ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF ENFORCEMENT
FIELD EVALUATION OF MOBILE LIDAR
FOR THE MEASUREMENT OF SMOKE PLUME OPACITY
Project No. NEIC-TS-128
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
DENVER, COLORADO
February 1976
-------
FROM
ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF ENFORCEMENT
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
BUILDING 53, BOX 25227, DENVER FEDERAL CENTER
DENVER, COLORADO 80225
T° Mr. Lloyd Kostow DA^ August 12, 1976
Region IX, S&A Division
Remote Sensing Specialist
SUBJECT: Transmittal of the Lidar Report: Field Evaluation of Mobile Lidar for
the Measurement of Smoke Plume Opacity
Attached are three copies of the subject report for your information.
The main body of this report describes the lidar tests that were
performed on the California ARB Smoke Generator, the Kaiser Permanente
Facility and the Industrial Facilities at Lathrop, California. As an
attachment to this report the technical report from the Stanford
Research Institute is also included.
If you have any questions or comments, please do not hesitate to call
me. I thank you for your assistance in setting up and making the contacts
with the industrial facilities prior to the initiation of this study.
Arthur W. Dybdahl
Enclosures
-------
Contents
I INTRODUCTION . . . . 1
II SUMMARY AND CONCLUSIONS ..... 3
III BACKGROUND ............ 7
IV DESCRIPTION OF THE FIELD
EVALUATION TESTS .... 9
Clean Air Tests 9
Screen Test 9
Smoke Generator Tests ...... 11
Lidar Tests at Kaiser-Permanente . 15
Lidar Tests at Occidental
Fertilizer Company and
Libby-Owens-Ford Co . 17
ATTACHMENT
Lidar Applications for Smoke Plume
Opacity Measurements
-------
I. INTRODUCTION
The primary function of the National Enforcement Invèstigátions
Center (NEIC) of EPA is the collection of various types of environmental
data for use as evidence in enforcement actions against polluters. Such
data collection involves the use of a wide variety of techniques and in-
strumentation. Where possible, remote sensing techniques are used to
complement traditional methods. Also, as new techniques are developed
that are more reliable and accurate, NEIC employs them to ensure that
the highest quality data are obtained.
With respect to emissions from stationary sources, plume opacity is
one of the primary parameters monitored. Opacity is specifically limited
by Federal, State or local regulations for practically all sources. The
traditional method of monitoring plume opacity is visible emission
observations (VEO’s) performed by a trained human observer. An instrumental
method involving the use of an optical transmissometer mounted in the
stack or associated ductwork is increasingly being used for continuous
monitoring of stack emissions. The two methods, however, are not directly
comparable because they measure opacity at two different locations: one
from the ducting leading to the stack and the other at the stack exit.
The VEO’s are also subject to interference by environmental conditions
that sometimes limit their accuracy and frequently preclude their use.
Research performed by and for EPA has demonstrated that a remote
sensing technique employing a mobile lidar* system is a reliable and
accurate instrumental method of measuring plume opacity. The NEIC is
procuring a mobile lidar system for use in collection of enforcement data.
* Lidar an acronym for Light Detection and Ranging, commonly refers to
laser radar.
-------
2
In the development of technical specifications for the mobile
lidar, other sources with similar equipment were investigated. The
research performed at Research Triangle Park, N.C. (RIP) by EPA staff
(1970 to 1974) pointed out the need forimprovements in the data handling
capability of the RTP lidar system; A mobile lidar system with amore
sophisticated data processing capability and with the additional ability
to track plumes of particulate emissions was contracted from the Stanford
Research Institute (SRI), Menlo Park, California. NEIC conducted a
limited-scope field study with the SRI lidar system to evaluate its
performance under a variety of field and test conditions comparable to
typical EPA enforcement data collection. Several calibration tests were
conducted, including opacity measurements of emissions from a smoke
generator used for certification of visible emission observers. Opacity
was measured for visible emissions from three industrial plants during
both daylight and darkness, and residual plumes were tracked.
The results of the SRI lidar study were used in the development of
design and performance specificaitons for the NEIC lidar system. This
report summarizes study conditions and the results of the study.
Additional details are provided in the. Attachment, the contractor’s
report, “Lidar Applications for Smoke Plume Opacity Measurements,
December 1975.”
-------
3
II. SUMMARY AND CONCLUSIONS
This field study was designed to evaluate the utility of various
features of a mobile lidar (laser radar) system in the monitoring; of
smoke plume opacity and optical backscatter. The study was carried out
with a contractor’s lidar system in the San Francisco Bay area and in
the San Joaquin Valley of California.
The following tests were performed with the lidar:
Clean Air Tests The lidar system electronics were calibrated
including the linear and the logarithmic video amplifiers and the
inverse-range-squared compensation circuitry.
Screen Tests An optical screen was used in conjunction with the
lidar in order to evaluate the feasibility of using such a device
as an opacity calibration target in the field. The use of screens
is feasible; however, specular reflections from the wire grid and
the degrading affects of field handling do present problems.
Additional study is required to establish the practicality of
screens for lidar calibration.
Smoke Generator Tests The lidar was used to interrogatejremotely
test) the black and white smoke plumes emitted by a calibrated
mobile smoke generator positioned 270 m (885 ft) away. The plumes
from the generator stack were puffy which was manifested as a
transient condition in the resultant plume opacity. The lidar
easily documented the variations. However, the magnitude of the
variations were not fully documented by the instack transmissometer
because its response time was too slow to closely follow the excursions
in opacity. Due to the low horizontal atmospheric visibility, this
-------
4
test also affOrded an evaluation of the lidar’s receiver gating
logic, for controling incoming signal intensities resulting from the
optical energy backscatter from the clean air in front of and
behind the plume. Signal. control for the backscattered energy from
within the smoke plume was likewise evaluated.
Lidar Tests at Kaiser—Permanente The lidar observed visible stack
emissions from cement kilns. The emissions contained water vapor
which condensed near the stack opening. The lidar located the
residual plume which contained only the particulate emissions after
the condensed water aerosols had evaporated. This test also demon-
strated the lidar’s capability of measuring exit plume opacity, even
when a descending plume is in the lidar’s line-of-sight to the
stack exit. Both day and night tests were performed.
Lidar Tests at the Occidental Fertilizer Company and ‘the Libby-
Owens—Ford Company Observations of particulate emissions were
made at night for opacity, dispersion and combination of small
plumes. The lidar was an effective implement in the measurement
and monitoring of these parameters.
These lidar tests revealed several very important features that
need to be included in the mobile lidar system which EPA/NEIC is procuring.
1. The minimum effective aperture of the receiving telescope
should be 20.3 cm (8 in) to aid in the detection of weak
return signals from the clear air beyond a dense smoke plume.
2. The minimum output energy per pulse of the ruby laser should
be 1.5 joules. Such a laser will provide sufficient peak
power to yield a re,turn (backscatter) signal for the clear air
behind a dense or nearly opaque particulate plume, and thus
-------
5
permit an accurate opacity determination. Calculation of
opacity for a plume of greater than 80% opacity is not prac-
ticable with a lower output laser.
3.. The maximum angular divergence of the laser beam should be 0.7
milliradians.
4. The ruby laser should have sufficient cooling capacity (for
the head) to sustain a firing rate of a pulse every second in
an ambient air temperature of at least 35°C (95°F). This will
also be adequate for sustained lidar operation in hot summer
weather.
5. The system must have a logarithmic video amplifier, in addition
to a linear video amplifier, within the lidar electronics.
The logarithmic amplifier will amplify weak signals such as
the clear air return signal from behind a dense smoke plume,
to a much greater level than the stronger signals from the
clear air in front of the plume as well as from the plume
itself. This will provide greater accuracy in measuring the
opacity of dense plumes. This is especially true when measuring
the opacity of a particulate plume within polluted areas
(areas of heavy atmospheric burden). In this case, the “clear
air” in front of a dense plume will have a much greater level
of optical backscatter than when the air is reasonably clean
(low pollution level). The return signal from the “clear air”
behind the dense plume, in this case, will be orders of
magnitude below that from in front of the plume and within the
plume.
6. A gating logic capability within the lidar electronics is
required to control backscatter return signal levels within
the receiver, from in front of and within the plume. The
-------
6
position of the gates (3 required) will be variable over the
eflecti.ve lidar range. This capability permits the detection,.
processing, and storage of the plume backscatter signal rather
than blanking it out as is done in the EPA/Research Triangle
Park lidar. There is valuable optical data in this plume
backscatter signal which will be effectively used in the near
future for plume tracking, the detection of the combining of
plumes, plume dispersion and eventually mass emissions rates.
-------
7
III. BACKGROUND
Smokestack’emissions opacity has been an air pollution monitoring
parameter in this country since the early 1900’s. In recent years, the
opacity of visible emissions has been determined by qualified visible
emissions observers being used to gather the required data during
daylight hours. The observers are tested every six months in accordance
with Method 9. They must demonstrate the ability to assign opacity
readings to black and white smoke plumes with an error not to exceed 15%
opacity on any one reading and an average error not to exceed 7.5% in
either smoke color category. For the most part they are limited to
reading emissions during daylight hours with good sunlight.
The lidar is basically a laser radar used in the remote sensing of
air pollution sources. In 1963, a lidar was successfully used to map
laser backscattered echoes (red light) from turbidity in the upper
atmosphere and to measure backscatter from the molecular constituents
and haze in the loweratmosphere. Backscattered light is that fraction
of light which is reflected back to the lidar from the aerosols present
in the atmosphere. The lidar consisted of a ruby pulsed laser as an
optical energy transmitter and a telescope/detector with the associated
electronics as a lidar receiver. The electronics processed the lidar
data and provideda usableoutput to the operator. It also contained a
timing mechanism to measure the round trip time intervals for the back-
scatter echoes from the time the pulse was transmitted from the laser to
the time each echo was collected by the lidar receiver. This provided
the range to each target.
In 1967, research was underway studying the use of lidar asa
remote monitoring instrument for air pollution. Much has been done in
-------
8
using the lidar as a remote monitor for particulates or aerosols being
discharged from smokestacks and for the mapping of aerosol.s which occur
in. the lower atmosphere as a result of ind drift of smokestack emissions.
Today, tunable lasers are being used in lidar systems, in addition
to a multitude of monofrequency lasers extending from the ultraviolet
into the intermediate infrared.
The research and development of the monofrequency lidar has progressed
to the point that it is ready for use in the EPA Enforcement Program to
monitor the opacity of stationary source emissions. Initially, •the
lidar will be used in the Enforcement Program to remotely measure
smokestack plume opacity. It is to that end that this field study was
designed and carried out.
It is noteworthy to mention that lidars are under development for
the detection and quantification of gaseous pollutants such as sulfur
dioxide, ozone, and nitrogen dioxide. They should be available for
field use within EPA in nearly two years.
-------
.9
IV. DESCRIPTION OF THE FIELD EVALUATION TESTS
The field evaluation tests for the lidar were designed to establis.h.
‘the performance characteristics of the lidar which includes calibration
ma clear air medium and to establish a viable means of lidar field
calibration which is easily reproduced from test site to test site. The
tests performed with the lidar are outlined below.
Clean Air Tests
The lidar was fired in the horizontal position on a test range to
obtain the necessary data to determine the state of calibration of the
lidar system electronics. The output of the lidar was displayed on an
oscilloscope. The curve on the scope displays a peak in the clear air
backscatter signal ‘intensity coincident with the convergence point of
the lidar transmitter and, receiver, i.e., they are looking at the same
solid angle in space. This curve then falls off toward zero as l/(range
from lidar to scatterer) 2 . The Mark IX lidar system electronics include
a linear video amplifier, a logarithmic video amplifier (amplifies weak
signals far more than the stronger lidar return signals), and an inverse-
range-squared compensation circuit [ l/(range from lidar to scatterer) 2 ].
The compensation circuit corrects the above mentioned fall-off curve for
range, along the line-of-sight from the lidar, yielding a straight line
on the oscilloscope parallel with the horizontal’axis. This test ,serves
as a performance check for the compensation circuit.
Screen Test
A means of lidar system calibration in the everyday field usage is
a necessity being easy to employ with set-up and take-down time of less
-------
10
than 5 minutes. This screen test must be consistently reproducible.
Field calibration provides a quick check and verification of the accuracy
with which the lidar can measure smokestack plume opacity. This capability
is essential in employing the lidar in the enforcement mission.
The screen used i.n this test was approximately one square meter in
size being constructed with a woodenframe and a fine wire mesh positioned
over the frame. At the test range, the screen was set up nearly 150 rn
(500 ft) from the lidar (at a distance greater than the system convergence
point where the laser and the telescope fields of view are coincident or
completely overlap) permitting a field calibration referenced to the
previously established opacity value of the screen. (Opacity is defined
as one minus the optical transmittance of the screen, plume or any other
target.)
This and manyother lidar screen tests have shown that this calibration
technique is not without problems which produce inaccuracies. Specular
reflection has been isolated as a rather significant source of error
when the lidar is measuring the opacity of a particular screen. Careful
fabrication of the calibration screen may reduce the error to a negligible
level. However, of major concern is the possible degrading effect field
handling can impose on the optical properties of the screens.
The calibration test ultimately adopted for the EPA/NEIC lidar
should yield a lidar calibration accuracy of 2 to 3% during both day and
night hours. It is quite probable that nighttime calibration may be
more effectual since optical scintillation induced by atmospheric turbulence
along the line-of-sight of the lidar will be at a significantly lower
level than that present during the daytime. Localized inhomogeneity in
the air along the lidar line-of-sight is also an error source in attempting
system calibration. Further study will be required to resolve these
issues.
-------
11
Smoke Generator Tests
These tests were performed in an open field during daylight hours
using a.mobile smoke generator provided and operated by the Air Resources
Board, State of California, Sacramento, California. The smoke generator
had a transrnissometer that measured smoke plume opacity across its stack
which was 4.6 in (15 ft) high. Black and white smoke plumes, generated
with known opacities as monitored by the transmissometer, were interrogated
(remotely tested) by light pulses from the lidar. The lidar was positioned
270 m (885 ft) from the smoke generator with the wind blowing nearly
perpendicular to its line-of-sight at 8 to 16 km (5 to 10 mph). The
horizontal visibility was less than 2 km at the test site.
The smoke generator emitted black or white smoke plumes over the
range of 20 to 100% opacity. The opacity values were varied in 5 or 10%
increments with each opacity value or level being interrogated with
approximately 10 lidar pulses. The smoke plumes from the generator were
not uniform but displayed a puffing characteristic. The wind also had
an affect upon the rise (height above stack) of the plume. The lidar
was aimed immediately above the opening of the generator stack to keep
the lidar beam completely within the plume.
The technical details of these tests are documented in Section IV
of theAttachment. The correlation values of smoke generator opacity
(transrnissometer) and lidar measured opacity for white smoke, are plotted
in Figures 5 and 6 of the Attachment. In these plots the opacity values
measured by the instack transmissometer were considered absolute or
having no variation along the abscissa. Careful analysis of Figure 3,
the smoke generator strip chart record, reveals that this indeed was not
the case. A temporal variation in opacity was present in the smoke
generator’s plume as evidenced by the oscillations in the strip chart
record. For the time interval given in Figure 5, Table 1 provides the
-------
12
variation in smoke generator opacity for each nominal opacity value
obtained from Figure 3.
Table
Nominal Value Range of Difference in
of Opacity Variation Opacity
( %) %) (%)
20
18—21
3
30
28-32
4
40
37-41
4
50
60
59-62
3
‘
70
68-71
3
80
78-81
3
90
89-91
2
100
99-100
1
t Figure 3 Data) White Smoke, 1325—1400 PST, Dec. 3, 1975
The spread ranges from 1% at the opaque end of the scale to 4% at
the lower levels of opacity. In addition to the above, there was also
the inaccuracy inherent to the smoke generator transmissonieter in measur-
ing the smoke opacity in the stack. Likewise, the response time of the
transmissometer to the puffing or changes in plume particulate density
is not of such a value to clearly document the full extent of the
variations in opacity. The lidar measures plume opacity in a time
interval of 30 nanoseconds (1 nanosecond = l0 seconds) while the
response time due to signal integration in the instack transmissometer
is in the order of 5 to 8 seconds. This time is more than io8 times
greater than that of the lidar. The lidar detects the instantaneous
plume opacity, be it in either a dense area or in a less dense area of
the varying smoke plume. The transmissometer averages the plume opacity
values over the 5- to 8-second time period and is not capable of detect-
ing the short lived peaks and valleys of the smoke plume opacity profile.
Thus, the magnitude of the plume variations was not fully documented by
the instack transmissometer.
-------
13
All of these parameters outlined above are effective sources of
error within the instack transmissomete.r and the effect is manifested in
the divergence of the smoke generator data and the lidar data. Likewise,
Table 2 depicts the variation in smoke generator opacity of the data
presented in Figure 3for the time interval given in Figure 6.
Table
Nominal Value
of Opacity
Range of
Variation
Difference
Opacity
in
•
20
30
27—32
5
40
37-43
6
50
47-53
6
60
•
56-63
7
70
67-73
6
80
78-82
4
90
.
89-91
2
100
99-100
1
t Figure 3 Data, White Smoke, 1415—1425 PST, Dec. 3, 1975
The variation pattern is nearly the same as that described for
the Figure 5 data. The variation pattern for the black smoke data
contained in Figure 3 is provided in Tables 3 and 4.
Table
Nominal Value Range of Difference in
of Opacity Variation Opacity
50 46-60 14
60 50-70 20
70 63-80 17
80 69-85 16
90 84-94 10
95 92-98 6
1- Figure 3 Data, Black Smoke, 1400 to 2410 PST, Dec. 3, 1975
-------
14
Table
Nomi
of
nal Value
Opacity
•
Range of
Variation
Difference in
Opacity
20
18-23
5
30
.
26-34
.8
40
37-42
5
50
45-54
9
60
51-66 .
15
70
63-75
12
t Figure 3 Data, Black Smoke, 2435 to 1450 PST, Dec. 3, 2975
The difference in opacity values about the respective nominal
values for the black smoke displayed a variation pattern similar to that
of white smoke. However, the magnitude of this difference was several
times greater than that for white smoke. Theblack smoke plumes also
appeared to be more turbulent or puffy than the white smoke plumes.
As was mentioned earlier in this report, the horizontal visibility
in this location during the. time of these tests, was less than 2 km.
This became a •significant factor in the smoke readings obtained by
visible emission observers.. The background sky was light gray due to
the heavy atmospheric burden (air pollutants). The VEO readings for
white smoke were consistently at least 20% lower than those obtained
with the lidar and the smoke generator transmissometer. Theblack smoke
presented a high color contrast against the background sky, and therefore
greater accuracy was achieved with fldar/VEO variations of nominally 5
to 10%.
Due to the low horizontal atmospheric visibility, this test also
afforded an evaluation of the lidar receiver gating logic for controlling
incoming signal intensities resulting from optical energy backscatter
from the clear air in front of, and behind the plume. Signal control
for backscattered energy from within the plume was likewise evaluated.
-------
15
Lidar Tests at Kaiser-Permanents
Both day and night tests were carried out with the lidar at the
Kaiser-Permanente Cement and Gypsum Corporation, Cupertino, California,
on Dec. 2, 1975. Three smoke plumes were interrogatedwith the lidar.
•The three stacks shown in Figure IV-1, were labeled #1 Stack, West Stack
and East Stack. These stacks discharge emissions originating from
cement kilns. Exhaust from #2 Kiln is discharged through the East Stack
and the #5 Kiln through the West Stack. These stacks are 64 m (210 ft)
high with 6.1 m (20 ft) exit diameters. The emissions from #1 Kiln pass
through an electrostatic precipitator (esp) and are discharged through
#1 Stack which is 38.1 m (125 ft) high with an exit diameter of 2.74 m
(9 ft). Emissions from #6 Kiln were discharged through an electrostatic
precipitator and out #6 Stack which is 30.5 m (100 ft) in height and has
an exit diameter of 2.44 m (8 ft). This stack was not in operation at
the time of this study. The remainder of the kilns, #3 and #4, discharge
into two American Filter baghouses each having ten units. The baghouse
emissions were not observed with the lidar.
The particulate plumes from these stacks drift a significant
distance into the valley below [ Fig. IV-2]. The stack emissions from
the East and West Stack were photographed on October 29, 1975 near
sunset during the preliminary test site visit.
The three plumes are known to contain a significant amount of water
vapor which condenses near the exit of the stack. The water vapor would
affect the VEO readings because water aerosols decrease light transmittance
(increase plume opacity) just as particulates do. The aerosols also
have a significant effect upon sunlight scatter as shown in Figures IV-3
and IV-4. In the former, the photograph was exposed from a hillside
above the plant looking east while the sun was low in the west, with the
valley terrain as a background. This arrangement would be in accordance
with the viewing requirements of Method 9.
-------
EPA /NEIC
WEST STACK
EAST STACK
/.
FIGURE IV-1 KAISER-PERMANENTE CEMENT CO. STACKS
#1 STACK
PHOTOGRAPHED OCTOBER 29, 1975 NEAR
SUNSET
-------
EPA/NE IC
FI6URE IV-2 KAISER-PERMANENTE
STACK EMISSIONS AND
CEMENT CO.
SUBSEQUENT DISPERSION
PHOTO6RAPHED
OCTOBER 29, 1975 JUST AFTER SUNSET
-------
EPA/N EIC
FIGURE IV-3 VIEW OF PARTICULATE PLUMES FROM THE EAST
AND WEST STACKS LOOKING EAST WITH THE SUN LOW IN
THE WEST (SUN LIGHT BACKSCATTERED OFF PLUME)
PHOTOGRAPHED OCTOBER 29, 1975
WEST STACK’ NEAST STACK
-------
EPA/NE IC
DUAL PLUME
IEW OF PARTICULATE PLUMES FROM THE EAST
STACKS LOOKING WEST WITH THE SUN LOW
(SUN LIGHT FORWARD-SCATTERED FROM PLUME)
EAST AND WEST STACKS IN-LINE
PHOTOGRAPHED ON OCTOBER 29, 1975
PARTICULATES+WATER AEROSOLS
EAST STACK PLUME
N
RESI
WEST STACK PLUME
FIGURE
AND
IN THE
IV-4 V
WEST
WEST
-------
16.
However, the line of demarcation cannot be clearly defined between
the region of the plume containing particulates and condensed water
aerosols, and the residual region of the plume where the condensed water
microdroplets evaporatedleaving only the particulates. This presents a
major problem because the water aerosols are not pollutants, but the
particulates are so classified. Regulations only apply to the particulates
as related to plume opacity. In Figure IV-4, the stacks were photographed
looking west with the sun low in the west. The plumes from the east and
west stacks appear much brighter and pronounced because the forward-
scatter of sunlight by the particulates and water aerosols is far
greater than solar backscatter [ Fig. IV-3]. Also shown in this figure
are the two regions of the plume mentioned above which were not discernible
in Figure IV-3 (these two photographs were recorded only minutes apart).
The plumes. from the East, West and #1 Stacks were interrogated with
the lidar positioned south of the stacks. The line-of-sight was nearly
perpendicular to the common-line of the East and West Stacks. The exit
plume’ opacity of each stack was measured and was found to have the,
fôllowiñg ranges of opacity [ Table 5]:
Table 5
K4ISER SMOKESTACK EXIT PLUME OPACITY
Stack
Opacity Range
Dominant Value
East Stack
35-75
65
West Stack
20-75
50
#1 Stack
20
20
The plume was drifting in a southerly direction due to a northerly
breeze present at the time of test. Many times during this study the
smoke plumes would drift as a slowly dispersing entity to elevation
levels below the exit height of the stacks and coincident with the
-------
line-of-sight of the lidar and VEO”(furnigating lapse condition). This
occurrence completely precluded the smoke reading of the exi.t’plume with
the VEO’s. However, the lidar is range-resolved; the range or distance
between the lidarand the plume(s) and beyond the plumes is segmented
into equal intervals or range cells usually on the order of 3 to 15
meters in length. Thus, the lidar can measure the exit plume opacity
even with a descending plume condition, if there is at least one range
cell between the descending plume condition, if there is at least one
range cell between the descending plume and the exit plume. This
condition is required in order to obtain a clear air return signal,
which is needed from in front of and behind the exit plume for the
opacity calculation [ Chapter II, Attachment]. This is depicted in the
lidar cross section [ Fig. 10, Attachment] provided as (a)2. One can
easily see the separation of the two plumes in the lower center of the
photograph. There was ample clear air return range between the two
plumes, permitting calculation of the opacity of the rear (exit) plume.
The ‘oscilloscope photographs of Figure 10 [ Attachment I] depict the
two aforementioned sections or regions of a plume, namely, the plume of
particulates and water aerosols, and the plume of only particulates.
The white area in each photograph is the particulate/aerosol plume while
the light gray/medium gray area is the residual plume, or the plume
consisting of only particulatés remaining after the water aerosols had
evaporated. These lidar cross sections were made after sunset into the
early evening.
Lidczr Tests at Occidental Fertilizer Company’ and Libby-Owens-Ford (LOF)
Company
After the smoke generator/lidar tests were completed, in late
afternoon on Dec. 3, 1975, the lidar was used to interrogate four stacks
located within the Occidental and LOF facilities in Stockton (Lathrop),
-------
18
California. These two facilities are adjacent to each other. One of
the stacks was on the Occidental facility (Stack 1) while Stacks 2, 3
and 4 were on the LOF facility. From after sunset i.nto darkness, over
120 lidar shots were made, measuring the, opacity of the plumes from the
four stacks and the az.imutha.l cross sections of- the plumes. from Sta ks.
2, 3, and 4 since they were in ‘proximity to each other. The data for
each stack plume is presented in Table 6.
Table 6
INDUSTRIAL SMOKESTACK OPACITY
Stack Opacity Average No. of Lidar Shots
No. Range Opacity
‘
1
19-42
30
24
2
40-95
71
30
3
83-99
90
36
4
90-98
92
33
Stacks 1 and 2 displayed a rather large variation in opacity
values. There was a puffing characteristic in the exit,plunies. The
exit plumes from Stacks 3 and 4 being larger in diameter than from
Stacks 1 and 2 were much more consistent in opacity values. The Cali-
fornia Air Resources Board (ARB) said during the tests that the plumes
from Stacks 2, 3 and 4 contained condensed water aerosols characteristic
of, industrial process at LOF. The same characteristic observed in the
Kaiser-Permanente facility is present in the azimuthal. cross-sectional
scan of Figure 12 of the Attachment. The white areas within the plume
structure had an opacity of nearly 100% when the scan was performed,
possibly consisted of both particulates and condensed water aerosols.
The light/medium areas were the residual’ plume wherein the aerosols were
not present due to evaporation.
-------
19
The lidar azimuthal scan also depicts where the three individual
plumes were combining, the direction of drift, (to upper right), and the
dispersion characteristics along the drift path.
‘The technical details of this test are given in Section VI of the
Attachment. ‘
-------
ATTACHMENT
LIDAR APPLICATIONS FOR SMOKE PLUME OPACITY MEASUREMENTS
STANFORD RESEARCH INSTITUTE
MENLO PARK, CALIFORNIA
-------
_ ©
1 bc U® °
1
Final Report December 1975
LIDAR APPLICATIONS FOR SMOKE PLUME
OPACITY MEASUREMENTS
By: EDWARD E. UTHE
Prepared for:
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
OFFICE OF ENFORCEMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
DENVER, COLORADO
ATTENTION: A. DYBDAHL
EPA PURCHASE REQUISITION NO. TS-129
SRI Project No. 4723
Approved by:
R. T. H. COLLIS, Director
Atmospheric Sciences Laboratory
RAY L. LEADABRAND, Executive Director
Electronics and Radio Sciences Division
-------
CONTENTS
I INTRODUCTION . . . . . 1
II BACKGROUND A1’ D METHOD OF APPROACH . . 2
III LIDAR CALIBRATION TESTS 6
IV SMOKE GENERATOR TESTS 10
V KAISER-PERMANENTE TESTS 19
VI OCCIDENTAL AND LIBBY-OWENS-FORD TESTS 22
VII CONCLUSIONS AND RECOMMENDATIONS . . . 25
APPENDIX A 28
REFERENCES .38
1
-------
I INTRODUCTION
The National Enforcement Investigations Center (NEIC), Office of
Enforcement, U.S. Environmental Protection Agency, Denver, Colorado,
contracted with Stanford Research Institute for remote sensing services
to evaluate and demonstrate the application of lidar instrumentation for
the collection of data in support of the enforcement monitoring program
for air quality. Because of budget constraints, the field program was
limited to three observational days, using the already existing capabili-
ties of the SRI Mark IX Mobile Lidar System. This report presents the
major results of this field investigation and of a limited data analysis
effort.
1
-------
II BACKGROUND AND METHOD OF APPROACH
SRI has pioneered the use of lidar in the remote observation of
stack plume optical and physical densities and plume transport and
diffusion characteristics. The first lidar measurements in the lower
atmosphere were made by SRI in the summer of 1963. Approximately a year
later, Dr. N.G.H. Ligda suggested a technique to remotely measure smoke
plume densities by comparing the lidar returns from the clear air before
and after the penetration of the plume by the laser energy. The plume
opacity 0 (actually one minus the transmission at the lidar wavelength)
is then given as:
0=1- 1 1 ) 1
where
10 = backscatter signal before the plume return
I = backscatter signal after the plume return.
The plume opacity technique was shown feasible in a series of tests
conducted for the Edison Electric Institute by Fernald and Collis (1965)
and a mobile lidar system was designed by Evans (1967) for application
of the technique. The system was constructed by General Electric for
the Environmental Protection Agency (EPA) and has been extensively eval-
uated on EPA research programs (Cook, Bethke, and Conner, 1972).
A difficulty with the opacity technique is that the large plume
return may saturate the sensitive photomultiplier detector and, because
of the detector’s relatively long recovery time, render accurate measure-
ments of the far side clear-air backscatter impossible. The EPA lidar
has been modified to overcome this difficulty by electronically gating
2
-------
out the plume return. Another difficulty is the accurate observation of
the clear air return from the far side of the plume for dense plumes
because of the relatively low intensity signals that result. The approach
taken by SRI is to use a video logarithmic amplifier that suppresses the
large plume return while amplifying the less intense clear air returns.
In addition, the receiver gain is electronically controlled in three
steps. For the plume opacity measurement, the first step is adjusted to
reduce the receiver gain out to the range of the plume. This first step
prevents the “clear air” return originating from near the lidar from
saturating the detector and log amplifier. The second gain reduction
step is triggered at the end of the first step and further reduces
receiver gain to prevent saturation by the plume return. The attenua-
tion and range intervals of the gain reduction steps are adjustable so
that both clear’ air and plume returns can be observed ‘and evaluated. At
the end of the second step, the receiver gain is turned on to full value
so that clear air return may be observed at maximum receiver sensitivity.
As shown in the following sections of this report, the log amplifier aids
in the observation of the signal from beyond the plume.
Another approach to plume density measurement is to relate the
backscatter from the plume to the plume density. In a series of experi-
ments conducted by dispersing plume particulates (fly ash) of known size,
shape, concentration, and refractive properties in a specially designed
aerosol chamber--allowing unobstructed lidar observations of aerosols in
realistic plume geometries--Uthe and Lapple (1972) were able to demon-
strate that the plume return for a 700-nm wavelength lidar is well
correlated with the plume opacity (one minus the transmission of light
with wavelength dependance identical to that of the eye response),
irregardless of the particle size. In addition, the plume return for a
1060-nm wavelength lidar is well correlated with plume physical density
irregardless of particle size. The plume backscatter technique would
3
-------
place fewer requirements on the lidar instrumentation than the clear air
opacity technique.
Multiple lidar signatures recorded while the lidar scans in either
evaluation or azimuth directions can be used to derive space and time
variations of plume density distributions. These data can then be
analyzed in terms of plume transport and diffusion characteristics as a
function of local topography and meteorological conditions. Initially,
the use of lidar in these roles relied on the handprocessing of lidar
signatures (signal intensity recorded as a function of range) into two-
dimensional isoscattering diagrams (Johnson, 1969). Later, the signa-
tures were hand digitized and computer analyzed using polar-coordinate
plotting routines (Johnson and Uthe, 1971). The computer processing of
digital records also facilitated the correction of the plume cross
sections for the attenuation effect and the computational inference of
cross-plume integrated densities (Johnson and Uthe, 1971).
The video disk technique of recording and processing lidar signa-
tures was introduced by SRI as an inexpensive means of electronically
generating pictorial displays of aerosol structure from recorded lidar
data (Allen and Evans, 1972). Johnson et al., (1973) used the video
disk technique to investigate plume geometric properties in both urban
and rural environments. While the technique gave a readily available
means to view plume structure in both space and time, the disk records
were not suitable for quantitative density evaluations because of band-
width problems.
The rapid development of commercial high-speed digitizers, mini-
computers, and digital display systems now make digital data processing
capabilities generally available at relatively low cost. Recently, SRI
has taken advantage of commercially available components to construct a
digital real-time lidar data recording, processing, and display system
4
-------
for its Mark IX mobile lidar facility (Uthe and Allen, 1975). The system
provides much higher signal proèessing accuracy than the video disk
technique and provides both real-time analysis and display of processed
lidar data. The system has been applied to the remote observation of
cirrus clouds and urban boundary layers but only recently has been
applied to plume studies. A journal reprint describing the capabilities
of the Mark IX digital system is reproduced in this report as Appendix A.
On 2 December 1975 the Mark IX lidar was operated at the Kaiser-
Perinanente cement plant located in Cupertino, California. The lidar was
operated without the gain reduction steps. All data collected on this
program were recorded on the digital system for both real—time and
subsequent data analysis. Inspection of the results obtained at Permanente
showed that the usefulness of the data obtained by the lidar without the
gain reduction steps was limited to plume opacities of 0.7 or less.
On 3 December the lidar was operated near Stockton, California.
The State of California Air Resources Board operated a calibrated smoke
generator in support of the lidar tests. Plumes from. 20 to 100 percent
opacity were generated with both white and black smoke. Later, the lidar
was used to observe plumes from Occidental and Libbey-Owens-Ford
manufacturing plants. The logarithmic amplifier and gain reduction
steps were used for all data collected on 3 December. In addition to
the plume opacity measurements, the lidar was scanned in both elevation
and azimuth angles to produce vertical and horizontal cross sections of
plume structure.
On 4 December the lidar was operated at the Stanford Field Site to
obtain clear air backscatter data for calibration of the log amplifier
response function and to calibrate the receiver gain reduction steps.
The results of these field observations are presented in the following
sections of this report.
5
-------
III LIDAR CALIBRATION TESTS
A receiver of large dynamic range is required to accurately observe
backscatter from the clear air on the near and far side of the plume and
from the plume particulates. Figure 1 presents the relationship between
opacity and the near and far side clear air returns (signal difference
in dB = 10 log I/Is) . For a plume of 90 percent opacity, a signal
difference of 20 dB (or a factor of 0.01) must be resolved by the lidar
receiver. In addition, the plume return may be as much as 30 dB above
the near side clear air return. For these reasons, a video logarithmic
amplifier and receiver gain reduction steps were thought necessary for
observation of dense plumes.
The response of the Mark IX logarithmic receiver and gain reduction
steps were calibrated by repeatedly firing the lidar into clear air
while reducing the receiver gain in known steps by inserting neutral
density filters in front of the detector. The Biomation 8100 digitizer
settings during this calibration were:
Voltage setting = 0.2 V
Voltage offset = - .79
Delay setting = .04
Sampling rate = .01 p s
Biomation counts recorded at the same range were averaged for 10 laser
firings. The results for various neutral density filters and receiver
gain reduction steps are shown in Figure 2. These results show that the
Biomation counts can be related to light input to the receiver by the
value 0.2 dB/count up to a value of 70 counts--above which a non-linear
response occurs. Since the noise level of the receiver was determined
as -105 counts, the receiver has a linear logarithmic response over a
6
-------
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.I
0
-20
SIGNAL DIFFERENCE — dB
FIGURE 1 PLUME OPACITY AS A FUNCTION OF THE SIGNAL DIFFERENCE (in dB units)
BETWEEN THE NEAR AND FAR SIDE CLEAR AIR RETURNS
C)
0
-IS -10 -5 0
-------
3 dB gate =
5 dB gate =
10 dB gate
15 dB gate =
I I I I
-40 -20
-100 -80 -60
0
BIOMATION COUNTS
20
40
80 100
N
N
N
N
N
N
-20
-15 —
-10 —
-5 —
0
•0
z
U
4.97 dB
7.48 dB
12.95 dB
17.28 dB
Gate Values
Noise level -105 counts
Amplifier linear range = 34 dB
N
N
0
N
N\
N
60
FIGURE 2 CLEAR AIR CALIBRATION OF LOG AMPLIFIER AND GAIN REDUCTION STEPS (4 December 1975)
-------
range of 34 dB. From the calibration data presented in Figure 2, the
gain reduction steps were d termined as:
Step marked 3 dB = 4.97 dB
Step marked 5 dB = 7.48 dB
Step marked 10 dB = 12.95dB
Step marked 15 dB = l7.28dB
Since the Biomation has a count range of -256 to +256, the voltage reso-
lution could be improved by adjusting Biomation settings; however, the
resolution and dynamic range is satisfactory for the plume studies
conducted under this program.
9
-------
IV SMOKE GENERATOR TESTS
The State of California Air Resources Board operated a calibrated
smoke generator in support of the lidar evaluation tests. Black and
white plumes were generated with known opacities as measured by an across-
the-stack transmissometer. Transmissometer records for the period of the
lidar tests are shown in Figure 3.
The Mark IX lidar van was located 270 meters from the smoke generator
and the lidar beam was aligned on the plume near the top of the generator
stack. Approximately 10 backscatter signatures were recorded for each
opacity setting. Typical backscatter signatures as recorded on tape and
displayed on the digital monitor (see Appendix A) are shown in Figure 4.
These signatures were digitally corrected for the inverse range squared
dependence. Figure 5 presents a plot of the lidar observed opacities
against the transmissometer opacities for the white smoke generated
during the time’ period 1325-1400 PST. These data illustrate that the
lidar technique produces values that are in general agreement with the
transmissometer values over the range of 20 to. 100 percent opacity. The
scatter in the data are primarily a result of plume spreading and a
resulting reduction in plume opacity downwind of the stack emission; the
data of Figure 5 show that the lidar derived values are generally lower
than the transmissomete values. Another major source of error can be
introduced by differences in the backscatter properties of the clear air
in front of and behind the plume. If the plume particulates increase the
clear air backscatter in front of the plume relative to that from behind
the plume, the lidar observed opacity will overestimate the true opacity.
If the clear aLr behind the plume produces greater backscatter than the
10
-------
I I 100
Black Smoke 90
80
70 c
a)
0
60
k mok
50 I
h te Sm ke
40
30
0
- 20
- 10
-0
1435 1425 4 5 1400
.- -- TIME - hours
I—. - 100
I••9°
__________ ______________________ ______ 80
______________________________________________________________ 70 c,
a
___________ ___________________ _____________ ___________ 0 -
___________________ ___________________________________________________ k
______ ______ ___________________ _________________ _______________________________ 60 a
______ _________________________________ __________________________ White Smok ____________________ 50
= ____________________________________________ 40
30
-____ ____ 0
_________ _________________________________________________________________________________________ 20
— ___________________ 10
- p - 0
1335 1245
- TIME - hours
FIGURE 3 STRIP-CHART RECORD OF PLUME OPACITY AS MEASURED BY TRANSMISSOMETER
-------
Rec.eivei gain reduction
CLEAR AIR
(No plume)
LOW OPACITY
(White smoke)
HIGH OPACITY
(White smoke)
LOW OPACITY
(Black smoke)
FIGURE 4 EXAMPLES OF LIDAi BACKECATTER SIGNATURES OBSERVED
ON DIGITAL ECREEN
15 dB dBI I
0 dE.
HIGH OPACITY
(Black smoke)
12
-------
100
90
80
70
60
50
40
30
20
10
10 20 30 40 50 60 70 80 90 100
OPACITY (transmissometer)
FIGURE 5 OPACITY OF WHITE SMOKE AS MEASURED BY
AGAINST OPACITY FROM TRANSMISSONETER
THE LIDAR PLOTTED
0
0
13
-------
clear air in front of the plume, the lidar observed opacity will under-
estimate the true opacity.
Figure 6 presents the lidar derived opacities plotted against the
transmissometer opacities for the data collection period of 1415-1425 PST.
These data points show more scatter and more values above the expected
45° line than for the data presented in Figure 5. It was visually
noticed that the plume was more “turbulent” in the later data run, and
this probably resulted in the lower correlation and greater lidar-observed
opacities (especially for the 30, 40, and 50 percent levels) presented
in Figure 6.
Figure 7 presents a plot of the plume returns (normalized to the
near side clear air returns) against the trarismissometer measured
opacities for the same data presented in Figure 6. These data also show
denser plumes for the 30, 40, and 50 percent opacities than expected
from the remaining data.
The plume returns plotted against the lidar-observed opacities are
shown in Figure 8. These data clearly show that the lidar derived
opacities are consistent with the lidar observed plume densities, and
therefore, the lidar opacity technique is considered valid, provided the
lidar observes a representative volume of the plume.
The black smoke plumes were more variable in terms of opacity than
were the white smoke plumes (see Figure 3). For this reason, the
opacities derived from the lidar are not presented here. However, Figure
9 presents the plume returns as a function of lidar-observed opacities
for the black smoke that was generated beginning at 1435 PST. It is
seen that good correlati,on is again obtained and that the plume returns
are approximately 5 dB (or a factor of 3) less than for white smoke.
Therefore, these data well illustrate the sensitivity of the plume back-
scatter to the absorption properties of the plume particulates.
14
-------
100
90
80
70
.H60
50
40
30
20
10
10 20 30 40 50 60 70 80 90 100
OPACITY (transmissometer)
FIGURE 6 OPACITY OF WHITE SMOKE AS MEASURED BY THE LIDAR PLOTTED
AGAINST OPACITY FROM THE TRANSMISSOMETER
WHITE SMOKE
1415 - 1425 PST
3 December 1975
S
S
S
.
.
.
S
S
I
S
I
S
I
S
S
S
I
I
S
I
S.
S
I
$
S
.
I
I
I
S
S
$
S
S
I
I
I
I
I
S
I
S
S
I
I
I
®= Two
or more data points
15
-------
I I I I I I
36 — WHITE SMOKE
1415 - 1425 PST
3 December 1975
35 —
34 —
S
S
.
: .
Two or more data points
26 I I
10 20 30 40 50 60 70 80 90 100
OPACITY (transmissometer)
FIGURE 7 LIDAR OBSERVED PLUME RETURNS (normalized to near side
clear air returns) PLOTTED AGAINST OPACITY FROM
TRANSMI S5OMETER
16
-------
I I I I I
.
.
33 - WHITE SMOKE
.
1415 - 1425 PST •
3 December 1975 • • •
5
32 —
• S
•. S
5 5 I
S •• •
31— • ••
• 5 •
. S
S S
30— • : U
S S
• •‘
29 —
28 —
S •
27 • = Two or more data points
S
26 I I I
10 20 30 40 50 60 70 80 90 100
OPACITY (lidar)
FIGURE 8 PLUME RETURNS PLOTTED AGAINST OPACITIES OBSERVED FROM
THE LIDAR (white smoke)
17
-------
I I I I, I
29- . —
BLACK SMOKE . .
S S
28 1415 - 1425 PST
- 3 December 1975 •
S S •S
27-
.•‘
S •
. •.
26- S
.
S
. . .
25 - .
S
.
24- •
.
23- ‘SI.
22 -
21 —
S
20
19 —
S
18 — 0 = T o or more data ‘points
S
17 I I I ‘
10 20 30 40 50 60 70 80 90 100
OPACITY (lidar)
FIGURE 9 PLUME RETURNS PLOTTED AGAINST OPACITIES OBSERVED FROM
THE LIDAR (black smoke)
18
-------
V KAISER -PE BMANENTE TESTS
On 2 December 1975 the Mark IX lidar was used to observe stack
emissions from the Kaiser-Permanente Cement Company, located in Cuper-
tino, California. These initial observations were confined to testing
the suitability of log and linear response receivers and analog range
correction. The receiver gain reduction steps were not used. Three
plumes were observed--one from a short stack known as Stack 1 and two
larger stacks known as West Stack and East Stack. Observations on the
East Stack plume gave opacities ranging from 35 to 75 percent with most
values near 65 percent. Observations on the West Stack plume gave
values of 20 to 75 percent with average values near 50 percent. Stack 1
values were near 20 percent opacity. These opacities are the real-time
values computed and printed by the lidar digital system.
Because of possible lidar saturation effects, difficulties in
observing the clear air beyond the plume due to a nearby hill, and the
limited analysis time devoted to this project, these returns have not
been reanalyzed. However, these tests firmly indicated that the log
receiver and gain reduction steps were required for plume opacity mea-
surements of dense plumes in low-visibility atmospheres.
In addition to the plume opacity observations with the lidar
pointing at a fixed location over the stack, scanning lidar observations
were made to derive both vertical and horizontal cross sections of plume
structure. The lidar signatures were recorded on DECTAPE with analog
range correction applied (see Appendix A). Elevation and azimuth pointing
angles were also recorded. The data were then played back through the
digital system and displayed on the TV monitor in polar-coordinate form.
19
-------
Figure 1OA presents two elevation scans made sownwind (to the west) of
the East and West Stack plumes. The vertical and horizontal scale marks
are drawn at spacings of 75 m. The maximum angle of the elevation scans
was restricted by the lidar van openings in the direction of observation.
These cross sections clearly show the rise ofthe plumes and their
transport towards the lidar van (the van was located on the south side
of the plumes). Figure lOB presents two azimuth scans from west (bottom
of the picture) to east. These data show the transport and diffusion
of the plumes to the west and in the direction of the lidar. Although
quantitative estimates of downwind plume densities can be made from
these data, this was not attempted in this program.
20
-------
FIGURE 10 LIDAR DERIVED VERTICAL (a) AND AZIMUTH (b) CROSS SECTIONS
OF PLUME STRUCTURE ( 2 December 1975; Kaiser-Permanente)
(a) VERTICAL CROSS SECTIONS
( 75 m between scale marks )
(b) AZIMUTH CROSS SECTIONS
( 75 m between scale marks )
21
-------
VI OCCIDENTAL AND LIBBY-OWEN -FORD TESTS
Following the smoke generator tests on 3 December, the lidar was
used to measure opacities of smoke plumes originating from four stacks
(called Stacks 1-4) located within the Occidental and Libby-Owens-Ford
manufacturing complexes near Stockton, California. The results are
shown in Figure 11. The opacities from Stack 1 varied from 20 to 40
percent with an average around 30 percent. Stack 2 observed opacities
were larger and showed more variation (from 40 to 95 percent) than those
for Stack 1. Assuming the 40 to 60 percent opacities are a result of
the laser beam not adequately positioned on the plume, an average
opacity of 80 percent is obtained. The plumes from Stacks 3 and 4 were
larger in diameter and, consequently, the lidar beam was more easily
positioned so that the full plume was sampled on each lidar firing.
The Stack 3 and 4 observations gave average opacities of 90 and 93
percent. These readings were obtained using the log amplifier and gate
reduction steps as employed on the smoke generator tests.
Figure 12A presents an azimuth scan (from left to right, but
displayed from right to left in the figure) across (and about 5 m above)
the top of Stacks 2, 3, and 4 with 150 m between vertical and horizontal
scale markings. Figure 12B presents the same data, but with 75 m between
scale markings. These data show that Stacks 3 and 4 effectively attenuate
the “clear air,” causing minimal plume returns from beyond the stack,
while this is not the case for Stack 2. This is consistent with the
data presented in Figure 11. The transport and diffusion of the stack
pollutants downwind are easily visualized from the presentations in
Figure 12.
22
-------
100
S
S S • S S • •.•
S • •
S •5.• •••••SSS•% • .1
90 • • ••S•. •. • • • • S •S • S •
• • S • •
S
• STACK4
S
80 - STACK 3
• S S
• S
I
70-
1 -4
60 -
H
S
I
S
N.)
H
L) 50 - •• •
0
• S
40 -
STACK 2
S ••S
S •
30 -
• S S S S
• S • 5
• S.
S
20 -
S
STACK 1
10
FIGURE 11 LIDAR OBSERVED OPACITIES FOR OCCIDENTIAL (Stack 1) AND LIBBY-OWENS-FORD (Stack 2,3, and 4) PLUMES
-------
(a) AZIMUTH CROSS SECTION
( 150 m between scale marks )
(b) AZIMUTH CROSS SECTION
(plume numbers indicated)
( 75 m between scale marks )
FIGURE 12 LIDAR DERIVED AZIMUTH CROSS SECTION OF PLUME STRUCTURE
( 3 December 1975; Libby-Owens--Ford )
24
-------
VII CONCLUSIONS AND RECOMMENDATIONS
This very limited study was conducted to investigate applications
of the lidar technique to the remote measurement of opacity of stack
emissions and the observation of plume density distributions downwind of
the stack. The SRI Mark IX mobile lidar system was used to observe
smoke plumes generated from both manufacturing plants and a calibrated
smoke source. It was demonstrated that the following specialized
features of the Mark IX lidar system greatly aided the plume observa-
tions:
• The log amplifier (with nearly 40 dB of dynamic range)
allows effective observation of both the clear air
and plume particulates.
• The use of receiver gain reduction steps can extend
the dynamic range and solves some electronic satura-
tion problems. The two-step gating allows plume
opacity measurements in low-visibility atmospheres.
• The advanced capabilities of the digital system
permit both real-time and subsequent data processing
and display of backscatter information for effective
and accurate data collection and analysis.
The results presented in this report demonstrate that the lidar
techniques of observing plume opacity (from the near and far side clear
air returns) and plume density (from the plume returns) are valid, and
can accurately be made with a lidar system such as the SRI Mark IX.
The possible difficulties that were evident in this study included:
• Obtaining a representative sample - The shot-to-shot
variation in plume opacity and backscatter was
primarily a result of plume density variations intro-
duced by the turbulent air flow near the top of the
stack. Averaging over many lidar observations would
tend to underestimate plume opacity if the beam does
25
-------
not intercept a representative sample on each firing.
• Clear air inhomogeneities - The “clear” air in low
visibility atmospheres (such as near manufacturing
complexes) can have large density (backscatter)
gradients. Since the opacity measurement requires
the assumption of clear air homogeneity in the
vicinity of the plume, large errors will result if
large irthomogeneities occur.
• Backscatter dependence on particulate properties - The
backscatter from a plume is sensitive to particle
adsorption (as evident from the results from the white
and black smoke observations presented in this report),
as well as to particle shapes and sizes. The inference
of plume opacity or mass concentration from plume back-
• scatter requires knowledge of the relationships between
optical and physical properties and their dependence on
particle size, shape, and absorption. Other problems
such as multiple scattering and beam defocusing also
require consideration.
It is recommended that the evaluation of lidar as applied to plume
opacity and density measurement be conducted using a plume of particles
with known number concentration sizes, shapes and refractive properties.
Only then can experimental accuracies and limitations be properly
assessed. The smoke generator used in this study was useful for demon-
strating lidar concepts, but is not suited for lidar calibration or
research purposes because of the small size of the generated plume.
Figure 13 illustrates a large-scale aerosol chamber designed
especially for lidar studies (Uthe and Lapple, 1972). It is suggested
that a program to investigate limitations and accuracies of the lidar
technique for various optical and physical plume properties could
economically be conducted with our existing lidar and aerosol chamber
instrumentation.
26
-------
BLACK TARGET
FIGURE 13
DIAGRAM OF EXPERIMENTAL SYSTEM FOR EVALUATING LIDAR
TECHNIQUES FOR MEASUREMENT OF PLUME OPACITY AND DENSITY
LIDAR
AEROSOL CHAMBER
Dust, feeder
Compressed
TA-653583-1 7
Air
FAN
27
-------
Appendix A
A DIGITAL REAL-TIME LIDAR DATA RECORDING,
PROCESSING AND DISPLAY SYSTEM
28
-------
Optical and Quantum Electronics 7 (1975) /21—129
Review
A digital real-time lidar data recording,
processing and display system*
E. E. UTHE, R. J. ALLEN
Stanford Research Institute, Menlo Park, California, USA
Received 21 November 1974
Application of laser radars to meteorological programs has been limited by the absence of
suitable data recording, processing and display techniques. This paper discusses a digital
data system that provides the means of performing real-time analysis and display of lidar
data. The system is constructed of commercially available components and makes optimum
use of available software. Application of the system to the real-time viewing of cirrus cloud
structure and inference of cloud density is presented. A review of previously used lidar
data handling techniques is also presented.
1. Introduction
Lidar (Light Detection and Ranging) uses laser
energy in radar fashion to effectively observe
remote atmospheric constituents. Light elastic-
ally backscattered from molecules (Rayleigh
scattering), suspended particulate matter, and
cloud drops and crystals (Mie scattering) can
be employed to derive atmospheric structure
over extended volumes. In addition, quantitative
estimates of atmospheric densities can be made
when certain atmospheric conditions exist or
when relationships between the optical and
physical properties of the scattering medium
can be assumed. The Mie scattering technique
applied to air pollution measurement has
previously been discussed in this journal by
Collis and Uthe [ I].
The application of Mie scattering lidar
techniques to meteorological probing has been
limited over the last ten years by a lack of
appropriate data recording, processing and dis-
play techniques, resulting primarily from the
wide bandwidth and large dynamic range re-
quirements of the electronics necessary to pro-
cess lidar backscatter signatures. Bandwidths
greater than 50 lvi Hz are needed to adequately
monitor returns from solid reflectors, aerosols
and clouds. The amplitude range of the output
signal from a lidar photomultiplier tube may
extend over more than four decades, but wide-
band logarithmic amplifiers and gain switching
techniques can be used to compress the dynamic
range of backscatter signals to be more com-
patible with data recording and display elec-
tronics. In addition to bandwidth and dynamic
range problems, the low pulse repetition rates
of laser systems have historically limited the
application of real-time radar display techniques
to lidar data.
The technique used in early lidar systems was
to photograph the face of a wide bandwidth
oscilloscope where backscatter signal intensity
was displayed as a function of range. Fig. I
presents an example of data collected in this
manner during 1969. These particular A-scope
displays of lidar backscatter signatures were
collected from a downward pointing airborne
lidar which was used to investigate properties
Development of this system was supported by the United States Air Force, Space and Missiles Systems
Organization.
© 1975 Chapman and I-fall Ltd.
121
29
-------
E. E. Uthe, R. J. Allen
w
0
F-?
jO
0
(-9,-
0
-J
HAZE LAYER SURFACE RETURN
1_lull
RANGE — 750 rn/DIV
Figure 1 Example of backscatter signatures recorded on
Polaroid film in 1969. Data collected from an airborne
lidar at an altitude of 3 km and an elevation angle of
— 6O .
of the Sahara dust layer that is transported
across the North Atlantic [ 2]. A video-logarith-
mic amplifier was used to compress the signal
amplitude for recording. Even when multiple
traces were recorded on a single photographic
print, large amounts of film had to be processed
and the meteorologist could not readily interpret
the data in terms of atmospheric structure in
more than a single dimension. However, with
much effort, the photographed backscatter
signatures could be hand-processed into presen-
tations depicting time and space contours of
backscatter intensity, and these displays of
atmospheric structure could be visually inspected
for information on atmospheric dynamic, physi-
cal and radiative processes. A further step was
the digitization of recorded signatures using
curve-following techniques, and the computer
processing of these digital records into contour
maps of backscatter itensity. Fig. 2 presents an
example of a computer generated display show-
inglidar derived structure of the Sahara dust layer.
Such presentations are required for effective
meteorological use of backscatter lidar tech-
niques.
For certain atmospheric conditions or when
appropriate additional information was avail-
able, digitized backscatter data could also be
11111 I III
I I I I II I
I I I I I
I I II
88.5 90.0 91.5 93.0 94.5 96.0 97.5 99.0
DISTANCE — km
cJ
-6.0
-5.0
-4.0
-3.0
-2.0
0. 0
1.0
2.0
3.0
DEVIATIONS FROM BEST FIT
EXPONENTIAL ATMOSPHERE, Relative DB Units
Figure 2 Example of a computer generated cross section of aerosol structure derived from digitized lidar backscat-
ter signatures. Marks above cross section indicate time of laser firing.
122
3.0
2.5
2.0
1.5
1.0
E
-
LU
0
D
I—
F-
-J
0.5
0
30
-------
Digital real-time lidar data recording, processing and display system
E 3.0
1.5
SQ
=
0
0 1.5 3.0 4.5
RANGE — km
Figure 3 Example of aerosol structure derived from video-
disc recording technique. Note the geometry of the smoke
plume and haze layer top.
used quantitatively to derive absolute aerosol
densities. Although lidar data collection rates
were relatively high, the substantial data hand-
ling requirements of both the qualitative pattern
analysis and the quantitative density analysis
typically resulted in the processing of only a
small percentage of any collected data set.
A data processing advance was achieved in
1970 when a video disc recording techniq 1
viewing in graphical form [ 3]. The video disc
allowed a format that enabled recorded data
to be played back through an electronic system
that generated intensity modulated pictorial
displays. Fig. 3 presents an example of data
processed by the video disc technique [ 4]. In
this example, the lidar was being scanned in
I I elevation in a plane nearly perpendicular to a
6.0 7.5 smoke plume which was rising in a surface
haze layer. Each lidar backscatter signature is
represented as an intensity modulated line
segment with brightness proportional to the
logarithm of received signal. In addition to
logarithmic amplification, it was necessary to
add inverse range squared correction electronics
to the photomultiplier detector in order to pro-
duce these intensity modulated displays. The
video disc system greatly increased the lidar
application to real-time and subsequent analysis
of time and space variations of atmospheric
structure over extended volumes. However, the
analog circuitry required to process the data and
the 4-MHz bandwidth limitation of the video,
was applied to the processing of lidar data for
256 A atta, Litt t
512 12; CH NCH POLAR O
MONITOR MONITOR
+ UNIT 0 UNIT 1
AT 11 Opttati , ,g 512 570 8
DISPLAY SYSTEM Sytttta ttd Ltda, oata B ,ts ,Tapt
Ra,ty,,k DX 1007200 ____________________________
512.. 256t4B ,IMt,,ttyy
4Ra 1tt .L,ttePI It
________________ I Tityt 1.06 ________________
591 1C
NT ER F ACE
MODULE
767770 ISIaIHII
767772 IOatpyt Datal
767774 IItp , ,I Datal
BR-4
O 8.8
St ylI — 2.5
to 3.2
I i 792YB 6811 A
ROM MM 11-L EXTENDED
BOOTSTRAP MEMORY ARITHMETIC
LOADER 16K UNIT
773100 Load Addtt,s CyoX T,t,tty
777344 Stall 0.9 tI
UNIBUS
777776 IStaIll TTY KtyboatdlRtadot 767760 ISI l ll
777560 IStat , ,I 767762 00otpt Data)
777562 IDatal 767764 Ilopot Data)
_______________ _______________ TTV PyottalPooh BR-i
CENTRAL TTY 777566 IDIaI REFRESHED DR tIC
PROCESSOR CONTROLLER OR-4 [ A-SCOPE MODULE
SWITCH TAPE POINTER FAST TRANSIENT H CONTROL
REGISTER PUNCH Bio ,oI,oy 8100 LOGIC
777570 (Data) READER KEYBOARD tLa a tF t
ASR 33 TELETYPE
10 that a Itt 1.17
LIDAR SYSTEM
IMat,toot Foa Rat 60 pptttl
Figure 4 Block diagram of the Mark IX lidar digital system.
DUAL DECTAPE TRANSPORT
TU56
t
TAPE 5K Vol. -
CONTROL LER
TC 11
777350 (Da’a ’
777340 ISlaE . . ’
BR-S
UN lB US
1767700
TTY THUMB DECKIT 11.6 iNP ’
SWITCHES INTERFACE
82
LIOAR THUMB
SWITCHES
16j 17 18 1t9
I6) Aaiotth AoqIt , , lot Shot Ntttb , ,tI
I7I Eltoaliot Atgl, , , 10, St,ot Ntob,tl
123
31
-------
E. E. Uthe, R. J. Allen
disc reduced the data resolution and degraded
the quality of the data for quantitative use.
For several years, therefore, many lidar pro-
grams remained mostly concerned with studies of
atmospheric structure rather than absolute
aerosol density.
Digital techniques obviously had much to
offer in handling lidar data and several groups
have, of course, developed systems for this
purpose, particularly in upper atmospheric
studies where the raw signal is normally in the
form of a series of photo-electron pulses.
The rapid development of commercial high-
speed digitizers and minicomputers now make
digital data processing capabilities generally
available at relatively low cost; in addition,
digital display systems are also available.
This paper describes a digital system con-
structed from commercially available com-
ponents that is capable of providing both
real-time computer analysis and real-time display
of processed lidar data. The system offers all
the benefits of digital handling and computation
while retaining the very valuable graphical
capability shown in the familiar two-dimensional
intensity modulated form of Figs. 2 and 3.
Large volumes of data may be recorded, pro-
cessed and displayed in real-time to facilitate
operational and research applications.
2. System hardware
2.1. Constraints
Digital data capabilities were added to the
existing SRI Mark IX analog lidar system which
is installed within a 6 m van complete with its
own power generators. About 2000 VA of
power and room for one additional 190 cm high
cabinet rack were available for the additional
equipment.
The Mark IX was constructed primarily
for air pollution studies where lidar operation
may proceed while the van is moving along
urban and rural traffic links. To be compatible,
the digital system had to be available to with-
stand the vibration associated with mobile
operation. In addition, new applications re-
quired that the new equipment had to withstand
hostile environments characterized by high
humidity and marine air. This required special
considerations for the newly added equipment
plus some rework of the existing analog system.
A further constraint was that the design,
124
construction and programming of the digital
system was to be completed within a limited
funded program that was scheduled to be field
operational within a few months.
2.2. Desired Capabilities
The goal was to incorporate a digital system
capable of real-time analysis utilizing a versatile
interactive gray-scale display system. It was to
record all data produced by the lidar, play back
the data into the computer for later processing
and display, and provide hard copy output sum-
marizing results of quantitative analysis. These
capabilities along with the programming re-
quirements are summarized in Table 1.
2.3. System description
The digital system that was judged to best meet
the desired requirements within the imposed
constraints is diagrammed in Fig. 4. It was
designed around the UNIBUS concept of the
PDP II computer system manufactured by
Digital Equipment Corporation. The bus physic-
ally consists of 56 high-speed (400 ns word
transfer) bidirectional and asynchronous com-
munication lines that allow the computer and
the peripherals to operate at their maximum
speeds. Devices connected to the UNIBUS
can send, receive or exchange data without
processor intervention and without buffering in
memory.
The computer uses 16-bit words but also
effectively processes 8-bit data bytes. Eight
processor registers allow effective processing
of structured word or byte data arrays. A 4-level
interrupt system is available for establishing
peripheral priority control of the UNIBUS.
The bus concept allows a single set of processor
instructions to be used for memory reference,
operational statements, and input and output-
of data from peripherals.
The peripherals consist of a Read Only
Memory (ROM) for rapid loading of bootstrap
instructions, a 16000 word core memory which
can be expanded to 28 000 addressable words,
a teletype unit with paper tape read and punch
capabilities, and an extended arithmetic unit for
hardware integer multiply and divide. The digit-
izer is the Biomation 8100 that can sample
and hold 2000 bytes at sample intervals down to
10 ns. The control logic was designed to allow
the lidar rather than the processor to control
32
-------
Digital real-time lidar data recording, processing and displai’ system
TABLE 1 Desired capabilities of the digital system.
1-lardware
Real-time analysis
Versatile display
Complete recording
Read capability
Hard copy output
Real-time processing and display, keyboard monitor
to control all computer functions of data processing.
A-scope, Z-scope, alphanumerics, time and height
marks, etc.
500 data points/trace, azimuth, elevation, time,
program control switch data and other data at rates
of 60 observations mm— 1 .
Play back of data into computer for later processing
and display.
Hard copy of results, for real-time operational
programs.
Software
Program library
High level languages
Control of branch points
and real-time data input
File manipulation, analysis, display, data acquisition,
and other programs.
Field operations by scientists and engineers familiar
with FORTRAN or BASIC languages.
Control type of analysis and input of heights, levels,
scales, rates, etc., via data switches, i.e., without stopping
program execution.
the data collection operation, thereby main-
taining the timing required by the analog video
disc recording system. In addition to the lidar
backscatter data, an input data interface to the
bus is used to transfer information to the pro-
cessor on date, time, azimuth, and elevation
125
Figure 5 Mark IX digital system. From top to bottom: dual Dectape, TV monitor, PDP-11 minicomputer, Biomation
transient digitizer and teletype with sense switches mounted on lower right.
33
-------
E. E. Uthe, .R. J. Allen
pointing angles of the lidar and manually set
switch data for input of supplementary data
and program control.
The display system consists of a Ramtek
GX200 and a TV screen with vertical raster
scans. The Ramtek chosen uses 256 raster scans,
each with 512 line elements. Each of the 256
by 512 picture elements displays 4-bit informa-
tion to generate gray scale pictures consisting of
16 gray steps. This equates to over haifa million
bits of display memory. Examples of data
displays will be presented below.
Magnetic tape was chosen for the recording
medium because magnetic discs were judged
to be marginal for use during mobile operation
of the van-mounted lidar. However, magnetic
tape recorders are also vulnerable to humidity,
salt spray and vibration. A dual DECTAPE
system was selected since it has redundant read!
write heads with a medium packing density.
The 10 cm diameter DECTAPE reels are pre-
formated into 578 blocks for quasi-random
access. One DECTAPE unit is used for data
storage with typically 490 8-bit samples per back-
scatter signature and 22 bytes of supplementary
data on each block of tape. The other DECTAPE
unit is used as a system device for Digital Equip-
ment Corporation’s RT-l I real-time operating
system.
Fig. 5 shows the digital system installed within
the Mark IX lidar van. The dual DECTAPE
unit is on top and has been fitted with a protect-
ive plastic cover containing three cylinders each
with a desiccant compound and indicator.
These were added to reduce the humidity around
the heads and magnetic tape. Below the DEC-
TAPE unit is the TV monitor with a 90° rotated
yoke for 256 vertical raster lines each with 512
elements of resolution. The PDP-l 1/10 mini-
computer is next with the Biomation 8100 shown
below the computer.
3. System software
The RT-1 I real-time operating software avail-
able from the computer manufacturer includes
a keyboard monitor, peripheral interchange
package, and various operational programs such
as an assembler, a BASIC language interpreter,
and a FORTRAN compiler.
The basic data handling programs are written
in assembly language and are linked with the
higher level language packages. These routines
126
L FIRING 1
I LOGaRITHMIC
0101501 L! !FCAC 1oN ANAlOG
PRIORIrv a
A
DIGITIZATION h ACCORD ON 010CR-DISC
I1 - -11
3 MAGNETIC TAPE I COMPOSITE COMPOSITE
PRIORITy 3 I A-SCOPE 2-SCOPE I
L RANGE CORRECT DATA A SCOPE
QUANTITATIVE ANALYSIS AND J . . . . . . . . . . . .JA OR Z.SCREEN
3 OUTPUT RESULTS TO TCLETVPE [ 1 ]
I- - - -J
Figure 6 Flow diagram of lidar signature processing and
display.
are callable from FORTRAN or BASIC data
acquisition and analysis programs. The user
and service programs reside on DECTAPE
to allow rapid building of a program library
of nearly unlimited size.
Fig. 6 illustrates the data flow for each lidar
firing. After logarithmic amplification, the signal
is routed to both the analog and digital pro-
cessing units. The digitizer control functions
may be set or adjusted either by the operator
or the computer program that is under execu-
tion. Between lidar firings, 2000 data bytes are
retained by the internal digitizer memory that
continually refreshes an oscilloscope display
of signal amplitude as a function of range. This
display provides the operator a real-time view
of the latest backscatter trace and also an indica-
tion of digitizer performance. A specified num-
ber of data bytes (typically 490) are transferred
to core memory and are immediately written
on magnetic tape along with 22 bytes of house-
keeping data as Priority A. Upon completion of
Priority A operations, the digital system per-
forms a set of operations designated as Priority
B. These operations are interrupted at any time
the lidar is fired in order to perform the Priority
A operations for the data resulting from the
lidar firing. Upon completion of Priority A
operations, the control of the digital system
reverts back to the processing of Priority B
operations at the point of interruption. This pro-
cedure allows real-time recording of all data and
further processing of a percentage of the data
determined by the length of the Priority B opera-
tions and the lidar fire rate. At a later time, the
recorded data can be played back with the
Priority B operations performed on all collected
data. Priority B operations are normally directed
34
-------
Digital real-time lidar data recording, processing and display system
to the display of lidar data and inference of
aerosol or cloud densities.
4. Examples of system applications
The capability of the digital lidar system can
be illustrated by the following examples of how
it is used to monitor cirrus cloud structure
and densities in real-time. The Priority B
operations include:
(1) Correction of the lidar trace for the inverse
range-squared dependence and instrument
response functions;
(2) computation of cloud densities over an
altitude interval determined by input
switch data. The computation assumes a
clear air density at the lower input
altitude, a mean crystal radius, and a
factor between the backscatter and extinc-
tion coefficients that includes a multiple
scattering correction [ 5]. An approximate
correction for cloud attenuation is made
on the basis of observed cloud thickness
and relative density as determined by the
maximum cloud-to-clear air ratio;
(3) output of the computational results to
the teletype;
(4) plot of the complete lidar signature on the
digital screen.
These operations could be performed on the
data resulting from each lidar firing with a
lidar fire rate of 6 min 1 and a density computa-
E
w
(9
z
15.0
12.0
9.0
6.0
3.0
1350
w
C ,)
z
0
0
cc
0
-J
Figure 7 Example of a digital A-screen plot showing two
cirrus cloud layers.
tion on over 100 data points i.e., for a cloud 3 km
thick.
The choice of Z-screen or A-screen and the
plot scaling factors (dB/gray step or dB/line
element) are specified by the sense switch settings
which are sampled for each lidar firing. The plot
data are referenced to the minimum digitizer
value (—128 counts) at a range of 9 km. In
addition to the display of lidar data, a sense
switch option allows the painting of a 16-gray
scale step function on the display screen. This is
useful for real-time evaluation of cloud densities
and for adjusting the screen brightness and con-
trast controls for viewing and photographing
the complete 16-gray step levels. Examples of
digital output are presented below.
Fig. 7 shows an A-screen presentation plotted
across the screen using every other point of the
490 data point array. This particular data
1360
Figure 8 Example of a digital Z-screen plot showing structure of a 6-8 km cirrus layer. Curved lines and data
indicate that the lidar was being scanned in elevation during the data collection.
127
ALTITUDE
TIME — h
SCROLL MARK
35
-------
E. E. Uthe, R. J. Allen
TABLE II Examples of I
idar digital data output.
Liquid
Column
H
Mm
S
Height
Density
Number
water
content
14
0
11
6.735
28.6438
48989.1
1.47748
2.92177
14
0
14
6.735
26.8712
32571.7
.982341
2.09537
14
0
17
6.735
27.7394
39779.7
1.19973
2.30663
14
0
20
6.735
27.7756
40112.4
1.20976
2.47013
14
0
23
6.735
28.7885
50648.9
1.52754
3.26136
14
0
26
6.765
28.3339
45476.3
1.37154
3.01024
14
0
29
6.735
27.4862
37526.5
1.13178
2.50087
14
0
32
6.735
26.7627
31767.8
.958098
2.05492
14
0
35
6.735
27.5224
37840.4
1.14124
2.41707
14
0
38
6.765
28.0445
42544.6
1.28312
2.34013
14
0
41
6.735
29.6929
62374.8
1.88118
3.79786
14
0
44
6.765
27.5019
37547.5
1.13241
2.21339
E 15
12
I —
-J
<3
TIME
— hours
1430 1440 1450 1500
1510
I
1
3 dBGR\Y SCALE
Figure 9 Example of Z-screen plot of 16-gray scale steps
and of a cirrus layer that is non-uniform in structure and
density.
illustrates the range corrected backscatter signal
from multiple cirrus layers. The detector has
been gated off for the first 1.5 km to prevent
electronic saturation by low-altitude boundary
layer clouds. Thus far we have not made
extensive use of A-screen plots and have not
yet addec vertical and horizontal scale markings.
Fig. 8 shows an example of a gray scale Z-
screen plot. Initially the lidar was being scanned
in elevation angle, and the curved lines represent
the height limits of integration as input by the
operator, in this case, from 5.5 to 8.5 km. Time
marks are determined from the digital clock
and in this case are spaced at 10 mm intervals.
The black raster line indicates the end of a 256
line plot and the screen has begun to scroll left
erasing the first raster line and adding the last
128
line for each lidar firing so that the last 256 lidar
signatures are always displayed.
Table II presents an example of the teletype
output. The first columns list the times of lidar
firings. In this case the lidar was being fired
every 3 s. In real-time, the cloud density analysis
was completed for every second lidar firing. The
next columns give the height of maximum cloud
density in km, the cloud-to-clear air density ratio
in dB, the inferred crystal concentration in crystals
per m 3 and ice water content in g m ’ at the
height of maximum cloud density, and the last
column gives a quantity related to the vertically
integrated liquid water content.
Fig. 9 presents a Z-screen with the 16 bright-
TIME — h
1230 1240 1250 1300 1310 1320
I l
E 18
115
w
o 12
I-
-J
6
Figure 10 Z-screen plot of a high-altitude low density
cirrus cloud.
1.5 dB/G RAY STEP
36
-------
Digital real-time lidar data recording, processing and display si’stem
E
w
a
12 —
9
6
3
Figure 11 Example of digital iso-contouring applied to the
presence of a high- and low-density cirrus cloud.
E
a
I-
I —
-j
1500
1200
900
600
300
0
1310 1400
TIME
Figure 12 Example of a digital Z-screen presentation of the
low altitude pollution layer over Menlo Park, California, 22
October 1974.
ne s levels plotted on the left-hand side of the
screen. In this case, the operator used 3 dB per
gray step. The lower area represents a range
interval for which the detector was gated off.
The difficulty of photographically reproducing
a TV screen presentation is shown by this figure.
The curved lines and reduction of gray steps were
introduced by the photography. A correspond-
ing loss of cloud structure as compared with that
observed on the actual screen occurs in all the
presentations here.
Fig. 10 illustrates the Z-screen presentation
for a low-density (2 x IO g m 3 ) high-
altitude cloud.
Fig. 11 illustrates the iso-contouring capability
of the Mark IX digital system. When the maxi-
mum brightness scale is exceeded, the data is
recirculated to the low brightness gray steps.
In this manner, a low dB/gray step factor may
be used to give good visual display of low-density
clouds while still getting a dB reading on high-
density clouds.
Fig. 12 presents a Z-screen example of the
structure of the lower atmospheric pollution
layer observed over Menlo Park, California.
The new digital capabilities of the Mark IX
lidar should greatly increase the possibility of
relating screen brightness to aerosol density for
lidar studies of the pollution layer [ 6]. It is
planned to investigate applications of the digital
system in connection with other remote sensing
instrumentation, such as backscatter acoustic
sounder and radio-acoustic sounding systems.
5. Conclusion
- In conclusion, a relatively low-cost lidar data
recording, processing and display system, pos-
sessing many capabilities required for efficient
1700 meteorological operation of lidar, has been
constructed and demonstrated.
References
1. R. T. H. COLLIS and E. E. UTHE, Opto-elcetronics
4 (1972) 87-99.
2. E. E. UTHE and w. n. JOHNSON, Lidar observations
of the lower tropospheric aerosol structure during
BOMEX, Final Report, SRI Project 7929, AEC
contract AT(04-3)-115 (1971).
3. R. J. ALLEN and w. E. EVANS, The Review of Scientific
Instru,nents 43 (1972) 1422-32.
4. W. B. JOHNSON, ‘Lidar Measurements of plume
diffusion and aerosol structure,’ paper presented at
the American Meteorological Society Conference on
Air Pollution Meteorology, Raleigh, North Carolina,
USA (April 1971).
5. C. M. R. FLAT T, Journal of Atmospheric Sciences
30 (1973) 1191-204.
6. E. E. UTHE and p. B. RUSSELL, Bulletin of the A,neri-
can Meteorological Society 55 (1974) 115-21.
3.0 dBI 1.5 dBI
GRAY STEP GRAY STEP
1500
(POT) —
1600
hours
129
-------
REFERENCES
Allen, R.J., and W.E. Evans, 1972: Laser radar (LIDAR) for mapping
aerosol structure. Rev, of Sci. Inst. , 43, 1422-1432.
Cook, C.S., G.W. Bethke, and W.D. Conner, 1972: Remote measurement of
smoke plume transmittance using lidar. Applied Optics , 11, 1742-1748.
Evans, W.E., 1967: Development of lidar stack effluent opacity measuring
system. Phase I and Ia Reports--Design Studies, Edison Electric
Institute, SRI Project 6529. Stanford Research Institute, Menlo
Park, California.
Fernald, F.G., and R.T.H. Collis, 1965: Report on experiments to explore
the feasibility of measuring the opacity of stack effluents by
lidar. Report submitted to Edison Electric Institute, SRI Project
5564. Stanford Research Institute, Menlo Park, California.
Johnson, W.B., 1969: Lidar observations of the diffusion and rise of
stack plumes. J. Applied Meteorology , 8, 443-449.
Johnson, W.B., R.J. Allen, and W.E. Evans, 1973: Lidar studies of stack
plumes in rural and urban environments. Final Report Contract
CPA 70-49, Environmental Protection Agency, SRI Project 8509.
Stanford Research Institute, Menlo Park, California.
Johnson, W.B., and E.E. Uthe, 1971: Lidar study of the Keystone stack
plume. Atmos. Environ. , 5, 703-724.
Uthe, E.E., and R.J. Allen, 1975: A digital real-time lidar data
recording, processing, and display system. J. Optical and Quantum
Electronics , 7, 121-129.
Uthe, E.E., and C.E. Lapple, 1972: Study of laser backscatter by parti-
culates in stack emissions, Final Report Contract CPA7O-173,
Environmental Protection Agency, SRI Project 8730. Stanford
Research Institute, Menlo Park, California.
38
GPO 833—343
------- |