 6. 78 10.01 -0.89 4.30 22.44 2. 53 . 0. :3:1 12. 95 33.08 92.39 1. 36 9.0 47.50 26.00 1. 03 0.32 0.24 7240 1172408 5. 50 l. 61 7. 12 8. 56 -0. 83 3.82 20.33 2.84 0. 5:r 11. 64 37.08 90. 17 5. 16 10.0 53.60 23. 50 1. 05 0. 52 0.27 72408i'2412 8. 44 l. 79 22. 45 6. 70 7. 04 2. 98 28. 56 3.04 0. 74 0.0 65. 57 151.02 33.92 0.22 70.0 47. 50 26.94 2. 23 0. 76 0. 23 7241272416 .7.08 .1.46 14.59 __:-:-0.66 14.23-4.0341.76___ .1.41 a. :m.-1.i. so 39. 53 __ 90..04 ____3_ 30 _Q. 14 -120. 0 35. 00 33. 50 1. 88 0. 17 0. 15 7241672420 5.81 1. 28 4. 77 4.20 7. 63 3.-16 20.04 0.90 0.61 15.92 30. 54 85.09 -0.99 0. 15 72. ~ 33.30 32.67 o. 90 0. 13 0 10 7242072424 6. 04 1. 40 6. 91 8.89 2.65 4. 78 21.83 l. 33 0. 4'J 16. 73 31. 27 112. 07 -7.38 0. 19 35. 0 41. 80 27. 50 1. 10 0. 31 0. 09 7250072504 6. 02 __ 1. :56.__9 __36 _ 9. 80 __Q, 41. .4. 60 21. 36 ____ 1.58 _ 0. 19 11. 33 39. 63 .. 91..80_____ 4. 62 o. 22_ 15..o 53. 30 24. 11 1. 32 o. sa o. oe 7250472508 6.05 1. 79 10. 02 9. 60 0. 61 4.48 19.45 1. 31 42.44 91.95 17.20 0.22 12. 5 62.CO 22. 11 1.39 0.80 0. 18 72:iOS72512 7. 40 2. 62 21.80 13. 15 9.4011.12 3. 10 2. 42 0. (~9 20. 43 59.92 140.00 2.31 0.24 77. 5 53. 50 25.83 2 26 1.25 0. 18 7251272516 _ 9. 92 ____2 ..33 .23-24 ...1L 59_ 16__ 5.7 12_.44 :z:L~4 ___LJ8 ___Q. 9:> 27.. 57 68. 27 144. 63._:___.4....13 ...0. 17__ 120~ 0_3.'i'. 30 33. 50 2. 75_ 0. 82 0. 16_
Table 10. Summary of Atmospheric Results (continued)
\. i ~--
TIME FINE COARSE FIN£ COARSE HN03 NH4 NH3 FINE COARS~ TOTAL FINE TOTAL R BAG oa RH TEMP BSP BSW BAP S04 GC4 N03 N03 CE c~ ~n MARR MA~R FTNE E-4/M PPB x_ .. G ~-4fM_ . __
,:__:.w-·.;_;;~;~,J 1/. 12 ~- S''i· i.-!. •Jb '/. 56 14. 3;- 11. 05 18. 62 l. :iJ. 0. 84 2~i. ~4 68. 08 134. 75 6. 'JO 0. 18 150. 0 38. €v ::;d. '1"4 c::. ::.:~ 0. -!:."+ v. l~ 7252072524 12. 98 2.31 22. 10 12. 35 0.68 12. 56 18.24 1. 71 0.25 20.00 75.00 138.99 7. 57 0.21 67. ~ 47.CO 2~39 2. 52 0.88 0. 11 72~0872612 8. 70 l.86 16.86 7.36 13. 04 5. 59 16. 79 3.26 0. 16 13.83 54.00 109. 80 11.65 0.23 40.0 54. 30 25. 72 2. 33 0. 71 0. 25 1261212616 2. 55 o. 11 o.99 o.39 9. 79 o. 11 a.63 o.63 o. 14 7.84 ~o. 50 6.43 o. 12 45.o 36.oo 32. 94 o.31 o. 13 o. 05 7261672620 9. 73 o.e6 11. 47 16. 45 1. e3 6. so 1e. 22 2.33 o.67 21. s2 63.24 103. 11 14. 12 o. 10 110.0 36.oo 32. 50 o. 90 1. 32 o 20 7262072624 7.44 l. 16 5.20 12.31. 3.57 5.01 17.35 2.33 0.32 15.97 44.91 103. 13 5_90 0. 25 25.0 43.00 27. 11 0. 90 0. 55 0. 20 7270872712 6. 40 1.91 20. 42 14. 27 8.20 10.23 19. 92 4. 55 0.64 19.26 63.8~ 155.39 11.28 0. 32 57. 5 39.00 26.00 1.66 1.27 0. 36 7271272716 9. 54 1. 61 26. 77 10. 13 11. 72 U. 42 48. 10 2. 59 ·O. 71 25. 18 64. 57 149. 06 10. 97 O. 18 130. 0 33. 80 33. 50 2. 13 1. 16 0. 24 7271672720 9. 60. 1. t,9 12. 49 10. 35 -12. 21 . 7. 24 :?3. 20 2. 42 _Q, 67 24. 58 ·- 58. 99 136. 84 --· 4 .. 26 .0. 19 105. 0 34. 50 32. 39 1. 43 0. 72 0. 21. 7272072724 5.69 1. 59 10. 19 10. 60 1.92 5. 66 19.95 2.34 0. 58 19.35 59. 75 117.05 15.22 o. 28 16.3 38.30 27.22 1. 14 0. 22 0. 19 7280872812 6.03 1. 40 11.00 5. 55 15.03 4. 45 9. 58 4. 56 0. 94 15.86 48. 56 115. 92 12. 53 o. 32 45.0 40.80 26.00 1.21 0. 47 7281272816 6. 53 1.46 ~3-~2 -0.26 17. 74 7. 50 38. 10 2.55 0. 77 20.62 60.32 126.22-- 15. 11 o. 18 117. 5 31.80 32.67 2. 11 0. 24 72S1672s20 6.22 o.ea 16.00 4.67 6.09 4.89 19.25 2.30 o. s1 19.82 41.32 95.23 3.e3 o.24 75.0 34.30 32.22 o.99 o. 43 0.21 7282072824 2. 61 0.96 1.83 3. 57 0. 67 0.80 13.49 1.03 0.35 10.25 22.88 54.35 4.84 0.20 15.0 42.30 27. 50 0. 27 0. 15 0. 10 7290872912 3. 74. 1.27- 5. 58 4.25 12.27 13.76. 4.05 1.22 14. 47 31.08 109. 56 ~-2.33 0.27 37. 5 43. 50 26. 56 o. 85 0.31 0. 40 7291272916 S. 65 0. 78 21. 11 4.99 13.64 6. 52 36.42 2.36 1. 12 22.08 43.47 125.05 1. 43 0. 18 110.0 32. 50 33. 78 1. 33 1.34 0. 20 7291672920 4. 88 1. 42 7.87 9.24 8. 69 4.62 27. 54 1. 98 0.68 21. 16 39. 51 111. 73 1.29 o. 19 75. 0 33. 30 33. 22 0. 75 0. 63 0. 17 7292072924 2. 69 1.04 3.20 -6.39. 0.65 -1.55 22.45 l. 59 0.60 14.35 23.67 77.25 -0.34 0. 20 22. 5 35.60 28.06 0. 42 0. 25 0. 13 7310073112 14. 25 2. 1s 13.83 12. 66 12. 1~ e. 55 4. 78 3. 4~ o.s2 19. 48 77. 78 139. 44 18.42 0.20 s2. 5 58. so 2s. 56 2. ss 1.00 o 34 7311273116 21. 52 1. 57 8. 01 48. 23 -4. 76 9.86 4.60 2. 75 0. 99 23 95 83. 12 132.40 17.85 0. 18 135.0 45. 30 30.28 2. 22 0. 99 0. 26 7311673120 16. 99 l. 78 4. 12 --7. 48 12. 56 5. 00 2. 96 1. 46. 0. 64 16. 45 56. 43 98. 12 6. 82 0. 20 82. 5 55. 30 27. 39 l. 41 1. 18 0. 08
~ 7312073124 18. 30 -0. 37 8.63 1.66 9. 73 7. 18 3. 25 1. 77 0.46 92.69 107.66 31.85 0.24 27. 5 70. 00 22. 78 3. 18 0.20 0. 14 ~ 8Gl0880112 27. 93 2. 07 9. 91 8. 56 2. 10 2. 43 0. 58 18.32 79.25 125. 72 9.93 45.0 65. 00 23.33 2. 34 2. 11 0. 19
8011280116 40. 18 1.64 12. 68 -o. 40 28.44 17. 45 4.90 2.20 0.48 19.36 108. 17 120. 70__ 20.08 0.23 102. 5 50.30 27. 11 2. 96 1. 58 0. 28 8011680120 23. 95 1. 41 2. 77 2. 47 10.85 5. 72 1. 55 l. 11 0. 50 12. 14 61.36 83. 47 11.60 0. 16 67. 5 56. 50 25.83 l. 57 1. 23 0. 10 8012080124 20. 25 1. 06 6. 69 4. 35 4. 26 7. 11 0. 59 1. 52 0. 18 11. 14 58. 11 82. 49 10. 44 32. 5 68. 80 21. 39 l. 71 2. 25 0. 17 8020880212 18. 26 1. 41 8.62 5.08 13.68 6.39 1. 57 4.64 0.88 18. 72. 64.85 118.83 7. ~7 0. 28 27. 5 61.00 22. 67 1. 78 1. 74 0. 50 8021280216 19. 36 1. 55 5. 28 -0.06 27. 71 6. 47 1.81 2.81 0. 94 20.01 52.45 122.03 -2.42 0. 22 67. 5 47.30 26. 56 1. 62 0. 94 0. 33 8021680220 13. 70 3. 16 3.65 6. 75 9. 10 3. 53 0. 70 1. 50 0.36 12.02 4B.o7 76.4B 12.77 0.20 42.5 53.00 25.28 1. 06 0.93 0. 18 8022080224 10.60 2.84 2. 67 8.66 1.41 2.46 0.69 0.98 0.33 6.45 33.48 58. 12 9.00 0. 19 22. 5 70.30 20. 44 0.83 1.03 0.07 80~0880312 11. 90 2. 79 8.29 7. 94 13.03 5. 18 2. 24 4. 11 1.31 17. 31 58. 58'111.41 12.02 0.28 25. 0 55.00 24.25 l. 49 1.29 0. 47 8031280316 10. 92 3. 22 5. 61 11.23 10. 46 3. 62 1. 59 1. 78 0. 71 15. 50 43. 54 91.21 6.98 0.21 45.0 47. 50 26. 95 1. 15 0. 59 0. 23 8031680320 10. 87 2. 91 3. 54 11. 15 7.31 3.69 1. 19 1.86 0. 51 14. 12 55.81 81.43 20. 54 0.23 35.0 48. 50 24.28 1. 01 0.62 0. 17 8032080324 10. 08 2. 69 3. 80 1C. 17 0. 42 3. 63 1. 19 1. 40 0.31 10.30 41.63 71.34 11.39 0.25 15.0 68. 30 21. OS 1. 02 1.09 0. 15 8040880412 15. 74 4. 11 28. 07 22. 69 14. 42 14. 09 2. 05 10.81 2.39 36 73 132. 94 210.33 32. 42 0. 41 40.0 57. 30 24.25 4. 21 1. 55 l. 38 8041280416 20. 14 3. 91 s. 12 18.88 33.09 1. 96 3.42 0.85 25.04 78. 16 130.45 14. 55 0. 18 80.0 40. 30 27.67 2 j4 0. 47 0.3~ 8041680420 9. 59 l. 40 4. 55 8. 74 13. 14 3. 12 1. 55 2. 16 0. 70 15.30 46.33 73.20 9.94 42. 5 38.30 27.22 1. 00 0. 45 0. 25 8042080424 4. 63 1. 44 3. 51 6. 5S 0.94 1.24 2. 16 2.06 0.39 11. 18 33.20 62.33 8.00 0. 19 8.0 61. 30 2~.05 0. 62 0. 41 0. 2i 8050880512 7. 94 1.88 13.85 -4. 50 30. 40 6.42 3.80 11.42 3. 13 36.63 91.25 182. 75 16.62 30.0 43.00 26.00 2.37 0. 70 1.29 8051280516 10. 64 1. 25 7.99 7. 14 38. 17 3.63 4.03 3.34 0.95 21.90 57.44 97.30 10.94 52. 5 35.30 30.28 1. 49 0.64 0.41 8051680520 6. 19 o. 55 4.82 6. 51 16. 11 2. 03 1.95 2.07 o. 55 14. 17 38.27 70.87 7.98 0.20 32. 5 33. 00 30. 28 0. o9 0. 16 0,22 ao52oso524 4. a2 1. 29 4. 72 9.25 1.9s 2.22 3. 67 4.23 o.sa 16.38 38.66 102.02 3.37 0.20 1s.o 51. 3b 24. 12 o. 67 o.36 o. 44 8060880612 9. 18 1. 18 15. 33 9. 32 28. 49 · 9. 64 11. 1'l 2. 58 29. 55 111. 81 166. 88 36. 71 22. 5 46. 30 27. 2:2 2. 46 0. 50 1. 3i) 8061~80616 8. 84 l. 53 11. 92 82. 92 -a. 54 4. 76 7. 17 6. 54 2.47 34.21 76. 14 169. 44 12.03 72. 5 31. 00 33.89 2. 31 0. 72 0. 65 8061680620 12. 54 1.31 8. 11 11.73 38. 56 6.48 1.84. 3.79 0.97 24. 53 68.20 125.93 11.29 0.25 57.5 39.00 30. 44 l. 66 0. 48 0. ~5 0062oao624 7. 20 1.60 6. 99 13. 13 3.26 3.83 6 71 3. ~o 1. oa 17. 49 44. 42 119.88 2.25 0.22 e.o 55.oo 25. 12 o.87 o. 53 o.3a 8070880712 6.62 2. 11 11.04 14.83 15.27 4. 52 4. 54 4.34 0.84 19.87 61.97 136. 79 12.88 25.0 60.00 24.61 1. 17 0.68 0.47 8071280716 8. 69 1. 82 8. 12-21.38 26. 51 2.93 4. 19 1.9<, 0.32 17.41 46. 53 89.02 6.98 0. 14 40.0 39. 30 29.89 1. 02 0.06 0. 19 8071680720 8. 57 0. 91 5. 11 6.68 13.46 2 80 2. 50 1. 53 0.35 13.44 40.97 73. 18 7.30 0. 15 32. 5 41. 30 29.33 0. 79 0.32 0. 17 8072080724 4.64 0. 44 3. 48 6. 41 l. 15 1. 35 3.66 2.40 0.47 12. 10 36.90 80. 12 10.01 9.0 66.30 22. 78 0. 69 0. 48 0. 27 8080680812 14.69 1.69 17.33 5. 58 13.46 6.32 3. 54 0.61 18.24 71.69 118.35 11. 74 30.0 58.80 24.83 2. 53 1. 15 0. 39 aos12eos16 11. s1 o.73 7.33 -2.38 2s.a4 4.21 3.32 2. 12 o. 56 10.32 54.79 01. sa 12.34 o. 15 45.0 42. so 20.a9 t.67 o.45 0.22 8081680820 4.21 1. 85 3. 93 l. 85 11. 49 1. 52 4.32 1. 3,J 0.43 12. 11 33. 55 67. 92 9. 42 0. 15 25.0 41. 50 28.89 0. 72 0. 17 0. 16 eoa2oso824 3. 77 1.01 3.23 s.92 0.25 1.aa 4.01 2.12 o.53 11. 11 34. 57 11.60 -23.79.o. 19. _a.a 61.00 23.06 o. 56 o. 57 o. 1a
Table 11
Explanation of Headings Used in Table 10
Heading
Time
Fine S0 4
Coarse S0 4
Fine NO 3
Coarse NO 3
Fine CE
Total CO
Fine Mass
Total Mass
BAG
BSP
BSW
BAP
Meaning
Designates the date and time of the 4-hr period. For example, 7101571019 indicates a starting date of July 10, star ting time of 1500 hours, ending date July 10, at 1900 hours.
Sulfate collected on a Teflon filter downstream of a cyclone with cutoff at 2.2 µm. µg/m 3 •
= Coarse sulfate calculated as the total S0 4 colle~ted on a Teflon filter without preceding cyclone less fine S04 • µg/m3 •
True particulate respirable nitrate measured with a Teflonc oa tad cyclone plus acid gas diffusion denuder ahead of Teflon and NaCl-impregnatad cellulose. fil tars. µg/m3 •
Total inorganic nitrate (TIN) collected with a Teflon and NaClimpregnated filter less fine nitrate collected with a TIN sampler preceded by a cyclone (cutoff 2.2 µm). i.ig/m3 •
Measured with a TIN sampler plus cyclone less Fine N0 3• µg/m3.
Fine elemental C. Mean results by RM and LTM methods on samples collected with quartz filters preceded with a cyclone ·(cutoff ca. 2.5 J,1ln). µg/m3.
Total organic carbon. Total carbon measured on hi-vol fil tar samples with baked glass fiber filters less mean C by the RM and LTM on hi-vol samples. µg/m3. e
Mass cone en tra tions measured on Teflon fil tars preceded by a cyclone (cutoff 2.2 µm}. i.ig/m3 •
Mass concentrations measured on Teflon filters without cyclone. µg/m3.
Light absorption coefficient due to gases ( i • e • NO 2 )" in unit - t+ - l
10 m •
Light scattering coefficient due to particles measured with a - t+ - l
heated integrating nephelometer at 550 nm, in units 10 m
~ght scattering coefficient for water estimated as the
difference between the reading of a nephelometer at ambient _ I+ - l
temperature and BSP, in units 10 m • 41:fht absorption coefficient measured by the IPM: in units 10-
m
- 42c -
Table 12
MEAN CONCENTRATIONS OF AEROSOL CONSTITUENTS AND METEOROLOGICAL PARAMETERSa
.S-peciesb San Jose Riverside Los Angeles overall
1.66 13.04 7 .. 17
3.38 13.54 7 .. 88
FC 1. 54 e
0.61 1.80
2.. 11 8.28 7.,93 6.r:n
0.30 6.09 ;.80
0.,25 0 .. 60 o.~
9.44 17 .. 58 18.31 15.03
22 .. 26 47.47 61 .49 43.54
TSP 51.49 115.70 90.85
4.33 24.54 3.20 10.63
IDTO 3 2 .. 13 7 .41 16.74 8.69
0 3 ( p-pb) 24 72 42 49
NO 2 ( ppb) 18 .. 79 50.44 47 .05 38-58
RH(%) 46.90 39.84 51 .73 46 .15
T (°C) 23.66 29.5g 26.0; 26.40
a. Exce-pt as noted, concentrations in µg/m3. Sampling between July 10 and August 8, 1982, generally during the hours 0800 - 2400.
b. F = fine, C = coarse
- 42d -
--
Table 13
SUMMARY OF OPI1ICAL MEASUREMENTS AT SAN JOSE, RIVERSIDE and LOS ANGELES (10-4m-1)
San Jose Riverside Los Angeles
Max. Max. Max. Parameter Mean %of b' t Value Mean 'I, of b' t Value Mean %of b' t Value ex ex ex
b Oe 17 ± 0.23 21.8 1.13 0.21 ± 0.10 8 .. 6 0.47 0.36 ± 0.32 12.0 1..38ap
b 0.47 ± 0.27 60.0 1.15 1.44 ± 0.70 59.,3 2.75 1 .61 ± 0.86 53.7 4.21 sp
b 0.050 ± 0.12 6.4 0.44 0.58 ± 0.45 23 .. 9 1.34 0.82 ± 0.53 27.3 2.25.i::,. SWN (D
b 0.090 ± 0.05 11. 5 0.29 0.20 ± 0.05 8.2 0.32 0.21 ± 0.06 7.0 0.41ag
b' 0.78 ± 0.38 2.43 ± 0.84 3.00 ± 1.06ext
each component of b t are illustrated in Figures 20-22. These ex
plots show, in general, that diurnal changes are correlated among
the four coefficients, with the poorest correlation for b ; this· ag
probably reflects the relatively small changes experienced in the
concentration of N0 2 • The onset of a relatively polluted period in
Los Angeles, beginning August 4 is clearly seen.
c. Diurnal Variations for the Dominant Aerosol Species and Ozone
Figures 2'3-25 show the average diurnal concentrations for all a
sampling days for the pollutants S0 4 , N03, C , and 03 at each e
site. In Riverside, two 24-hour runs were performed. The data
points shown at 0200 and 0600 hours are means for these two
episodes only. Thus conclusions drawn based on differences between
these data points and those for other time periods averaged over
eight days may be in error because of a statistical bias. At both
San Jose and Riverside, sulfate concentrations showed relatively
little diurnal change. In Los Angeles the sulfate concentration
was higher and exhibited a sharp maximum in mid-afternoon,
paralleling that for ozone, and similar to previous observations
(21). At both Los Angeles and Riverside fine particle nitrate
exhibited a morning maximum. This contrasts with earlier findings
(9) which found a large particle N03 maximum in the morning and a
fine N0 3- (< 0.5 µm) maximum in the afternoon. This earlier study
was hampered by both positive and negative sampling errors for
nitrate, a factor which might account for this difference. Four
hour average ozone maxima at the sampling sites were, o.o;, 0.13
and 0.08 ppm at San Jose, Riverside and Los Angeles, respectively.
Thus the study was performed dui,ing periods of generally light
photochemical smog.
D. Size Distribution for Aerosol Constituents
Figure 26 compares mass measurements with the fine (~ 2. 2 um) and
total particulate samplers. The data show moderate correlation
- 43 -
t
........ N 00 a, ....;
,{ (F)
w ~
w ::) (.f)
C)__J ...., < > <
z U1
OJ
0 N
QJ s... :::s C') ,,-
'-'-
r---··-----~-----~I --------- ---1 -------
1
I ! I I i I
I I
I j
l j
.....
013 0..0.. IJ0)0,0
...0 ..c ...a ...c i ' J
! I i
>< 0
t-0!) 9 - 43a -
r---------------------------~-- ---i
[J"'J ~ 0.. Cl.. 0 CQ 0) 0
..c ..a ...0 _o
~! j
rl I
....__, I
(f)
w w Cl:J t-t
_j CJ')
< 0:: w > >
CD ,.... N
(lJ !'.,.
:::s 0, ......
LL..
/ iJJ\t- 9
- 43b -
Figure 22 B VALLJES
' LOS ANGELES (1982) -------------- -------------------------- --- ---- -----·-------1
A 4. 01 I\
i \ =- t:~ !
I3.5 --·-·- bspI \ ----~ bap
3.0 t· I \ l \ A 1· .\
,......,. \ . .\ ,. .-i E 2. 5 ~ \ I \ / \ / \ . A
! I • \ i . I'· I . ,t,. ~ w • i \ 1 • - :-... 1 \ I\ · \ 1\() l'Sl
2. e!- d i\ L ...._ I \ I"\ I \ I \ '-.../
i ~
\ I . ·. I • • • • / / \ / •I ·, •: . \.,;:.--·":\ I \ I "\ • '\' J • • '- II
.· \~· :· ·.. V. : \\ . . . ; / ·, ; .....0 1 5• ,__ • : ·. :• ·. t\. ! • • \ I . i I \
f • • : \ • \ . • • . • I J • I •I . . . . . . . ; : -: : \ ,< .·. \ ,· . . /~ "\ / \ . / ,\ \ / . I • •\ -'• o • • / 'i •• • • ./ • • • I • •
I1. 0 i- ...... : ; : \ . . . . . : ,.__J ~ • I \ \ f I ' . /A· . .-·: \ 1 1! : ; ••• \/ :_ ! l \ / / i / \ ~-✓ \ / .-· '. ·,
I • • • • I . \Ii\ !Jf \ \... l • • • • " v• \ ', ; • • , : : •. • .. • • • ~- • 0 • e ." 0 Cl e •
! •. . • • • •• I f. I• · · \.• I . · ,.\. • • . · · · · . . · · . \0 5 : • t- . . ,. r 1· \" .· .. · . . . .
! • • '\ \ / • • • • • • • •• / • • • • ' • -- \:" • • •• ••• / ,,, _..--~., •• - •• "'I. • .... .. \ / ,,....- ', ' ,. . . --· .. . .
! ," •• ,,. / -- .., -- \ __,,- \ / ": J.· • . --- . . \ •. ... --- , I"- . ..- ---- ·~ -~ ,'- -- - --~ ---- ·-1 . ........ . . --- . ---- 1 • '\ .•. . ·~y..,,,- -, -·- - ~ --- ,----- - ..._ ·-----------· __, -~- ..___ .\ •...,/' ' ,,,.-· _::...,_ ', / ;,--' ✓ • -~-- . - - -- . __J.--'
0.. 0 L ---·- -·- .. _j ·,¥, --- -· _ _J ·-·-- --·-··---. _j -- _ ---- __j_ -------- -··- -- ... _, ____ ---·-- ---··· l_________________ J______ ···--·---1- .... . ... ---JUL 31 AUG 1 AUG 2 AUG 3 AUG 4 AUG 5 AUG 6 AUG 7 AUG 8
DAY
c================-c,r.a=c;c~~ ~ ---=--~•, ~-- -
------
------
~ w p,
Cl')
0
H 0
G----t
a ,_q P-i P-i
('I')
E , ....
m :J
6,--
5~
I I l
4l-I 1 i l l
31r
2'r-,
11-!
, 0 L
0
o••••••••••••
-·-·-·-·-
_______ J __________1 _______ 2 4
Figure 23 Diurnal Variations in Mean Pollutant -C-oncentrations (San Jose)
----7
FINE SULFATE FINE NITRATE FINE Ce OZONE
----------- J_ ·-·------···--.l.-
6 8
! ' '
0 • ~ 0 0 O O O O O D 0 O O O O O O O O O e 1
.r---•--·---·~---·~ ,,,. ..... ✓/ ~0
/' ·.. . -~ ✓ -~
/ /. ·---~.,- - ,..,..:-. - - -- ' .. ·':·' ' ~------- -~. __ /· -· ------ ----......_ . . '-
/ ~--.' ·--.•.-........
·--,._ '·--,..________.------=------'-- - -- --
l ___________J_ ____________L ___________l__________j -----------------1- _---------'------ _____ j_____________j
10 12 14 16 18 20 22 24
TIME OF DAY _..-----~.J•=~·-- -
figure ~4 uiyrnal ·Variations in Mean Po11utcint Concentrat1ons (Riverside)
RIVERSIDE20,. ·- ----- ---7
18~
! ~
16 1
r(Y) \
0
H 0 14r
G--i
i! Pi
12 L-
(1')
E J::,. w '--. CD rn
:)
10r 1
!•
8 ~-! ' ;
6 ~-!
4 l-
2~-
! i
0 L .. ________ 0
FINE SULFATE FINE NITRATE FINE Ce OZONE
e e • • • • e • • o • • • o • • e o o • •
: .................... .
-...,..1··"" / ./ '·-' .
/ '-/ -.......
/ ,/ ·,~ ~I ... •. '.~
. ·,•..' I ..
I / ' .
,.✓ . ,_
·,,_ ---------- -- --~---~ ------·//.----~-...____......- .......---··-------~\_
\....__
I "'"- ....""°'
\
\, •✓- ·,
/
,/ ,- --- --- --- '\_ ··~
~ -- ....____ \ '·- - ---·/ ......... ,._
~ ~
✓--/ ...._• - -- - -- -- ~. -- - -- -~ -·--~✓•
.F
--~------·---•--·-·✓
J ____________ i ___ . ____ L .. ___ 1.. ________ L_ _J -- -- _----- _. L ____________ .L . ______ .__ L _____ ... ___ ..J ___________ J___ ___ ___ __j 2 4 6 8 10 12 14 16 18 20 22 24
TIME OF DAY
,...------,
Figure 25 Diurnal Variations in Mean.Pollutant Concentratiorii (Los Ange,les)
r-- /A.'\_ . ---------~ --·-- --· ---~-· .. ·------· --··· -- . -- --- - -·----,.
l
16 ! '-,
r FINE SULFATE
,.,,-·
14 t-· I
FINE NITRATE FINE Ce
. ./-·
1/. "\,_
·,'\ __
' I OZONE ,, (Y)
0
H 0
G--t
i 12~
j ·,
......... .....
.§ Pi Pi 10~
... _ ',.
(T} ! ....'-....
E .i::. w '-----.. Ha m
:.) 8 ~-
./'-,.
BL.
I
' / .
,/
/ '- '- .,(.✓
,I' '·--... •••........__ · .. ......... ....... ·. ....... ·.
'·· .. ~'"'(_· •••••••
4~ i
2~ i
/./
,.
/✓ 0 '- '- '-
'-
-.... -..... -.... ..._ -..... .__ -··
'·
-•··
'· '-,·,., ' -- ---"• ..._
- ",
i
0 l .. __ j _____ . J - - J ______________L ___ _ -· .1 ··--· . - ... j .... I . .L . _________ L ----- ___ .L . . -.. J ___ . . ·--I
_J 0 2 4 6 8 10 12 14 16 18 20 22 24
TIME OF DAY ::.b-----....:J:..L..i_.,_ ~ ...w........ ........._____,
'F!i,.gure 26
MASS COMPARISONS 1982 ARB VISIBILITY STUDY
lSfJ r I /
i / /
125~ / I
I ~ / +} 1ool
+
+ +
+ ++ ♦
+
+
~I
U1 I
(/) I I'✓ ' ,< j I e-· ~
t4 75i I ,W i /
_J i t.J t
• I +t-1 ii-0::: +ssL /
I
t.0*•I + +
CL < I I+ + + +\
... .,,.,*♦ ..._._I ♦ ♦ .,,At, + -F ♦ ....,.w z / ++# ♦ t--t + 1/ .....+ ...
25! / / ~,.LL r /+ ~·· + •♦ ♦~-;It♦ +I+ r + +,.J;>'+ +
( / +.,,..... 0 ~-1+ I I __l-~-~- I I .. J_.
0 25 S0 75 ll!e 125 15S 175 200
TOTAL PARTICULATE MASS (ug/m~
NASS1V2SYMBOL•+ LINETYPE•---f;.i~">< as:e. 5417 ••s:3. 1600 b-8.4735 e,,a:8."317 •,..•13. 6868 ra=e. 8242
( r = 0.82). Higher correlation cannot generally be exnected since
the proportion of coarse particles can vary with meteorological
conditions. On average, the fine fraction represented about half
of the total particulate mass.
The proportion of the total sulfate which was in the fine fraction
was assessed in similar manner (Figure Z7) • More than 90% of the
total sulfate collected with Teflon filters was in the fine
fraction. Thus either fine or total -particulate sulfate could be
employed for data analysis in assessing contributors to light
scattering. With nitrate, however, a substantially larger fraction
of coarse nitrate was observed. Linear regression between fine and
total particulate nitrate yielded the equation:
Fine N03- • -0.625 + o.620 (Total N03-)
m 0.917
Thus nearly 40~ of the nitrate was coarse material.
Samples collected on pre-fired Pallflex quartz 2500QAO fil tars
following a cyclone (outpoint about 2.5 um) were compared for total
carbon retained to that on pre-fired glass fiber fil tars using a
conventional hi-vol sampler. The face velocities for the two
units were nearly identical (49 cm/sec for the fine particle
sampler, 47 cm/sec for the hi-vol). Assuming no significant
difference in carbon retention by the quartz and glass fiber
fil tars, differences reflect directly the proportion of carbon in
the fine fraction.
For ten paired samples collected in Riverside and downtown Los
Angeles the fine fraction carbon (ug/m3) averaged 62 ± 13% of that
in the· total particulate (hi-vol) sample. Linear regression
yielded the result:
Fine C = -3.93 + 0.789 (Total C)
- 44 -
Figure 27 Comparison of Fine and Total Particulate Sulfate
···--····•~
40_
+ s. J. ◊ RV
35_ 0 L.A. f;;' E ...,~ en 30_ :)
'-' t .,..
a 25_tn
, CJ.,_,.. 20_ f--0:: < CL 15-.. w z ........ LL 10_
y = 0.691 + 0.929x r = 0.995
5_
.0 J,t
__ ... .1 J . J. 1 l l l J
!
?t:0 5 10 15 20 b..,J 30 35 40 45
D RTT~i _ TOT f.L. , A ~ c. S04 (· ·o/m-r,~-; I. ,,
.,,,,.,.,
- 44a -
The results for percent fine carbon compare to the range 56 to 66%
obtained with dichotomous samplers at three sites.
Precedin~ results suggested that the proportion of coarse particle
carbon obtained with dichotomous sampler may be erroneously high
because of the relatively low face velocity for the coarse particle
filter (34). The present data fail to support this. The
proportion of C in the fine fraction can be assessed from Figuree
28 which "?lots fine against total Ce. The results are highly
correlated and indicate about 20% of the elemental carbon to be in
the coarse fraction. The fine Ce is compared to the total carbon,
measured by combustion in Figure 29. The results suggest, on
average, about 18% of the total carbon to be elemental carbon.
Figures 30 and 31 test linear regressions between fine particle C ,e
fine and total suspended particle mass, respectively. The moderate
correlations observed (r = 0.69 - 0.73) suggests that elemental
carbon represented about 6% of the fine particle mass.
Eo Calculated vs. Observed Ammonium Ion Levels
If one assumes all fine particle S0 4 = to be ( NH 4 ) 2S0 4 and true,
fine particle nitrate to be NH'4,N'03, then expected concentrations of
NH 4 + can be calculated. Regression equations relating observed and
calculated NH 4 values are as follows:
San Jose Cale. NH4 +
= 1.66 + 1.77 (obs. NH 4+) r = 0.81
Riverside Cale. NHt+ + = 2. 56 + o. 96 (obs. NH 4
+) r = 0.96
+ +Los Angeles Cale. NH4 = 3.92 + 0.98 (obs. NHt+ ) r = O. 95
In all cases, calculated values exceed observed values. At
Riverside and Los Angeles where high correlations were observed,
the difference was relatively constant at 2.5 to 4 µg/m 3 • This
- 45 -
1--== ~-:,.
Figure 28 Comparison of Fine and Total Particulate Elemental Carbon
12 -----~-- ------· --·---·------------ ....... --- ---------------·----·-
11
/"',.. (T)
E '-.. m :)
'-.../
c./J'
10
9
8
7
6
5
4 ---
3
2
+ s. J. ◊ RV o L.A.
◊
0
j I
J ...... L L .. L .. L. 1 ... L ... J ... L .. _L .. .L.... L ...L_J 2 3 4 5 6 7 8 9 10 11 12 13 14 15
CARBON SYMBOL=·
I LINETYPE= ---------it.W y=a+b*xg1 i--I < n=l06
_J a=0.0441 s.==0. 0421 :::J b=0. 7977 s-.==0. 0109u 81-•c0. 2685 r~0.9904 ~
t-c~ < o_
LLJ z ..........« LL
TOTAL PARTICULATE Ce (ug/m~
Figure 29 Fine Elemental Carbon vs. Total Carbon in 4-Hour Samples
VISIBILITY STUDY ,·-·-- ·--··-··-----•-·•-·· -•-·· ----
0
1 ·7l 10~
,-·I + s. J. .-,·'
◊ RV /
I /o L.A. /
l l ! ,.
/
~~·...
~ ~"1(,'i, /
~'i,v, ,/CD ), /L
~o,,V-'o, /'·, /C) /=1 ./ '--' ✓•
~ 0, 6~ / tr I
llJ u lJJ z 4t----, LL
/ .,..... +
/ /
-o,..,. /~ .A ...e (:)V.• E) y = -0.924 + 0~1aox.,...,,5 + + e
r = 0.84 0 / ee/◊ -,,,.• ❖ +
,..--l:) ◊ + e e / ◊◊ ◊
/ ◊:~~◊◊ 2 ~- / e 1¢,
✓• ◊ +·- + s ◊
◊ ◊
i /+
0 L- .,.¥..~~ .._ ~----- ◊. ...1___________________ ...L--'- ____________ - __, - j__________________ -------·-·•·-.J~ 0 10 20 30 40 so
TOTAL CARBON (UG/M3)
'""I
Figure 30 Fine Particle Mass vs. Fine Particle Elemental Carbon
--- --- ··- --------- ----A -------~- --·--·-----·--··-·:-
11 +
10 + I ·------ -·•- --- ----- ------------------I I
9 +
8 + I
---- ---- . -- ·----- ----- ··- ---------------
7 +
6..,..Fine C e
I
Iii::. (µg/m3} . I -l.11 () 5 +
4 + I•··· -
3 +
•I 2 +
II-.
A l +
. - ----·v•AA~AA
- ---- --
A A A
-- .
A
A
A A
A A
A A
--- - ------------- A ---------- ---------------~------
A
A ---·
A
B y = 0.008 + 0.0568x
A AB A A . A----A------- -------- -------•---------------------A----
A · - - r = 0 .69
___.. -A---------------------·--------------------
A A A A A
A BB A .. - AA- --- -------A--------------------·--------------··-····-------------A A
A A A
I ... ~BA~B AI A
0 ·---- --- ----- A - - ··-·- . /.. ---- -- -- -- --- - ...
-+-------+-------+-------+-------➔ -------+-------+-------+-------+-------+-------+-------+-------+-------+-----~ O 10 20 30 40 50 60 70 60 90 100 110 120 130
Fine Particle Mass (µg/m 3)
Figure 31 Total Particle Mass vs. Fine Particle Elemental Carbon
A I.. -----------------------A-----
~o•I
A
-11 +
9 +
ti Pi
Fine C e
(µg/m3)
El
7
6
~
4
3
2
1
0
+ I
I... +
A
+ --
-----A---·-- - -- -•. - -----------+
A -- -----------A------------- ---- -----
A A A + A BA
- -·----- . A - -A
A A + ----- - -- --- -------------A------¥----~O..-S94--i---0....-O-l3-7-x----- --- --
A AAB A A A A r = 0. 732A
A - --B ... A--A--AA-- --A- -- . --- A-------------. --- --- - - ---- --- --- -- --- --- -· .. + A A
A.~ A A A A - -- A-.-----·-----------···---- -----A------·•·· ----------- -------------·-··--
A A + A AA
A A
A +-
-+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+---O 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210
Total Particulate Mass (µg/m 3)
(
difference may be indicative of acidic aerosols although other
cations besides NH4+ and H+ may be significant.
F. True ~article Carbon (PC) and Total Low Volatility Carbon (LVC)
Figure 32 compares total results for quartz filter samples
collected with a preceding Al 2o3 -coated denuder (sampler 4)
against total carbon results without the denuder. The results show
about 17% less carbon on the filter with the denuder, which might
infer that a corresponding amount of carbon was retained in the
denuder of sam-pler 4. However, this assumes that the presence of
the denuder does not influence the loss of carbon from the filter
by volatilization. If the denuder does remove vapor-phase,
material in equilibrium with material in particle phase, then
volatilization could be enhanced, decreasing retention of carbon on
the filter.
As discussed in Appendix F, water retained bv the A.1 203 and its
subsequent elution with CH2Cl2 - MeOH appeared to g.reatly decrease
the precision of the subsequent carbon measurements. In contrast to
prior work using CH 2Cl 2 alone for elution (40) tests of the
consistency of these results with those from other carbon
measurements have caused us to reject these data from further
analysis. Accordingly, no PC and LVC data can be considered.
G. Atmospheric Nitric Acid
I t has been prop o s e d and demons t rat e d ( 5 3 , 5 4 ) t h a t t h e
concentrations of atmospheric HN0 3 and NH 3 can be influenced by the
equilibrium:
The dissociation constant is strongly dependent on temperature. At
relative humidities below that needed to cause NH 4N0 3 to deliquesce,
- 46 -
Cua/ m3)
Figure 32 Comparison of Filter Collected Carbon With and Without 30 ·--· ··-- ___ :_______ Di_~~~~ i o~- ~~~-u~~r _!o~-~~-~~on_a._~-~-o~s _y a~~_rs _.. .. .. . ... .. _
I I-
a ___ _
-
( -;;" e
' m ::) ~
25_
~
20_
+ ◊ 0
s. J. RV L.A.
◊
A
'°'llJ I
r""Y'.... w r-·-_j,__. LL
~
15_
~
a: w _J CL ~ ·<v'1 r--lJ
... 5_
... y
r = =
0.691 0.995
+ 0.929x
J.. l . 1 L L. l 1a
l J i. L,. 15
1 ! • 1. 2C
l l . 1 1 i L.i 32
..,_J
- 46a -
a knowledge of the temperature and corresponding dissociation
constant, together with the ~H3 level, should permit calculation of
the corresponding equilibrium values for the HNO 3• Employing
reported values for the temperature dependent dissociation constant
(54), and 4-hour average measurements of NH 3 and temperature,
observed and calculated HN03 were compared. Observed HN0 3 values at
San Jose, Riverside and Los Angeles, measured by the denuder
difference technique (21), ranged from 0.25 to 8.5, 0.5 to 16.5 and
1 to 60 µg/m 3, respectively. However, only at San .Jose was a
significant correlation observed (r =- 0.73) between calculated and
observed values. The corresponding regression equation was:
- 47 -