EPA-650/4-73-002
October 1973
ENVIRONMENTAL MONITORING SERIES
-------
-------
EPA-650/4-73-002
LIDAR STUDIES
OF STACK PLUMES
IN RURAL
AND URBAN ENVIRONMENTS
by
Warren B . Johnson, Jr., Robert J . Allen, and William E. Evans
Stanford Research Institute
333 Ravenswood Avenue
Menlo Park, California 94025
Contract No. CPA 70-49
Project Element No. 1AA009
EPA Project Officer: Francis Pooler, Jr.
Meteorology Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
October 1973
-------
This report has been reviewed by the Environmental Protection Agency and
approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Agency, nor does
mention of trade names or commercial products constitute endorsement
or recommendation for use .
11
-------
BI3LIOCRAPHIC DATA
SHEET
1. Report No.
EPA-650/4-73-002
3. Recipient's Accession No.
4. Title and Subtitle
LIDAR STUDIES OF PLUMES IN RURAL AND URBAN ENVIRONMENTS
5- Report Date
October 1973
6.
7. Author(i)
Warren B. Johnson, Robert J. Allen, and William E. Evans
8- Performing Organization Kept.
No" 8509
9. Performing Organization Name and Address
Stanford Research Institute
333 Ravenswood Avenue
Menlo Park, California 94025
10. Pro|ect/Task/Work Unit No.
1AA009
11. Contract/Grant No.
EPA CPA 70-49
12. Sponsoring Organization Name and Addreaa
Meteorology Laboratory
National Environmental Research Center
Research Triangle Park, N.C. 27711
13. Type of Report It Period
Covered
Final
14.
IS. Supplementary Notes
16. Abstracts
Experimental results are presented from field studies of smoke plume diffusion and pollution-layer structure
in both rural and urban areas, using the Mark VIII mobile lldar (laser radar) system. This new system was
first applied to study the behavior of smoke plumes from the 250-m stacks of a large coal-burning power plant,
the Homer City Generating Station, in western Pennsylvania. Examples from the Homer City study of character-
istic changes in plume diffusion and low-level aerosol structure resulting from time-varying meteorological
conditions are presented in the fora of vertical plume cross sections. Helicopter-measured 802 cross sections
and the lidar-obtained (moke cross sections are compared on a case-study basis. Estimates of diffusion
parameters derived from the lldar observations are compared with those used in a fluctuating plume diffusion
model.
The mobile lidar observations in urban areas (San Jose, California, and St. louis, Missouri) reveal significant
variabilities in the pollution-layer structure associated with urban effects, transitional meteorological con-
ditions, and apparent convectlve Influences.
An analysis of the eye-safety aspecta of lldar use in the atmosphere is also Included. It is pointed out that
the probability of eye damage from lidar usage under current operational procedures is extremely small. How-
ever, remedial suggestions are offered for reducing the eye hazard even further.
17. Key Words and Document Analysis. 17o. Descriptors
Lidar
Laser radar
Plume dispersion
Mixing-layer structure
Diffusion modeling
Air pollution meteorology
17k. Identificrs/Open-Ended Terms
17c. COSATI Field/Group
It. Availability Statement
Unlimited
19. Security Class (This
Report)
UNCLASSIFIED
20. Security Class (This
Page
\JNCLASS1F1ED
21. No. of Pages
109
22. Price
POMM NTIS-»t INEV.
USCOMM-OC t4fS2-P7J
-------
CONTENTS
LIST OF ILLUSTRATIONS vii
LIST OF TABLES ix
ACKNOWLEDGMENTS xi
I INTRODUCTION 1
A. Background 1
B. Nature and Objectives of This Study 2
II LIDAR TECHNIQUES AND INSTRUMENTATION 5
A. Techniques 5
B. Instrumentation 7
III HOMER CITY STACK PLUME EXPERIMENT 11
A. Objectives 11
B. Experimental Techniques 11
C. Data Summary 15
D. Observed Plume Characteristics 15
1. Plume Evolution with Time During Transitional
Periods 17
2. Time-Averaged Cross Sections 21
3. Plume Trapping: Relation to Ambient
Atmospheric Stratification 23
4. Plume Condensation 26
E. Estimates of Plume Dispersion 29
F. Comparison of Lidar Data with Plume SO2 Cross
Sections Observed by Helicopter 32
IV SAN JOSE URBAN POLLUTION-LAYER TEST 37
V ST. LOUIS URBAN POLLUTION-LAYER EXPERIMENT 43
A. Objectives 43
B. Experimental Techniques 44
-------
C. Data Summary 45
D. Results 45
VI RELEVANT SUBSEQUENT LIDAR APPLICATIONS 55
A. Diurnal Variation of the St. Louis Mixing Layer
(NSF/SRI METROMEX Projects) 55
B. Spatial and Temporal Variations of the Los Angeles
Mixing Layer (EPA Project) 57
VII CONCLUDING REMARKS 61
REFERENCES 63
APPENDIX A—Mark VIII Litiar System Details: Reprint of "Laser
Radar (LIDAR) for Mapping Aerosol Structure,"
by R. J. Allen and W. E. Evans A-l
APPENDIX B—Lidar Data Summary, Homer City Experiment B-l
APPENDIX C—Lidar Data Summary, St. Louis Experiment C-l
APPENDIX D—Design Considerations for an Eye-Safe Lidar
System D-l
VI
-------
ILLUSTRATIONS
1 SRI/EPA Mark VIII Lidar System Installed in Van for
May 1970 Study of Diffusion of Stack Plumes from
Large Power Plants in Western Pennsylvania
2 Locations of Lidar Sites Used for Observations of
the Stack Plume from the Homer City Generating
Station 13
3 Schematic Plan View of Typical Lidar Observational
Technique for Plume Studies 14
4 Time Series of Smoke Plume Cross Sections observed
by Mark VIII Lidar 2.5 km Downwind of 250-m Stack
of Homer City Generating Station, Pennsylvania on
9 May 1970 18
5 Time Series of Vertical Plume Cross Sections on
9 May 1970, Except for Upper Left Photo that is a
Zenith Scan Through the Haze Layer 20
6 Quasi-instantaneous (15 to 30 sec) Plume Cross
Sections (top 10 photos) on 11 May 1970 and Corre-
sponding 45-min Composite ("average") Cross Sections
(bottom) at Downwind Distances of 1 km (left) and
2 km (right) from the Stack 22
7 Composite Cross Sections Showing Plume Trapping at
1500 m Height 24
8 Example of the Trapping of a Plume Rising in a
Turbid Ambient Haze Layer 25
9 Time Evolution of Plume Structure at 2.5 km
Downwind Within a Multilayered Haze Structure 27
10 Example of Cloud Formation Within the Smoke Plume
3 km Downwind from Stack 28
11 Comparison of Helicopter SO2-(left) and Lidar
Particulate-(right) Cross Sections for Two Cases
During Homer City Experiment 34
12 Three of the Helicopter SO2 Cross Sections Used for
the Plume-Height Comparison in Table III-4 . . 35
VII
-------
13 Height/Distance (Photos A, B, and C) and Height/Time
(Photo D) Cross Sections Obtained During Special
Tests of the Modified Mark VIII Lidar System in the
Menlo Park, California Area 38
14 Lidar-Observed Time/Height Cross Sections of the
Urban Haze Layer over San Jose, California on
11 December 1970 . 39
15 St. Louis Lidar Route and Checkpoints 46
16 Time Changes in Pollution-Layer Structure Approxi-
mately Along Wind Through East St. Louis 49
17 Vertical Pollution-Layer Structure in an Approxi-
mately Crosswind Direction Along 1-70 in St. Louis 51
18 Vertical Pollution-Layer Structure in the Downwind
Direction Along 1-55 from the Arch in Central
St. Louis 52
19 Vertical Pollution-Layer Structure in a West-East
Direction Along 1-270 Downwind of Central St. Louis .... 54
20 Height/Time Cross Section of the Aerosol Structure
over St. Louis on 13 August 1971 as Observed by
SRI/EPA Mark VIII Lidar System 56
21 Time Evolution of the Los Angeles Aerosol (Smog)
Layer as Observed During an Early Afternoon Period
by the Mark VIII Lidar over a South-North Route on
the East Side of the Basin 58
22 Location/Time Contours of Lidar-Observed Aerosol
Mixing Height (in Meters) Along a 70-Mile Freeway Loop
over the Los Angeles Basin on 21 September 1971 59
23 Mobile Lidar Routes, Los Angeles Basin, September 1972 ... 60
Vlll
-------
TABLES
1 Characteristics of the Mark VIII Mobile Lidar
System
2 Chronological Summary: Homer City Lidar
Experiment 16
3 Power Plant Precipitator Operation During the
Homer City Lidar Experiment 17
4 Comparison of Observed and Calculated Values of
Plume Radius-to-Dispersion Ratio (R/a ) 31
5 Comparison of Plume Heights (above stack base)
Obtained from Helicopter SO2 Cross Sections and
Lidar Particulate Cross Sections 33
6 Comparison of Lidar-Observed Haze-Layer Tops over
San Jose, California, with Mixing-Depth Estimates
from Helicopter Profile Data and a Mixing-Depth
Model 41
7 Data Summary: St. Louis Lidar Experiment 47
IX
-------
ACKNOWLEDGMENTS
The authors would like to give credit to a number of individuals
who have made significant contributions to this project. With regard to
personnel at Stanford Research Institute, Mr. Norman Nielsen played a
key role throughout the course of the work. He helped to design the
Mark VIII lidar system and was in charge of fabrication and testing of
the system. In addition, he participated extensively in each of the
field experiments. Mr. Russell Wolfram designed and supervised construc-
tion of the range-compensation module of the lidar system. Mr. Albert
Smith participated in two of the experiments and assisted in the data
analysis, as did Miss Joyce Keoloha. Dr. Walter Dabberdt conducted the
comparison of the lidar and helicopter mixing-layer data from the San
Jose tests. Mr. Ronald Collis furnished supervision and support throughout
the course of the project.
We also wish to express appreciation for the technical and financial
support furnished on behalf of the sponsoring agency by Dr. Francis Pooler
and Mr. Lawrence Niemeyer of the Meteorology Laboratory, U.S. Environmental
Protection Agency. In addition, the help of Mr. Frank Schiermeier of the
same group in furnishing data from other elements of the LAPPES program
was particularly valuable.
xi
-------
I INTRODUCTION
A. Background
Prediction of near-ground pollutant concentrations is a central
factor in air pollution control and air resource management. In carrying
out this role, however, the air pollution meteorologist and chemist must
deal with the fact that pollutants move, disperse, and change character
in a three-dimensional, dynamically open atmospheric system. Achievement
of a fuller understanding of these processes, as needed to develop
techniques for accurately predicting near-gound concentrations, requires
that detailed observations be obtained in three dimensions within the
atmospheric medium. In this regard, one of the classic problems has
been to obtain detailed measurements in the vertical. Such measurements
have been particularly needed in studies of diffusion from elevated
sources such as tall stacks.
Consequently, increasing attention has been given to the capabilities
of atmospheric indirect sensing techniques for supplementing conventional
direct sounding systems that use balloons or instrumented aircraft. A
variety of types of remote probing systems are under active development,
including acoustic sounders (Shaw, 1972; McAllister, 1972; Hall, 1972;
Little, 1972; Uthe and Shaw, 1973), passive radiometric systems (Hosier
and Lemmons, 1972), and various laser techniques. Of the latter, lidar
(laser radar) has proven to be especially promising. Lidar applications
in air pollution studies have been discussed extensively in the literature
(Johnson, 1969a ; Collis, 1970; Barrett and Ben-Dov, 1967; Hamilton, 1966,
1967, 1969). The advantages of the lidar technique basically stem from
its ability to obtain useful measurements remotely and at a high density
in space and time.
1
-------
This report describes the results of an observational program that
were obtained with the van-mounted Mark VIII mobile lidar system developed
by Stanford Research Institute (SRI) with partial support from the U.S.
Environmental Protection Agency (EPA) under this contract. This lidar
system was designed particularly for air pollution applications and has
a number of advanced features that are described in Section II and in
Appendix A.
B. Nature and Objectives of This Study
The observational program included lidar experiments at one rural and
two urban sites;
Homer City Generating Station (HCGS), Pennsylvania (rural)
San Jose, California (urban)
St. Louis, Missouri (urban)
The Pennsylvania experiment was the major part of the study and entailed
lidar measurements of smoke plumes from the 250-m stacks of a coal-burning
power plant (the HCGS). Results from a similar lidar study of nearby
Keystone Generating Station have been previously reported (Johnson, 1969b;
Johnson and Uthe, 1971). The Keystone and Homer City investigations have
been carried out in conjunction with a coordinated EPA program called the
Large Power Plant Effluent Study (LAPPES). As a part of this program,
ground and airborne measurements of sulfur dioxide (SC^) concentrations
and wind and temperature profiles were taken quasi-concurrently with the
lidar observations. The LAPPES data collected during the several experi-
mental periods have been summarized by Schiermeier and Niemeyer (1970)
and Schiermeier (1970, 1972a, 1972b); preliminary analyses have been re-
ported by Pooler and Niemeyer (1970). One of the specific aims of the
LAPPES was to develop and validate transport and diffusion models for
calculating ground-level concentrations of effluents from large power
-------
plants with tall stacks, since the results from previous studies of
plumes from short stacks cannot simply be extrapolated with any degree
of reliability. To accomplish this goal requires a better understanding
of plume behavior; therefore, this was the main purpose of our lidar
studies of the smoke plumes from these tall stacks. Although particulate
emission is not the major air pollution problem associated with power
plants (because of the efficiency of modern fly-ash removal devices),
the smoke serves as a convenient tracer for study of diffusion processes.
Because the particles are small, their diffusion should be essentially
identical with that of a gas such as SO > the primary pollutant of
^
interest.
The urban studies in San Jose and St. Louis were conducted to test
the feasibility of a lidar system for mapping the three-dimensional struc-
ture of the urban haze (pollution) layer, including plumes within this
layer. Such information is needed to aid in the further development of
air quality simulation models for urban areas.
-------
II LIDAR TECHNIQUES AND INSTRUMENTATION
A. Techniques
The Mark VIII lidar system is basically composed of a laser trans-
mitter that emits a very brief, high-intensity pulse of coherent light,
and a receiver that detects the energy at the laser wavelength back-
scattered from the atmospheric aerosol as a function of range. The fac-
tors affecting the strength of this received signal are best discussed
in terms of the lidar equation, which may be written in the form:
2 r R l
P (R) = K E R~ p(R) exp - 2 J cr(r) dr , (2.1)
r L o J
where
P = power collected (at a given instant) by the primary
receiver optics from atmospheric backscatter of laser
radiation
R = range
K = constant based on lidar characteristics
E = energy transmitted into the atmosphere
a = volume extinction coefficient
0 = volume backscatter coefficient.
This relation assumes a constant energy density across the beam, randomly
distributed scatterers within the effective scattering volume, Bouguer's
law of attenuation, and neglects multiple scattering effects. These
assumptions are normally valid because of the very narrow field of view
of lidar systems 0^1 mrad).
-------
Equation (2.1) shows that the lidar signature is dependent on the
range distribution of two optical parameters: the volume backscatter
coefficient (3) and the volume extinction coefficient (a). The exponen-
tial term containing Q represents the two-way transmittance to the back-
scattering volume, and it can be seen that |3 and a act in opposite direc-
tions on Pr. The backscatter coefficient is a function of the particle
number concentration, size distribution, and optical properties, as is
the extinction coefficient. However, since p enters into Eq. (2.1) as a
direct factor (rather than in an integral expression, as does a), it is
apparent that rapid changes in the received signal can be ascribed pri-
marily to p.
When atmospheric attenuation can be ignored, as in the case of a
relatively clear atmosphere (Uthe and Johnson, 1971), the lidar back-
scatter signature provides a rapid means of qualitatively describing the
distribution of atmospheric particles along an observing path. Multiple
lidar observations can then be assembled to provide knowledge of the
spatial or temporal variations of aerosol structure.
With the Mie scattering theory, quantitative mass concentration es-
timates can be derived from lidar data if the size distribution and op-
tical characteristics of the aerosols are known or assumed. This has
been successfully demonstrated but is a rather complex procedure (Johnson
and Uthe, 1971).
The variation in optical scattering properties may be the result of
either changes in the particle number density, or changes in particle
size distribution, or optical properties of the particles. Although this
nonuniqueness exists in single wavelength lidar measurements, lidar
observations of particulate plumes or layers show boundaries and varia-
tions of meteorological interest regardless of the physical or chemical
nature of the aerosol. The relative optical density of particulates as
-------
observed by the lidar technique has been related to meteorological fea-
tures by a number of investigators (Johnson, 1971; Johnson and Uthe, 1971;
Uthe and Johnson, 1971; Gambling and Bartusek, 1971; Olsson et al., 1971;
McCormick and Fuller, 1971; Viezee and Oblanas, 1969).
In lidar studies of smoke plumes or urban pollution layers, a fre-
quently observed relationship is that between the vertical structure of
temperature and aerosol density. Atmospheric stability associated with
a temperature inversion normally limits convective penetration of aero-
sols across the inversion boundary; hence, strong vertical aerosol gra-
dients that are easily identified on a lidar backscatter signature may
form at the top of the mixing layer. The height of the mixing layer is
one of the more important parameters in modeling the diffusion and trans-
port of pollutants of near-surface origin. Pollutants are often effec-
tively trapped beneath an elevated inversion, resulting in a greater aero-
sol density within the mixing layer. Experimental results along these
lines are presented in Sections III, IV, V, and VI of this report.
B. Instrumentation
The Mark VIII lidar system in its initial form is shown in Figure 1,
while Table 1 lists the characteristics of the system. A more detailed
*
description is furnished in Appendix A. The Mark VIII and Mark IX
lidars are unique for their capability of video disc recording of large
quantities of lidar data and rapid electronic processing of the data to
obtain range-corrected, intensity-modulated CRT displays of particulate
distributions in real-time. Also, these lidars use coaxial transmitter
and receiver optics to minimize geometric effects in the backscatter
*
The SRI Mark IX mobile lidar system, which is a similar but improved
version of the Mark VIII, was recently designed and built by SRI with
internal funding.
-------
LU
? H
O
r-.
05
<
in
DC
Q <
LU -I
to o
? IT
U-
LU m
>i
DC _
IJ h-
to
§s
< t: >-
CL _ tO
LU Q Z
"^~ 7
DC "- UJ
to O Q.
DC
D
-------
Table 1
CHARACTERISTICS OF THE MARK VIII MOBILE LIDAR SYSTEM
Characteristics
Description
Transmitter
Laser rod
Wavelength (A)
Beamwidth (mrad)
Optics
Pulse energy (joules)
Pulse length (nsec)
Peak power (MW)
Q-switch
Maximum PRF (pulses/min)
Cavity cooling
Receiver
Optics
Field of view (mrad)
Predetection filter passband
width (A)
Detector
Frequency response
Automatic gain compensation
Signal conditioning
Scanning and firing
Elevation scan (degrees/shot)
(mode 1)
Azimuth scan
Firing rate (shots/min) (mode 2)
Mobile capability (Mode 3)
Firing density (shots/mile)
Vehicle speed while firing (mi/hr)
Data recording and display
Recording and storage
Direct recording
Displays
Ruby (3/8 x 3 inches)
6943.0 ± 0.4
0.8
Lens (2-inch), coaxial with receiver
telescope
0.9
30
30
Pockels cell
30
Refrigerated water
Newtonian reflector (6-inch)
Adjustable, 1 to 5
10
RCA-7265 PMT (S-20 photocathode)
Direct writing: >100mHz
Recording: 4.5 tnHz
Inverse range-squared gain control of PMT
Log or linear amplification
Motor driven and selectable between
0.5 and 10
Motor driven
Selectable, 1 to 30
Selectable, 10 or 20 (controlled by
vehicle odometer)
Up to 60
Magnetic video disk recorder
CRT oscilloscope photographs
Intensity-modulated range height indi-
cator (RHI), height/time or height/
distance, cross-sectional displays;
also conventional A-scope (amplitude
versus time)
-------
signature, which permits valid data to be obtained at a shorter distance
from the lidar than is possible with bistatic systems.
As indicated in Table 1, three basic modes of lidar operation are
possible with the Mark VIII system:
• Mode 1—Van stationary; programmed lidar elevation scan. This
produces a radar-type intensity-modulated, RHI fan-shaped dis-
play on an oscilloscope.
• Mode 2—Van stationary; lidar firing at programmed increments
of time while pointing upward (or in any other direction). This
produces an intensity-modulated, height/time cross-sectional
display.
• Mode 3—Van moving; lidar firing at programmed increments of
distance while pointing upward. This produces an intensity-
modulated, height/distance cross-sectional display. Large
areas can be mapped in this fashion.
For this study, successful field operations with the system were con-
ducted under Modes 1, 2, and 3 at Homer City, San Jose, and St. Louis,
respectively.
During the plume experiment at Homer City, the fast firing rate, in
conjunction with the automatic elevation scanning feature, permitted de-
tailed cross-sectional scans of most plumes to be completed within one
minute. In addition, the RHI display capability has been an important
advantage. It not only eliminates the need for costly and time-consuming
digital calculations but also furnishes a quick look at the data in the
field for experimental control in real-time. The resulting vertical
cross sections can be analyzed individually or as time-average composites
through multiple displays and photographic exposures. Hamilton (1969)
has also carried out lidar plume studies along these lines, but with
less advanced instrumentation.
10
-------
Ill HOMER CITY STACK PIAME EXPERIMENT
A. Objectives
As mentioned in Section I, the objectives of this experiment were
to:
. Test the improved capabilities of the newly developed
Mark VIII lidar system for plume studies
. Use the lidar data to investigate the characteristics of the
behavior and diffusion of the smoke plume from the Homer
City Generating Station (HCGS)
• Compare the lidar plume measurements with SO measurements
from an instrumented helicopter operated by the EPA.
These objectives have been successfully achieved through the results of
the field program, which will be discussed later in this section.
B. Experimental Techniques
The field program was conducted in the vicinity of Homer City,
Pennsylvania, about 80 km northeast of Pittsburgh, during the period
29 April to 15 May 1970. Since at this time the lidar system had not
yet been installed in a van, the equipment was boxed and transported by
air to Pennsylvania, where it was installed on a temporary basis in a
rented van for the duration of the experiment. Two motor-generator sets
mounted in a pick-up truck provided the electrical power for field
operation.
Before the data collection period, a field survey was conducted
within a 10-km radius of the HCGS to locate suitable lidar sites that
Located 5 km SW of Homer City.
11
-------
were accessible and would provide relatively unobstructed viewing of the
HCGS stack plume. Of these predetermined sites, a total of six were
used in the experiment, as indicated in Figure 2. The individual sites
used for each observational period were selected on the basis of the wind
direction prevailing at that time and the resulting plume orientation
relative to the various sites.
The ideal geometric arrangement of plume and lidar for this situa-
tion is to have a horizontal distance, perpendicular to the plume center-
line, of approximately 2 to 5 km between the lidar and the downwind plume
portion of interest, as illustrated in Figure 3. This is because the
basic lidar observational mode used for plume measurements consists of
scanning the plume in the vertical plane, thus producing cross-sectional
displays.
The geometric arrangement mentioned above has the plume close enough
to the lidar to subtend a sufficiently large elevation angle to permit
detailed measurements at 1° elevation angle increments, yet far enough
away so that plume cross sections from elevation scans at other azimuth
angles can also be obtained, as indicated in Figure 3.
During the Homer City experiment, the vertical plume scans were
carried out automatically by the motor-driven, programmable elevation
scan feature of the Mark VIII system. Such scans were usually carried
out at a lidar firing rate of 20 shots/min and at elevation angle incre-
ments of 0.5°, 1°, or 2°. This permitted a plume scan covering a 20° to
90° elevation sector to be completed within 30 sec to 2 min. Occasion-
ally, overhead scans covering a 90° to 135° sector were conducted to
observe the general ambient aerosol structure in the mixing layer.
Where possible, the lidar was positioned at each site so that it
could be aimed in a direction approximately perpendicular to the plume
axis, as in Figure 3. (On a few occasions when the wind direction
12
-------
SITE
38
SITE
20
SITE
40
HOMER CITY
GENERATING
STATION
STACKS
u
FIGURE 2 LOCATIONS OF LIDAR SITES USED FOR OBSERVATIONS
OF THE STACK PLUME FROM THE HOMER CITY
GENERATING STATION
13
-------
LIDAR
ELEVATION
ANGLE SCANS
\
TA-8509-2
FIGURE 3 SCHEMATIC PLAN VIEW OF TYPICAL LIDAR OBSERVATIONAL TECHNIQUE
FOR PLUME STUDIES
14
-------
changed substantially, the lidar van had to be moved to another site.)
Vertical cross sections through the plume were normally obtained at
several azimuth angles and thus at several downwind distances from the
stack so that spatial differences in plume behavior and diffusion could
be documented.
C. Data Summary
A chronological field activities and data summary is given in
Table 2. A more detailed data summary is included in Appendix B.
As may be noted, weather conditions were far from ideal during the
experimental period, with snow occurring on 6 May and rain occurring on
several days. In addition, data were lost in a few cases because of
equipment malfunctions. (This was perhaps to be expected during the first
field trials of a completely new and sophisticated electro-optical system
that had been completely designed and built 'from scratch in less than
four months immediately preceding the experiment.)
Despite these problems, a large amount of high-quality data was
obtained, particularly on 3, 9, 11, and 15 May. These experimental re-
sults will be presented in the next section.
D. Observed Plume Characteristics
In this section, we present a number of the lidar cross sections of
the smoke plume from the HCGS obtained during the May 1970 observational
period. These data have been selected to illustrate certain features of
the plume structure and behavior, some of which are observable only by
means of the lidar technique. Throughout the observational period, the
power plant's stack-gas particulate removal system (using electrostatic
preclpitators) was operated at least at 75% capacity (see Table 3)
so that a very "clean" plume resulted with regard to particulates.
15
-------
Table 2
CHRONOLOGICAL SUMMARY: HOMER CITY LIDAR EXPERIMENT
Date
(1970)
29 Apr
30 Apr
-2 May
3 May
4 May
5 May
6 May
7 May
8 May
9 May
10 May
11 May
12 May
13 May
14 May
15 May
Day
Wed
Thurs
-Sat
Sun
Mon
Tues
Wed
Thurs
Frl
Sat
Sun
Mon
Tues
Wed
Thur
Fri
Site
Number
—
—
36
30
37
20
37
30
37
—
37
36
37
38
38
20
20
20
—
40
36
Observational
Periods (EST)
—
1601-1755
0756-1000
0701-0709
0922-0951
1113-1141
0835-0905
1131-1501
—
0656-1452
1551-1653
0548-0651
0858-1452
0933-1053
1230-1700
1044-1506
0820-0830
1419-1503
1008-1253
1414-1704
*
C.S.
Numbers
—
—
1-8
9-24
25-26
27-33
34-38
39-42
43-69
—
70-153
154-164
165-174
175-217
218-235
236-318
319-391
392-395
396-408
409-443
444-493
Number of
C.S.
—
—
8
16
2
7
5
4
27
—
84
11
10
43
18
83
73
4
13
—
35
50
Remarks
Unpacked equipment
Instrumentation check
and recalibration
Upper inversion apparent
Very light winds, plume
ill-defined; generator
outage
Wind direction changing
Wind direction changing
Plume looping; equipment
malfunction
Weather poor, snow flur-
ries; Advisory Committee
meeting
Demonstration for Advi-
sory Committee during
AM
Laser down for repairs
Upper inversion apparent
Wind direction changing
Rain started 1355 EST
Low ceilings earlier
Low ceilings earlier :
data improperly recorded
due to electronic mal-
function
Low ceilings, light rain
Equipment repaired; rain
started 1605 EST
Poor weather all day
Rain earlier
Very hazy, 2-3 mi vis-
ibility
Total cross-sections recorded 493
Total lidar observations recorded 20,000 (approximate)
C.S.—Cross Section.
16
-------
Table 3
POWER PLANT PRECIPITATOR OPERATION
DURING THE HOMER CITY LIDAR EXPERIMENT
(May 1970)
1
3
4
5
6
7
8
9
10
11
12
13
15
Time
(EST)
0500-1100
All day
0500-1100
0500-1300
0500-1300
0800-1700
0500-1700
0500-1700
All day
0500-1700
0500-1700
0500-1700
0500-1700
Percentage of Precipitator
Sections in Operation
No. 1 Stack
75
100
83
92
Stack off
Stack off
83
83
100
83
83
75
75
No. 2 Stack
100
100
100
92
83
83
Stack off
Stack off
Stack off
Stack off
Stack off
Stack off
Stack off
This plume was frequently virtually invisible to the eye at a short
distance downwind from the plant, but was nevertheless consistently
detected by the lidar at distances of several km.
1. Plume Evolution with Time During Transitional Periods
Figure 4 shows a time sequence of 15 plume cross sections
during the period from 0656 EST to 1003 EST on 9 May 1970. In this
instance, the lidar vertical scan intersects the plume at an angle of
approximately 25° at about 2.5 km downwind of the stack. Because of
this rather oblique viewing angle, the plume cross sections shown in
this series appear about 2.5 times wider than the true plume width at
17
-------
2250-
0656 EST
0811 EST
0919 EST
0946
1003
1500
3000
1500 3000
DISTANCE m
1500
3000
TA-8509-3
FIGURE 4 TIME SERIES OF SMOKE PLUME CROSS SECTIONS OBSERVED BY MARK VIII
LIDAR 2.5 km DOWNWIND OF 250 m STACK OF HOMER CITY GENERATING
STATION, PENNSYLVANIA ON 9 MAY 1970
18
-------
the 2,5-km downwind distance. The power plant stack was located to the
right of the lidar so that the cross sections in the figure are viewed
roughly in the downwind direction.
The sky was clear during the early morning, with an estimated
7 mile visibility and with scattered cumulus clouds forming by 0730 EST.
Surface winds were from the southwest at about 5 mi/hr, increasing to
10 mi/hr by 1000 EST.
The time change in the plume shape is quite dramatic as the
atmospheric stability decreases during the course of the morning. The
plume shape progresses from the typical fanning, tilting shape in the
early morning—caused by the wind veering with height in stable
conditions—to a partial ground fumigation by 0906 EST and full downward
mixing by 0946 EST. Portions of the plume apparently begin to be mixed
downward into the rising convective layer by 0827 EST. During the period
from 0912 to 0935 EST, one area of the plume (on the right side of the
cross sections in Figure 4) retains its identity as its height decreases,
finally reaching the ground at 0935 EST. The plume just after exit from
the stack was observed to be barely visible by eye at 0946 EST, indicat-
ing that the atmosphere dilution had become pronounced.
A related sequence of cross sections is depicted in Figure 5.
This sequence was observed on the same day and during the same time
period as those in the previous figure. However, in this case, the
lidar was pointed at a different azimuth angle, such that it intercepted
the plume axis at an angle of approximately 50°, at a downwind distance
of about 3.5 km from the stack. The plume-width exaggeration in these
cross sections is about 1.3.
19
-------
30 45
RANGE — km
60 75 0
30 45
RANGE — km
60 75
TA-8509-18
FIGURE 5 TIME SERIES OF VERTICAL PLUME CROSS SECTIONS ON 9 MAY 1970,
EXCEPT FOR UPPER LEFT PHOTO, WHICH IS A ZENITH SCAN THROUGH
THE HAZE LAYER
The cross sections are oriented at 55° to the plume axis, 3.5 km downwind
from the stack.
20
-------
Unfortunately, the plume turned out to be closer to the lidar
than was desirable and caused somewhat poorer definition of the plume
structure than in the previous sequence. However, the beginning of the
fumigation at 0914 EST and the rather complete downward mixing at 0938
EST are still readily apparent at this distance from the stack (1 km
farther than in Figure 4).
The first frame in Figure 5 shows the results of a zenith scan
to measure the height of the haze layer in which the plume is embedded,
which turns out to be 1900 m above the lidar. Scattered clouds later
form near this haze top and are apparent in the plume photographs at
0748, 0814, and 0830 EST.
2. Time-Averaged Cross Sections
Figure 6 illustrates the averaging capability of the display
system. The two bottom pictures represent time-averaged plume cross sec-
tions over approximately a 1-hr period, at downwind distances of 1 km
(left) and 2.5 km (right). Five of the six individual quasi-instantaneous
cross sections that were superimposed to develop the average cross sec-
tions are shown above each of the latter. These and other data from
11 May 1970 have been used to estimate plume dispersion as needed for
input to a fluctuating-plume dispersion model, as described later in
this section.
In view of the late-afternoon observational time of these data
and the partly cloudy, light wind weather conditions prevailing at the
time, the atmospheric stability was probably neutral to slightly unstable.
The plume is well above the surface at 1 km downwind and quite compact.
The hole in the plume bottom apparent at 1541 and 1600 EST is probably
caused by the vortex-roll bifurcation phenomenon described by Scorer
(1958) and previously observed in the lidar observations of the plume
21
-------
30 45
RANGE — km
30 45
RANGE km
FIGURE 6 QUASI-INSTANTANEOUS (15 TO 30 SEC) PLUME CROSS SECTIONS (TOP
10 PHOTOS) ON 11 MAY 1970, AND CORRESPONDING 45-MINUTE
COMPOSITE ("AVERAGE") CROSS SECTIONS (BOTTOM), AT DOWNWIND
DISTANCES FROM THE STACK OF 1 km (LEFT) AND 2.5 km (RIGHT)
22
-------
from the Keystone Generating Station (Johnson and Uthe, 1971). At 2.5 km
downwind, however, the plume has diffused considerably and occasionally
reaches the ground.
Very little plume-width distortion is present in this series of
photographs because the vertical scans at 1 km downwind from the stack
intersect the plume axis at approximately 90°, while the scan at 2.5 km
downwind intersects the plume axis at an angle of about 60°.
3. Plume Trapping: Relation to Ambient Atmospheric Stratification
As was illustrated in the first frame of Figure 5, lidar
returns are obtained from the ambient aerosols in the clean background
air as well as from the particulates in the smoke plume. Since the up-
ward diffusion of the ambient aerosols, as well as the plume, is con-
trolled by the ambient thermal stratification, the lidar observations
during the Homer City experiment frequently revealed features of the
general meteorological regime within which the plume diffusion was taking
place.
Figure 7 gives an example of plume trapping by a thermal
lid at 1500 m height. The signal returns from scattered clouds at the
top of and above this layer are also apparent. The composite cross sec-
tions pictured here show that plume mixing was taking place over the
entire layer between the surface and 1500 m height during the two periods
of 22 and 27 minutes duration.
Another example of plume trapping is presented in Figure 8.
Here, the plume is seen to rise through a turbid, ambient haze layer and
to be trapped, along with the ambient aerosols, by a thermally stable layer
at 1500 m height. Near-neutral stability conditions within the layer cov-
ered by the rise of the plume probably existed at the time of these obser-
vations. The "hole" in the bottom of the plume caused by bifurcation is
again apparent.
23
-------
3.0 -
1.5 -
0 J
H-
LU
1 3.0
1.5 -
0 -I
TA-8509-4
FIGURE 7 COMPOSITE CROSS SECTIONS SHOWING PLUME
TRAPPING AT 1500 m HEIGHT
Each composite represents nine individual cross sections
at a downwind distance of 2.5 km. Clouds are visible
at and above the top of the trapped layer.
) 1 5
1 1 1
3.0
RANGE —
1
4.5
— km
1 1
6.0
I 1
7.!
24
-------
3.0 -
1.5 -
0 -1
I-
I
LU
1 3.0
1.5 -
0 -1
0.8 km
DOWNWIND
3.5 km
DOWNWIND
1.5
~T
3.0
RANGE
—I—
4.5
km
T~
6.0
7.5
TA-8509-5
FIGURE 8
EXAMPLE OF THE TRAPPING OF A PLUME RISING
IN A TURBID AMBIENT HAZE LAYER
The vertical extent of both the plume and the ambient
haze is limited to 1500 m by a stable layer.
25
-------
On 15 May, the plume observations showed an unusual multi-
layered atmospheric haze structure, as illustrated in the time sequence
of photographs in Figure 9. Up to six distinct haze layers are visible
in the lower 3000 m of the atmosphere. The stack plume is embedded
in the lowermost of these layers, which extends up to a well-defined
top at 750 m. The tilted and elongated (but intact) shape of the
plume indicates that the atmosphere within the plume layer was probably
still somewhat stable, giving rise to vertical shear in the wind direc-
tion (veering). The presence of the cloud layer at 2700 m may be partly
responsible for the continuation of this stability condition into the
late morning hours.
4. Plume Condensation
On frequent occasions, condensation of the water vapor in the
smoke plume formed visible clouds at the top of the plume. Figure 10
shows two lidar cross sections of this phenomenon that were obtained
during the morning of 11 May. These scans were made in a direction
approximately normal to the plume axis, at about 3 km downwind from the
stack. As presented, the pictures are viewed in the downwind direction.
The helicopter temperature profile at 1036 EST indicated a
lapse rate that was nearly dry adiabatic from the surface up to 950 m,
with high relative humidities near the top of this layer. The lidar
observations show that at 3 km downwind the plume top has already risen
to an elevation of 1400 m above the lidar (1150 m above the stack top).
Thin clouds caused by condensation are clearly apparent near
the bottom of the plume and along the outer boundaries of the upper
portion of the plume where the plume gas has been cooled by entrainment
with the surrounding air. Visual observations at this time confirmed
that the cooling tower plume was not intercepting the smoke plume.
26
-------
0 J
I 1.6
30-
30 45
RANGE — km
75
30 4.5
RANGE — km
TA-8509-12
FIGURE 9 TIME EVOLUTION OF PLUME STRUCTURE AT 2.5 km DOWNWIND, WITHIN
A MULTILAYERED HAZE STRUCTURE
The plume is contained in the lower layer with top at 750 m. A cloud layer
is visible at 2700 m (first two photographs), and lower clouds are forming
at 1300 m at 1236 EST.
27
-------
11 MAY 1970
0 J
1.5
I I
3.0
RANGE — km
1015
EST
4.5
FIGURE 10 EXAMPLE OF CLOUD FORMATION
WITHIN THE SMOKE PLUME 3 km
DOWNWIND FROM STACK
28
-------
E. Estimates of Plume Dispersion
Gifford (1971) has suggested that a "top-hat" fluctuating plume
model can furnish reasonable estimates of peak-to-mean concentration
ratios. This simple model is appealing in that it incorporates concepts
of buoyant plume rise and assumes a uniform plume-concentration distribu-
tion that conforms reasonably well to the quasi-instantaneous distribu-
tions observed by lidar.
Gifford's model uses as a parameter the plume radius-to-dispersion
ratio defined as
R/CT * = (2.96 X 10~5) 0 1/3U 1x2/3CT X (3.1)
y « y
where R = average value of instantaneous plume radius
a * = standard deviation of plume centerline location
-1
0 = stack heat emission in cal sec
U = mean wind speed
x = downwind distance
«. = crosswind standard deviation of the mean Gaussian
y
concentration distribution.
Previous work (Singer and Smith, 1966) has resulted in the
following well-known estimates of CT as a function of downwind distance
(x) and atmospheric stability:
4/5
ft = 0.32 x (neutral conditions). (3.2)
y
13/15
CT = 0.36 x (unstable conditions). (3.3)
Using these estimates, Equation 3.1 becomes
29
-------
-5 1/3 -1 -2/15
R/CT * = (0.95 X 10 ) Q U x (3.4)
y n
(neutral conditions)
-5 1/3 -1 -1/5
R/CT * = (1.03 X 10 ) Q U x (3.5)
y "
(unstable conditions)
A preliminary check of the validity of these expressions has
been prepared with the aid of the lidar data from 11 May, using values
* 6
of U = 6 m/sec and Q = 18 X 10 cal/sec as applicable for the after-
noon period. The results of a comparison between the observed and cal-
culated values of R/CT * are given in Table 4.
As indicated in the table, the observed R/n * values for the
"y
time period 1500-1700 EST compare reasonably well with the values given
by Eq. (3.4) if a near-neutral atmosphere is assumed.
However, the data for the earlier time period (1325-1458 EST)
give larger values of R/a *, even though there is good reason to believe
that the atmosphere was more unstable; e.g., cr */cj is larger than for
the later period. The reason for this discrepancy is unknown, but the
effect of plume trapping may be involved (see Section III-D-3). It
should be noted from Eqs. (3.4) and (3.5) that small changes in U can
significantly change the calculated values of R/CT *. This could explain
«7
the discrepancy between calculated and observed values if the actual
wind speed occurring at plume level during the period 1325-1458 EST was
smaller than that estimated on the basis of the surface wind measurements.
*
The surface winds, measured at the Jimmy Stewart Airport about 16 km
north-northeast of the lidar site, were reasonably steady at approx-
imately 6 m/sec from 215° during the period 1400-1700 EST. The upper-
level wind measurement nearest to the observing time was at 1250 EST at
the same airport. This sounding showed a speed of about 6 m/sec at the
plume elevation (500 m above the surface), with little vertical wind
shear. Hence, the best estimate for U was taken to be 6 m/sec.
30
-------
Table 4
COMPARISON OF OBSERVED AND CALCULATED VALUES
OF PLUME RADIUS-TO-DISPERSION RATIO (R/n *)
y
Downwind distance (m)
Number of cases
f
Surface wind direction
Surface wind speed^m/sec)
Plume rise (m)
R (m)
aR (»)
rr * (m)
y
rr * (m)
uz
*z*V
R/CT * (observed)
y
R/n. * (calculated) :
y
Neutral
Unstable
Time Period — 11 May 1970
1500-1656 EST
1000
13
215°
6
175
207
63
123
149
1.21
1.68
1.46
0.80
1503-1658 EST
2500
13
215°
6
340
467
103
400
189
0.47
1.17
1.28
0.67
1325-1458 EST
2500
21
215°
6
240
546
125
242
197
0.81
2.25
1.28
0.67
As observed at Jimmy Stewart Airport, Indiana, Pennsylvania.
31
-------
F. Comparison of Lidar Data with Plume SO2 Cross Sections
Observed by Helicopter
Only particulate distributions are observable by the lidar, and the
question of how well the diffusion of the plume particulates reflects
that of the plume gases, such as SO2, has been of interest for some time.
In an attempt to furnish additional information on this point, in this
section some of the lidar plume cross sections are compared with the
S02 cross sections obtained by plume traverses with the LAPPES-instrumented
helicopter.
These comparisons are necessarily limited in number because only a
small number of sets of lidar and helicopter cross sections were obtained
sufficiently close together in space and time to make the comparison
meaningful. Most of the helicopter observations were in the early morn-
ing, when the ground-based lidar measurements were frequently restricted
by ground fog. An additional complicating factor was an occasional lack
of coordination arising when communications were lost. Finally, there
were a few days when one or the other observational system (lidar or
helicopter) was not operating.
Nevertheless, some quasi-concurrent data were obtained, and the
results regarding observed plume heights are presented in Table 5.
Figures 11 and 12 show the helicopter data and some of the lidar data
used in Table 5. Cases 1 arid 5 (see Table 5) give reasonably comparable
average plume heights (within 50 m) for both helicopter and lidar data.
In Case 2, the lidar gives a considerably lower plume height than does
the helicopter, while the opposite situation holds for Case 3.
These results are much too sparse to enable any conclusions to be
drawn, particularly since a considerable time difference exists between
some of the helicopter and lidar cross sections. Overall, however, there
is a slight indication that the particulate plume heights are lower than
the S02, which might be explained by some settling or sinking of the
32
-------
X!
CU
H
QJ
E
G
r-t
osesi
i— f
o
n
o
•H
CJ
0>
en
w
en
O
U
rH
03
•H
T
O
•H
CJ
0)
01
tn
o
o
^
o
a
o
H
i— 1
OJ
W
*o C,
G
'!> &
G _i
£ C
0 ™
Q 01
H
Q
<» 3
be OJ
co E ^
SH 3 t,
^ ^ fat)
< ^ 'S
K
5
£ ^^
^
ol m
1-4 CO
CQ
(U "s
rH &
&. o
0 C
£
H S
C ^
QJ
S u
1 G
o 2
Q M
•H
Q
bD QJ
CO E .,
h 3 S
< 0)
X
ID 'e
s '
Q)
^ nj
«
B S
i — i &
Pn O
•'— <
•H CO
H 5
0) O
CO CJ}
Q d
Q>
o ^
in
TP
0
00
m
m
r-
T^
O
o
o
00
0
o
^'
o
CvJ
o
CO
rH
LO
m
CO
i— i
rH
O
m
CO
o
O
CM
o
m
m
CM
m
CM
00
m
CO
O
i
CO
CO
o
d
3
rH
CO
m
CO
o
m
CM,
O
O
in
CM
CD
O
1
CM
rH
to
O
&
m
rH
0
in
lO
CM
CM
t-
o
CO
CO
9
o
r-t
00
o
0!
ID
H
s
c
s
o
•o
33
-------
o
r~
O)
<
8
'8
o
o
in
(N
o
o
8
§ m
(a O
•- Z
<
O 2
h8
o
o
in
8
UJ
O
8
H
co
LU
o
(^
O>
<
o
§
o £
8 §
in
o 2
g £
CN
8 E
CO
8°
111
o
8 <
oc
o
LL.
CO
Z
O
LU
CO
O
cc
o
OC
<
Q.
CC
<
Q
_ I-
1- Z
LU LU
LU 5
O £
co X
LU
£t
£G
y oc
_l UJ
LU S
I O
LU ^
O o
Z 2
O cc:
co D
E Q
£ w
^ LU
S CO
O <
O O
cc
c
O
c
o
cu
8
c
o
CM
O a) o>
co .c '5
y o «,
s s
o"g £
CN « a)
D q-
_- OJ
O to »-
r— 03
- 5 £
LO _ •£
lg|
•2 E
S ^r
cu 00
R- c
Q. o
T OC .2
§ -g
cu
CO
C
0
cu
5 -^ co
2 I-
0- ^
OJ o>
•- 5 -t:
^ E >
^ c- co
2 o 2
C CM
O LO
O v ^—
CO
CN S C
O _ o
I- 05 §
o o o o o o
o o o o o
O CO CD "3-
-------
CROSS SECTION NO. 324
559-631 EST
5-4-70
1000
800
600
400
200
CROSS SECTION NO. 333
839-858 EST
1000
800
600
I—
5 400
I 200
0
1000
800
600
400
200
0
CROSS SECTION NO. 337
612-629 EST
5-11-70
5-15-70
500 1000 1500 2000 2500 3000 3500 4000 4500
DISTANCE — m
TA-8509-16
FIGURE 12 THREE OF THE HELICOPTER SO2 CROSS SECTIONS
USED FOR THE PLUME-HEIGHT COMPARISON IN
TABLE III-4
The contours represent SO2 concentrations in pphm: 5,
10, 20, 40, 80, 120, 160, etc.
35
-------
particulates. Not much of the latter, would be expected, though, if the
power plant precipitators are removing the large particulates as they
are designed to do.
36
-------
IV SAN JOSE URBAN POLLUTION-LAYER TEST
While the Mark VIII system was being prepared for the urban pollution-
layer observations in St. Louis, mobile and stationary tests were conducted
in the Menlo Park area and along local freeways. The results of several
of these special tests are presented for general interest in Figure 13.
These height/distance and height/time cross sections were among the first
such observations obtained by the newly modified Mark VIII lidar system.
At about this same time, another field project concerned with the
*
evaluation of an urban diffusion model was being conducted in nearby San
Jose (15 miles southeast of Menlo Park). Since this latter experiment
included the measurement of carbon monoxide (CO) and air temperature pro-
files over San Jose by means of an instrumented helicopter, it seemed worth-
while as part of the lidar system checkout to conduct a brief series of com-
parative measurements.
On several days during December 1970, the Mark VIII was situated in
the vicinity of Spartan Stadium near central San Jose and operated in the
vertically pointing stationary mode. Figure 14 presents a time series of
time/height cross sections obtained by lidar observations spaced 12 sec-
onds apart throughout most of one day. Concurrent temperature and CO
profiles obtained by the helicopter are given in the lower right portion
of the figure.
*
Jointly sponsored by the Coordinating Research Council and by the EPA.
t
Some minor technical problems occurred during this series. The horizon-
tal wavy lines at low heights and the occasional vertical white bands on
the photographs were caused by electronic ringing and intermittent noise.
However, these do not affect the general validity of the data.
37
-------
LU
O
LLJ
o
<
Q
I \ I
O Lf)
UJ^ — ±HOI3H
I
o
o
en
T~
DC
UJ
00
UJ
O
o
CM
1
0)
o
°?
h
— CD
_
f-
O)
Ic
to
-0
c
cu
cu
-
Q)
CD
^_
0)
O
I
o
I I
in
- 1H9I3H
$ CC £
^ Q "S
— 00
— r^-
— CD
^
CC
<
^
Q
LU
LL
5
o
^
LU
X
h-
LL
O
OO
CO
LU
1 —
_J
^
0
LU
Q.
a
o
"*""*
0)
j;
+-J
03
C
0)
03
a.
Q.
03
CD
CD
. .
V5
(U
a>
03
CD
~
0)
J5
0
O
QC
n
a
38
-------
1705-1741 PST
6 12 18 24 30 36
ELAPSED TIME —mm
1206 PST
1651 PST
6 12 18 24 30
ELAPSED TIME — mm
25 50
CO — ppm
0 5 10 15 20
TEMPERATURE — °C
TA-8509-13
FIGURE 14 LIDAR-OBSERVED TIME/HEIGHT CROSS SECTIONS OF THE URBAN
HAZE LAYER OVER SAN JOSE, CALIFORNIA, ON 11 DECEMBER 1970
Concurrent helicopter profiles of carbon monoxide (CO) and air temperature
are shown at the bottom right.
39
-------
The noon helicopter sounding shows temperature inversions at two levels,
approximately 450 m and 750 m above the surface, while the late afternoon
sounding indicates that the lower inversion has dissipated. The lidar data
reveal a haze layer persisting during the morning hours, with the haze top
varying between 400 and 600 m. (Scattered clouds initially present at
400 m height have dissipated by 1045 PST.) During the period from about
1250 to 1320 PST, the lower inversion is apparently penetrated by an
aerosol-laden air mass, which in turn is trapped by the upper inversion at
750 m. The lower haze layer immediately becomes reestablished but expands
at about 1425 PST to become a rather uniformly mixed layer up to a height
of about 900 m. By the time of the 1651 PST helicopter sounding, the top
of this layer corresponds closely to the 750-m height of the temperature-
inversion base. This rather close correspondence indicates that the lidar-
observed aerosol layers are indeed reflecting true mixing heights. In ad-
dition, the rapid variation in these heights clearly illustrates the poten-
tial lack of representativeness of a single temperature sounding.
The helicopter profiles of temperature and CO concentration have been
used to estimate the depth of the San Jose urban mixing layer for the period
7, 9, 10, and 11 December 1970. In Table 6, these estimates are compared
•with values obtained from the haze tops observed by the Mark VIII lidar.
As indicated in Table 6, the overall agreement in mixing depths obtained
by the two methods is reasonably good.
The helicopter and lidar data often indicate the presence of multiple
stable layers in the lowest 1000 m above the surface. The CO profiles on
these occasions show some penetration of CO through the near-surface mixing
layer. This occurs when the lid is not particularly strong (as indicated by
the temperature profile), but the concurrent lidar observations frequently
indicate intermittancy in the occurrence height of the lid. This intermit-
tancy may be the result of local (convective) or advective effects and seems
to represent a transitional stage in the lower atmospheric structure.
40
-------
Table 6
COMPARISON OF LIDAR-OBSERVED HAZE-LAYER TOPS
OVER SAN JOSE, CALIFORNIA, WITH MIXING-DEPTH
ESTIMATES FROM HELICOPTER PROFILE DATA
Date
7 December
9 December
10 December
11 December
Time
(PST)
1240
0800
1200
1700
0800
1200
1700
0800
1200
1700
Mixing Depths
(m)
Derived from
Helicopter Data
215
100
125, 675^
200, 725
125
425
500, 750
60, 550
500, 725
600
Lidar-Observed
H<
Haze-Layer Tops
(m)
450, 850^
300-500 (variable)
450-600 (variable)
750
300
450*
250
Low Clouds
350-700 (variable)
600-750 (variable)
Indicating the height variation over a one-half hour period.
t
Multiple values for the helicopter and lidar data represent
the tops of the surface and elevated mixing layers.
*
Lidar observation made at 1430 PST.
Source: Johnson et al. (1971).
41
-------
V ST. LOUIS URBAN POLLUTION-LAYER EXPERIMENT
A. Objectives
This experiment was basically a feasibility test of the usefulness of
lidar in urban areas with regard to air pollution applications. To the
best of our knowledge, it constituted the first field trial of a truly
*
ground-mobile lidar system —one that could take observations from a ve-
hicle while in motion. (Previous lidar systems had been only portable or
transportable and were thus capable of taking observations only in the fixed
mode.)
The reason for developing the mobile system was to permit observations
of the spatial, as well as the temporal, distribution of particulates in
urban pollution layers. As described by Clarke (1969) and a number of other
investigators, the structure of the pollution layer over and downwind of
cities, as determined by the relevant geophysical and meteorological varia-
bles, is a central concern in air pollution meteorology. The nocturnal
urban mixing layer and the diffusion of any pollutant plumes released within
it depend significantly on the urban heat output, urban roughness,
ambient thermal stratification, wind speed, and long-wave radiational pro-
cesses .
A truly comprehensive study of the modification of the lower atmosphere
as the air moves over a city would require measurement or estimates of all
of these variables. This first urban lidar experiment was a brief and rather
modest effort that was mainly concerned with determining the capabilities of
*
As of that date, three air-mobile lidar experiments using fixed-wing
aircraft had previously been conducted by SRI.
43
-------
lidar in this context, so that its future role in later studies entailing a
greater variety of types of measurements could be better defined. Thus, the
objectives of the study were limited to obtaining spatial distributions of
mixing depths over and downwind of St. Louis, by means of lidar observations
in a moving frame of the vertical particulate-layer structure over the city.
B. Experimental Techniques
The system was prepared for the experiment by installing the Mark VIII
lidar in a specially modified van with a 16-ft box-type body. Two 5-kW
motor-generator sets (operated from the van fuel system) were installed in
the front portion of the van to provide electrical power to the lidar while
the van was in motion. This area was partitioned off from the lidar section
and suitably vented to minimize noise and fumes in the lidar operating area.
A hatchway was installed in the roof of the van to permit vertical observa-
tions by the lidar. (Additional elongated trap doors were also installed
in the roof to permit elevation scanning while stationary.) A special modi-
fication was made to the van odometer so that it could control the lidar
firing rate. In this way, lidar observations could be taken at uniform in-
crements of distance, regardless of the speed of the van. Provision was
made for operator selection of a firing density of either 10 shots/mile or
20 shots/mile. An intercom was installed to permit communication between
the driver of the van and the lidar operator.
During the experiment, the van was driven over street routes in and
around St. Louis while lidar measurements were taken. The route followed
at any particular time was selected to give spatial coverage appropriate to
the wind direction prevailing at that time. Generally the maximum shot
density, 20 shots/mile, was used. The lidar data were cross-referenced
to spatial location by means of a written log in which specific road check-
points, disk recorder track numbers, and times were recorded. Each disk
44
-------
recorder track covered 160 lidar observations, or a route distance of
8 miles. Sometimes fewer shots/track were recorded to permit ending a
track at a convenient checkpoint.
The routes covered and the checkpoints used are indicated in the
St. Louis map depicted in Figure 15.
C. Data Summary
A general summary of the data collected during the experiment, along
with the prevailing weather conditions, is given in Table 7. A more
detailed summary of the individual runs, or route segments, is presented
in Appendix C.
D. Results
As indicated in Table 7, the wind speeds encountered during the ob-
servational period were rather consistently moderate to strong, as would
normally be expected in St. Louis in March. This prevented any significant
pollution-layer buildup during the observational period. The strength of
the return signal from this quite clean air was frequently too weak to per-
mit delineation of the pollution-layer structure.
In addition, a number of equipment malfunctions occurred that resulted
mainly from the low air temperatures and the rather severe road shock to
which the optical components and electronic modules were exposed. These
conditions had been anticipated, and heaters and heavy-duty shock mounts
had been installed and tested, but these precautions unfortunately proved
to be insufficient. Consequently, much of the resultant data contain
excessive electronic noise.
45
-------
-
01
o
<
I-
8
01
g
O
CL
V.
O
LU
I
O
Q
o
cr
-------
Table 7
DATA SUMMARY: ST. LOUIS LIDAR EXPERIMENT
1971
Date
1 March (Mon. )
2 March (Tues.)
4 March (Thurs. )
5 March (Fri.)
7 March (Sun.)
8 March (Mon.)
9 March (Tues.)
Period
(CST)
0938-1148
1315-1525
1623-1627
0820-1153
1252-1612
0450-0850
1002-1247
1355-1405
0437-0730
0759-1013
0424-0649
0638-1041
1219-1253
0431-0826
0859-1228
Run
Numbers
32-54
56-78
80
82-112
114-144
146-176
178-200
202
204-232
234-246
248-270
272-302
304*
306-336
338-370
Number
of
Runs
12
12
1
16
16
16
12
1
15
7
12
16
1
16
17
170
Weather
OlO 060°/15-20 32°
SOCDlSOlDlO 060° /10
901D/910 060°/10 50°
15007 360/10-15 31°
150010 360° / 15
OlO 270° /5 22°
/®7H 220° /5 36°
OlO 200°/5-10 45°
/OlO 180° /5 34°
OlO 200° /10
SOlDlO 270/15 31°
O15 300° /5 25°
O15 320°/10-15
100O/O10 130° /5 30°
80010 130° /10
Stationary run with van parked near the National Weather Service
Environmental Meteorological Support Unit (EMSU) balloon release
site; all other runs are mobile.
47
-------
It should also be noted at this point that an electronic problem in
the video disk recorder during the St. Louis experiment prevented the col-
lection of valid data within the first 150 m from the lidar. Thus, only
data above 150-m height are shown in the figures in this section. This,
of course, is unfortunate because of the important meteorological processes
and associated structure that often occur below this height within the noc-
turnal urban boundary layer.
The net result of these weather and equipment problems was that, out
of the rather large amount of data collected, only a small amount turned
out to be both of meteorological interest and sufficiently free of techni-
cal difficulties to be useful. Thus this first St. Louis lidar field study
can be considered as only partly successful. The potential of lidar for
studying urban air pollution was successfully demonstrated, but insuf-
ficient valid data were collected under conditions of interest to permit
any definitive conclusions to be reached regarding the structure and
behavior of the urban mixing layer.
In this section, we present and briefly discuss examples of the St.
Louis lidar data. The purpose of this is not to shed any revelations
about the air pollution meteorology of the St. Louis area, since the data
are not sufficient for that, but rather to illustrate the nature of the
capabilities of a mobile lidar for use in future studies of urban air
pollution.
The fundamental advantage of the mobile lidar as an observational tool
rests in its ability to obtain spatial as well as temporal distributions
of aerosol-layer structure. This is well illustrated by Figure 16, which
shows the time evolution—from early morning to mid-morning—of the verti-
cal aerosol distribution over two approximately nine-mile routes through
East St. Louis on 4 March 1971. The portions of each cross section marked
48
-------
9-1
6-
3-
E
8 0J
PT. 35
I 9 -i
uu
I
6-
3-
0-1
I
0
PT. 6
4 MARCH 1971
0643-0702 CST
(LOOKING TO SE)
4 6
DISTANCE — miles
PT. 30
1002-1013 CST
(LOOKING TO NW)
•
8 T 10
PT. 36
TA-8509-8
FIGURE 16 TIME CHANGES IN POLLUTION-LAYER STRUCTURE
APPROXIMATELY ALONG WIND THROUGH EAST
ST. LOUIS
49
-------
"a-b" are over a common route segment (see Figure 15) but are separated
in time by about three hours.
The early morning observations (top picture in Figure 16) indicate a
rather uniform aerosol layer topping at about 300 m, overlaid by several
plumes from industrial sources. The buoyancy of these plumes has appar-
ently permitted them to penetrate into the stable air above the lower well-
mixed layer. At the time the sky was clear, with an estimated surface wind
from 240° at 5 knots.
The mid-morning cross section (bottom picture in Figure 16) shows that
convective mixing has become dominant and has created a relatively uniform
aerosol layer extending to a height of about 600 m. This was also the
height of the tops of the plumes observed earlier and thus is probably the
height of the base of a general, ambient, elevated inversion layer. The
meteorological conditions at this time were much the same as earlier. The
surface winds were from 220 at 5 knots.
A 16-mile cross section from extreme northwestern to central St. Louis
is presented in Figure 17. This route lies approximately crosswind to
the surface wind at that time, which was from 240 at 5 knots (see Fig-
ure 15). The main feature of interest in Figure 17 is the change in the
height of the aerosol layer from about 250 m over northwestern St. Louis
to about 500 m over central St. Louis. These observations contain con-
siderable electronic noise, but the shape of the pollution layer is reason-
ably well-defined. Variations on the order of 100 m are apparent in the
top of the mixed layer near central St. Louis.
Figure 18 shows a similar vertical cross section, except it extends
from central St. Louis to the extreme south-southwestern edge of the city
during the early afternoon on 1 March 1971 (see Figure 15). The surface
winds were from 060 at about 10 knots. Thus, the route represented in
50
-------
GO
CJ
s-
CO
o
-
CJ
Lt
5
V
t/3
O
CO
CM
(O
s
ID
00
LU
_ o
CM '
CO LU
t" o
CO
CD
CM
CM
- O . CM .
£~ <-
Q
Z
to
CO
O
DC
O
DC O
0. Z
a.
< O
o ^2
D D
to
cc
LU
g "
H O
D Z
<2
o i-
> Q
tr
D
O
001 — ±HOI3H
51
-------
1426 CST
15 -i
12 -
9 -
I-
I 6 -
LU
I
3 -
0 J
I
0
PT. 1
(ARCH)
1 MARCH 1971
1441 CST
1 1 T
6 T 8
DISTANCE — miles
T
10
12
PT. 13
(1-55 AND
1-244)
TA-8509-9
FIGURE 18 VERTICAL POLLUTION-LAYER STRUCTURE IN THE DOWNWIND
DIRECTION ALONG 1-55 FROM THE ARCH IN CENTRAL ST. LOUIS
(LOOKING TO SOUTHEAST)
52
-------
Figure 18 lies approximately in a downwind direction from central St.
Louis. In this connection, it is interesting to note the slightly higher
top and the greatly enhanced density of the pollution layer as it moves
downwind from the central part of the city. This can be readily explained
on the basis of the additional pollutants that would be added to the mixing
layer from the many sources to the southwest, and thus downwind, from the
arch in central St. Louis.
During the field program, attempts were made to measure the structure
of the overall urban pollution plume or layer as it moves downwind from
St. Louis. Figure 19 depicts a series of cross sections obtained approx-
imately 10 miles downwind of central St. Louis near noon on 4 March 1971.
The route followed is in a west-east direction (see Figure 15). Since
the surface wind during this period was from 180° to 220° at 5 to 10 knots,
the lidar route is approximately in a crosswind direction. As shown in
Figure 19, during this mid-day period the mixing layer was apparently
well developed, having a top at an average height of 700 m, with height
variations of ±200 m. The shift in the urban-pollution plume position
during the period indicates that the winds were becoming more westerly with
time.
The plume width is difficult to judge but appears to be about 6 to 8
miles, except for the middle cross section where it is apparently somewhat
wider. This 6-to-8-mile plume width seems to be reasonable, since this
would encompass most of the heavily industrialized areas near the Missis-
sippi River in St. Louis and East St. Louis.
53
-------
L CN
2 t
I
§
in
•T
00 LU
O
t/3
Q
DC
HI
ID
•3-
O
Q.
O
UJ
cc
O
O
UJ
CO
cc ^
Z> D
I- O
O .J
h- CO
CO
_)
EC <
O LL
O a
Q- 5
< O
O Q
UJ CN
> JH
0)
UJ
cc
3
(D
54
-------
VI RELEVANT SUBSEQUENT LIDAR APPLICATIONS
As previously mentioned, when the Mark VIII lidar system was applied
in the three experiments sponsored under this contract, it was in each
case the first test of the newly fabricated system in a particular opera-
tional mode or configuration . On the basis of these tests , the system
was rapidly refined, resulting in significantly improved data quality for
later experiments. In this section, we briefly present samples of some
recent results from subsequent field studies to illustrate the current
*
capabilities of the system.
A. Diurnal Variation of the St. Louis Mixing Layer
(NSF/SRI METROMEX ProjectsT
Figure 20 presents a height/time cross section of aerosol structure
recorded over central St. Louis, obtained with the Mark VIII lidar (Uthe,
1972) . This figure shows the urban pollutants trapped within the stable
air near the earth 's surface . Clouds and haze are visible aloft and thus
hinder the passage of the solar energy to the surface. The figure also
shows that beginning about noon, a rapid deepening of the polluted layer—
apparently caused by vertical air motions resulting from surface heating—
effectively dilutes the urban pollutants. The convective cloud formation
observed by the lidar during the afternoon could be related to precipita-
tion anomalies that have been observed downwind of urban areas.
*
These studies were not supported by this contract.
f
Study sponsored by the National Science Foundation
55
-------
(D
O
C/3
D
CO
QC
LJJ
LLJ
to
_j
O
DC H
LU —
o <
_ D.
Z LU
O ^
H ^
G w
LU ^
w m
CO D
O LU
CC >
CJ CC
,,, LU
LU LO
co
<
CD
LU
I
LU
DC
D
g
LL
— 30nili1»
s ja)auio | i)| — 3001111*
56
-------
B. Spatial and Temporal Variations of the
Los Angeles Mixing Layer (EPA Project)
Figures 21 and 22 are examples of the results obtained with the
Mark VIII mobile lidar system in a field study of mixing heights over
*
the Los Angeles Basin in September 1972. The 70-mile triangularly
shaped freeway route over which the lidar observations were made is shown
in Figure 23.
Of particular interest in Figure 21 is the rapid and substantial
change in the structure of the mixing layer on the east side of the Basin
as the marine air penetrates eastward during the afternoon. An upper
aerosol layer is apparent near Azusa at 1550 PDT, possibly the result
of vertical air currents along the heated south-facing slopes of the
San Gabriel Mountains . A wind reversal from WSW winds at the surface to
NE winds at 1000 m elevation was observed by the El Monte EMSU station
at about this time. This vertical wind shear could have caused the waves
apparent on the aerosol layers.
Noteworthy features of the mixing-height spatiotemporal pattern in
Figure 22 include the enhanced vertical extent of the mixing layer near
South El Monte at 1300 PDT and the eastward propagation of the marine
layer (sea breeze) along Legs 2 and 3 between 1200 and 1400 PDT, as in-
dicated by the large contour gradient of mixing height. In addition,
larger mixing heights are observed over downtown Los Angeles at 1100
PDT—probably because of urban effects—and also in the Long Beach area
at 1200 PDT, apparently because of vertical motions resulting from hori-
zontal convergence of the air on the leeward side of the Palos Verdes Hills
This Palos Verdes Convergence Zone is a typical feature of the sea-breeze
circulation in the Los Angeles Basin.
*
Study conducted by W. B. Johnson, T. J. Lemmons, L. A. Knight, and
R. E. Chalfant of the Meteorology Laboratory, EPA.
57
-------
U-
> o
Z LU
0 2
Q O
UJ LU
LU O
£§
o i
'I
O <
O
CO 1-1 •*
LU K '
Z DC
< < I
CO 5 -°
O LU Ifl
E ^
Z <
cc co
5 1- LU
S LL I
I- < H
DC
D
O
UJ
o
cc
— 1HOI3H
sjaiaoi — 1HOI3H
58
-------
Oi
CO *~
— LU D
5t 2
x •
Lu ;
*2
CD co
z S
< «
c
co g
| CD
\H S
DQ ft" flJ
O LU ^
0)
QQ.
H O
O
co <
2 LU
O LU
I M
9 H
11 ™
i- O)
il
o •— o
OLU -D o
—I .= CD
< 2
O O
O -I
-I <
CN
CN
LU
IT
II
59
-------
CM
P~
CD
oc
LU
CO
O-
LU
CO
CO
<
CD
CO
CO
O
D
O
CE
CC
<
Q
CD
O
CO
CM
LU
IT
60
-------
VII CONCLUDING REMARKS
This project entailed the development and application of a sophis-
ticated and unique atmospheric remote-probing system incorporating state-
of-the-art electronic and optical technology. The technical problems
that inevitably accompany the field testing of complex new systems of
this nature were not completely avoided, but in every case these problems
were overcome. The overall results of the field programs appear to con-
firm the usefulness of lidar systems of this type for air pollution and
meteorological studies.
Perhaps the most significant feature of the Mark VIII system is its
capability for analog processing and display of the data in real-time.
This permits the inherent high data-rate advantage of the lidar technique
to be exploited and largely eliminates the need for costly and time-
consuming post-experiment data analysis.
"Seeing" the atmosphere in cross section has only recently become
possible through the development of lidar. This new perspective should
greatly help meteorologists to better understand the atmosphere and its
capacity for diluting pollutants. Although the conventional lidar tech-
nique using direct backscattering has some basic limitations for applica-
tions beyond those in this study, a number of these can be overcome with
advanced techniques currently under development, such as the use of
multiple wavelengths. In addition, the operation of lidar systems in
conjunction with other types of remote sensing devices offers consider-
able promise for the achievement of more comprehensive atmospheric mea-
surements in days to come.
61
-------
REFERENCES
Allen, R. J., and W. E. Evans, 1972: Laser radar (LIDAR) for mapping
aerosol structure. Rev, of Sci. Inst. , 43, 1422-1432.
Barrett, E. W., and 0. Ben-Dov, 1967: Applications of the lidar to air
pollution measurements. J. Appl. Meteor., C5, 500-515.
Clarke, J. F., 1969: Nocturnal boundary layer over Cincinnati, Ohio.
Mon. Wea. Rev., 97, 582-589.
Collins, R.T.H., 1970: Lidar. Appl. Optics, j}, 1782-1788.
Gambling, D. J., and K. Bartusek, 1970: Lidar observations of tropospheric
aerosols. Technical Report, Dept. of Physics, the University of
Adelaide, Adelaide, Australia.
Gifford, F. A., 1971: Peak to mean concentration ratios according to a
"top-hat" fluctuating plume model. Proceedings, AMS Conference on
Air Pollution Meteorology, Raleigh, N.C., April 1971.
Hall, F. F., 1972: Temperature and wind structure studies by acoustic
echo sounding. Remote Sensing of the Troposphere, Chapter 18,
V. E. Derr, editor, Government Printing Office.
Hamilton, P. M., 1966: The use of lidar in air pollution studies. Int.
J. Air/Water Poll., 10, 427-434.
Hamilton, P. M., 1967: Plume height measurements at Northfleet and Tilbury
Power Stations. Atmos. Environ., JL, 379-387.
Hamilton, P. M., 1969: The application of a pulsed-light-rangefinder
(lidar) to the study of chimney plumes. Phil. Trans. Roy. Soc.
Lond., A265, 153-172.
Hosier, C. R., and T. J. Lemmons, 1972: Radiometric measurements of
temperature profiles in the planetary boundary layer. J. Applied
Met., 11, 341-348.
63
-------
Johnson, W. B., 1969a: Lidar applications in air pollution research and
control. J. Air Poll. Contr. Assoc., 19, 176-180.
Johnson, W. B., 1969b: Lidar observations of the diffusion and rise of
stack plumes. J. Appl. Meteor. , 8^, 443-449.
Johnson, W. B., 1971: Lidar measurements of plume diffusion and aerosol
structure. Proceedings, AMS Conference on Air Pollution Meteorology,
Raleigh, N.C., April 1971.
Johnson, W. B., and E. E. Uthe, 1971: Lidar study of the Keystone stack
plume. Atmos. Environ. , J5, 703-724.
Johnson, W. B., W. F. Dabberdt, F. L. Ludwig, and R. J. Allen, 1971:
Field study for initial evaluation of an urban diffusion model for
carbon monoxide. Comprehensive Report, Contract CAPA-3-68 (1-69),
NTIS Accession No. PB-210820, Stanford Research Institute, Menlo
Park, California.
Little, C. G., 1972: Status of remote sensing of the troposphere.
Bull. Amer. Met. Soc., 53, 936-949.
McAllister, L. G. 1972: Acoustic radar sounding of the lower atmosphere.
Mathematics of Profile Inversion, NASA Technical Memorandum, NASA
TMX-62, 150.
McCormick, M. P., and W. H. Fuller, 1971: Lidar applications to pollution
studies. AIAA Paper No. 71-1056, Joint Conference on Sensing of
Environmental Pollutants, Palo Alto, Calif., November 1971.
Olsson, L. E., W. L. Tuft, and W. P. Elliott, 1971: Observational study
of the haze (mixing) layer in Western Oregon using laser radar, in-
strumented aircraft, and meteorological balloons. Paper presented
at the 64th Annual Meeting of the Air Pollution Control Association,
Atlantic City, New Jersey.
Pooler, F., and L. E. Niemeyer, 1970: Dispersion from tall stacks: an
evaluation. Paper No. ME-14D presented at the Second Intern. Clean
Air Congress, Washington, B.C., December 1970.
Schiermeier, F. A., and L. E. Niemeyer, 1970: Large Power Plant Effluent
Study (LAPPES)--Vol. 1, Instrumentation, Procedures, and Data Tabu-
lations (1968). NAPCA Publication No. APTD 70-2.
64
-------
Schiermeier, F. A., 1970: Large Power Plant Effluent Study (LAPPES)—
Vol. 2, Instrumentation, Procedures, and Data Tabulations (1967 and
1969). EPA Publication No. APTD-0589.
Schiermeier, F. A., 1972a: Large Power Plant Effluent Study (LAPPES)—
Vol. 3, Instrumentation, Procedures, and Data Tabulations (1970).
EPA Publication No. APTD-0735.
Schiermeier, F. A., 1972b: Large Power Plant Effluent Study (LAPPES)—
Vol. 4, Instrumentation, Procedures, and Data Tabulations (1971).
EPA Publication No. APTD-1143.
Scorer, R. S., 1958: Natural Aerodynamics. Pergamon Press, London and
New York, 211-212.
Shaw, N. A., 1972: Acoustic sounding techniques for monitoring atmo-
spheric structures. Argonne National Laboratory, Annual Report
1971-72 (ANL 7860).
Singer, I. A., and M. E. Smith, 1966: Atmospheric dispersion at Brook-
haven National Laboratory. Intern. J. Air and Water Poll., 10,
125-135.
Uthe, E. E., 1972: Lidar observations of the urban aerosol structure.
Bull. Amer. Met. Soc., 53, 358-360.
Uthe, E. E., and W. B. Johnson, 1971: Lidar observations of the lower
tropospheric aerosol structure during BOMEX. Final Report, Contract
No. AT(04-3)-115, Project Agreement No. 83, Stanford Research In-
stitute, Menlo Park, Calif.
Uthe, E. E., and N. A. Shaw, 1973: A comparison of atmospheric structure
as observed with lidar and acoustic sounder techniques. Paper pre-
sented at the Fifth Conf. on Laser Probing of the Atmosphere,
Williamsburg, Va., June 1973.
Viezee, W., and J. Oblanas, 1969: Lidar-observed haze layers associated
with thermal structure in the lower atmosphere. J. Appl. Meteor.,
8, 369-375.
65
-------
Appendix A
MARK VIII LIDAR SYSTEM DETAILS —
REPRINT OF JOURNAL ARTICLE* ENTITLED
"LASER RADAR (LIDAR) FOR MAPPING AEROSOL STRUCTURE,"
BY R. J. ALLEN AND W. E. EVANS
*
Rev. Sci. Instruments, 43, October 1972.
-------
Reprinted from:
THE REVIEW OF SCIENTIFIC INSTRUMENTS
VOLUME 43. NUMBER 10
OCTOBER 1972
Laser Radar (LIDAR) for Mapping Aerosol Structure*
R. J. ALLEN AND W E. EVANS
Stanford Research Institute, Menlo Park, California 94025
(Received 16 February 1972; and in final form, 20 March 1972)
This paper describes a third-generation experimental Q-switched, ruby laser radar or lidar useful for air pollu-
tion mapping and meteorological studies. This system differs from its predecessors in that it incorporates (1) auto-
matic inverse-range-squared receiver gain compensation, (2) coaxial transmitter and receiver optics, (3) automatic
scanning and pulsing at equal angle increments or time intervals for fixed operation, and at equal distances traveled
for mobile operation, and (4) video disk storage and instant replay of scan sequences. Details of these improve-
ments are included. Photographs of typical displays are also included to illustrate atmospheric returns recorded
during range-height mapping of a smoke plume from an industrial plant, showing a cross section of the plume's
structure; plan position mapping of the effluents from several industrial plants, indicating those emitting particu-
lates; and distance-height mapping, showing differences in pollution and changes in the height of the inversion
layer obtained while driving through a city.
INTRODUCTION
The development of the laser radar or lidar, as it has
come to be known, has indeed provided a useful research
tool for the study of the atmosphere, as predicted by
Northend, Honey, and Evans1 early in 1966. A third-
generation ruby lidar has now been built—mounted within
a truck (Fig. 1)—and is in operation mapping effluents
from smoke stacks, smog layers above urban areas, the
height of the inversion layer, and other related air pollution
and meteorological phenomena.2 This lidar system not only
provides an improved instrument for air pollution and
meteorological research but also a useful daytime or
nighttime means for mapping the sources of effluents over a
portion of a city from a single vantage point.
The new lidar system, called the SRI/NAPCA Mk VIII,
was developed jointly under the sponsorship of Stanford
Research Institute and the National Air Pollution Control
Administration (now the Environmental Protection
Agency). The primary improvements of the Mk VIII over
its predecessors are described in this paper, along with the
over-all system design and typical displays obtainable
during field observations. The basic lidar system equations,
receiver noise considerations, and general system parame-
ters important in the design and use of the lidar were in-
cluded in a previous paper1 and should be reviewed for a
greater understanding of the principles of lidar systems.
I. SYSTEM CONSIDERATIONS
For SRI lidar systems developed prior to the Mk VIII,
it has been the practice to display and photographically
record each individual oscilloscope trace or family of
oscilloscope traces.3 The log amplitude of the return signal
shown in the photographic record is read manually at
various ranges, as dictated by slope considerations and
points of inflection in the trace. These data are analyzed
and subsequently plotted automatically by special com-
puter programs that account for I//?2 corrections.4 The
plots produced by the computer include those that show
the vertical cross section of particulate matter intercepted
by the transmitted laser beam and indicate the gross
relative concentrations.
One Mk VIII design goal was to incorporate a real-time
display that would indicate in raster or map form the
location, spatial distribution, and approximate concen-
trations of particulates relative to clear air and thus would
allow an on-the-spot inspection of the atmospheric area
being surveyed. At the same time, the data obtained were
to be preserved for later computer processing and refine-
ment. This goal presented some unique requirements due to
the limited prf (30 pulses/min) and narrow transmitted
beamwidth (under 1 mrad) satisfactory for obtaining the
required performance during field observations. The
bandwidth requirements to handle the narrow pulse widths
(under 30 nsec) also presented some major limitations.
The system selected incorporates some unique signal
processing circuitry and control logic, together with a
television-type instant playback disk memory. This allows
real-time display, as well as data recording (with some
bandwidth limitations) for off-line processing. Either a
radar type PPI (plan-position indication) or an RHI
(range-height indication) intensity-modulated display
format can be selected. A-scope and other derivations of
the radar type display are also obtainable, as indicated
later.
To obtain a uniform low-brightness two-dimensional
presentation representing a clear air return at all ranges, it
was necessary to correct for the inverse-range-squared
(1/.R2) decay of the received usable signal. In principle,
this could be accomplished in either the photomultiplier
tube (PMT) or in the post-detection amplifier. Since a
large dynamic range (>5 decades) was involved, the
former method was selected to minimize PMT and ampli-
fier saturation effects, especially for operation within a
foggy atmosphere when the dynamic range of the useful
signal information is very large. Over-all PMT gain can be
A-3
-------
LASER RADAR
1423
controlled by varying the gain of individual dynodes5 or of
combinations of dynodes. The objective is to control the
total gain for a series of k dynodes (5*0 in such a manner
that the ratio of Sk(t)/R2 will remain constant.
Recording of an elevation or azimuth angle label along
with each lidar shot was circumvented and the display
electronics were simplified by assigning addresses to ap-
propriate locations on the disk. These details of the lidar
design are described in Sec. II.
II. LIDAR DESIGN
A. System Description and Timing
The lidar system (Fig. 2) consists of the laser transmitter
and receiver, parts of which are mounted on the pedestal—
together with the coaxial optics—and called the "lidar
head"; the head and pedestal position control; the video
processing and storage section; and the displays. Position
control of the head and pedestal and firing control can be
either automatic or manual. In the automatic mode, the
firing rate can be controlled in increments of time or
increments of the distance traveled by the truck in which
the Mk VIII is mounted. Between selectable limits, the
pedestal can be moved automatically from left to right, and
the lidar head can be moved automatically from horizontal
to vertical and back to horizontal, in equal increments of
degrees between shots. After the scan has been completed,
the lidar can be programmed either to stop or to return
immediately and repeat the scanning sequence. The head
and pedestal can also be locked in a fixed position with the
firing maintained at an automatic rate.
In the manual mode, the head can be stepped, and the
firing can be controlled at will either from the head location
or remotely from the electronics console. After a firing
sequence has been completed the position control pro-
grammer can be used for sequentially addressing the video
disk for immediate playback and display of the recorded
signals. A volume control and speaker allow the program
control pulses to be heard, a feature which has been found
useful in monitoring the progress of an experiment.
The time between the manual or automatic command to
fire the lidar and the actual lase is varied slightly to main-
tain proper phasing with the disk. Four main times are
identified in Fig. 3 as GO, MATCH, FIRE, and LASE.
The GO pulse occurs when the command to fire a shot is
given. Whenever the disk address matches the externally
specified address, the MATCH time occurs. Since the disk
is rotating at 30 rps, the MATCH time can be any time
from 0-66 msec after the GO time, depending upon the
disk position at the time of GO.
The time between MATCH and FIRE is adjustable be-
tween 100 and 1000 ^sec by a front-panel control. This
adjustable MATCH-FIRE delay is provided so that the
entire firing and subsequent recording process can be
{•I VAN EQUIPPED WITH Mk VIII LIOAR SYSTEM
Ibl ELECTRONICS AND GYRO-MOUNTED VIDEO DISC
IcI LASER HEAD AND PEDISTAL ASSEMBLY
FIG. 1. View of van and Mk VIII lidar system.
delayed with respect to the disk-lidar coincidence time.
The delay is used (1) to compensate for phase shifts
introduced when a recorded disk is removed and later
replaced at a slightly different angular position, (2) to
offset a new data sequence slightly so that records will
interleave between previously recorded records for fuller
utilization of the available disk storage capacity, and (3)
to provide time for the disk circuits to transfer between the
read and write modes when the display is operated in a
read-between-the-shots manner.
The FIRE time occurs when the signal is given to the
A-4
------- |