REMOTE SENSING OF HYDROCARBONS AND TOXIC POLLUTANTS:
(WORKSHOP MINUTES)
Freeman F. Hall, Jr. (Editor)
Environmental Research Center
University of Nevada, Las Vegas
4505 Maryland Parkvay
Las Vegas, Nevada 89154
Cooperative Agreement No. CR814002
Project Officer
James L. HcElroy
U.S. Environmental Protection Agency
P.O. Box 93478
Las Vegas, Nevada 89193-3478
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
LAS VEGAS, NEVADA 89193-3478
-------
NOTICE
Although this workshop was sponsored in part and funded in part by the
U.S. Environmental Protection Agency under Cooperative Agreement No. CR8U002
and personnel from the Agency participated in it, this report has not been
subjected to Agency review. The report, therefore, does not necessarily
reflect the views of the Agency and no official endorsement should be
inferred. The report represents only the informed views of the participants.
ii
-------
FOREVORD
In April 1989, a workshop was held in Las Vegas, Nevada, to consider how
emerging remote sensing technology could be used for the detection and mapping
of toxic air pollutants and hydrocarbons in the atmosphere and applied to the
solution of priority environmental problems. The meeting vas sponsored
jointly by the University of Nevada-Las Vegas's Environmental Research Center
and the Environmental Monitoring Systems Laboratory-Las Vegas. The attendees
were scientists and engineers from these groups, other Federal agencies,
universities (domestic and foreign), and private companies (domestic and
foreign) actively involved in research and development of the relevant remote
sensing technologies. Representatives from several Regions also attended the
workshop to act as observers and to provide the perspective of potential
clients for the technology.
This report is the result of that workshop. It summarizes the relevant
technologies and the capabilities of candidate remote sensors; a representa-
tive list of such sensors, their capabilities and characteristics, and their
geographical locations are presented for reference. The report also compares
the technologies for future performance capabilities, ease of operation, and
cost over the next five to ten years. As such, information contained in the
report should aid the Agency in formulating plans for sensor research and
development for dealing with important regulatory and research issues facing
it.
The report was assembled from inputs developed at the workshop and after
it by all of the participants. A draft of the report was circulated for final
comment by the participants. The result is, therefore, a group effort by the
attendees, who are listed in the appendices. The views expressed are the
informed views of the attendees and not necessarily those of the
organizations they represent.
James L. McElroy
Project Officer
iii
-------
Contents
Foreword ill
Preface 1
1. Executive Summary 3
2. Objectives of the Report , 6
3. Remote Sensing Advantages and Characteristics 9
4. Conclusions 14
5. Recommendations 15
Appendices
1 Glossary of Acronyms and Remote Sensing Terms 16
2 Remote Sensors Capabilities List 17
3 Spectroradiometers 29
4 Laser Remote Sensors 39
5 Examples of Remote Sensor Successes 46
6 Future Improvements Possible in Remote Sensor Performance 49
7 Remote Sensing Detection Levels for Hydrocarbons and Toxics .... 57
8 Attendees at the 1989 Las Vegas Workshop 58
References 62
iv
-------
Preface
This is a report on the significant advantages of remote sensors for
detecting and mapping airborne hydrocarbons and toxic pollutants. The report
is essentially the minutes of a workshop on this topic held in Las Vegas,
Nevada, on April 6-7, 1989. All of the participants at the workshop have
contributed to this report. A glossary of acronyms and remote sensing terms
is given in Appendix 1. Sensor performance described in detail in
Appendices 2-5 is achieved vith remote sensors that already exist in 1989.
The sophistication of the remote sensing discipline is not widely appreciated
because of its rapid maturity in the past ten years. This growth has been
possible thanks to:
* Spectacular advances in digital computers and data processors, and
* Improvements in optical components and laser technology, largely the
result of DOD and NASA programs.
Remote sensing as a monitoring tool is available and ready to assist the
U.S. EPA in documenting and solving many of its most pressing air-related
problems. This is a problem-focused report. For example, the many cities
with Clean Air Act nonattainment for ozone levels likely have emissions of
precursor hydrocarbons that differ considerably from (and are generally higher
than) those predicted using engineering models for inclusion in emission
inventories. The California tunnel experiment (Ingalls, 1989; Stockberger
et al., 1989) demonstrated that tailpipe hydrocarbons were two to seven times
higher than expected. Grab samples in the open environment can never provide
the complete picture of the merging plumes from natural and anthropogenic
sources into the ambient background. A properly designed suite of remote
-------
sensors can paint the big picture, providing real-time, spatially resolved
air quality data!
-------
1. Executive Summary
Remote sensing technology is now ready and able to play a significant
role in EPA programs such as TSCA, EMAP, RCRA, and Clean Air Act compliance.
By their very nature, remote sensors unravel many of the difficult pollution
monitoring problems by:
Measuring representative, volume-averaged samples;
Documenting the extent of hydrocarbons or toxic gaseous pollutants;
Providing reduced data to the investigator in the field in real time;
and,
Eliminating the need to capture samples that need to be returned to
the laboratory for analysis.
Remote sensors utilize the absorption, emission, or fluorescence spectra
of pollutants (and each pollutant has its own, unique, identifying spectrum)
to identify unambiguously the chemical constituents in the atmosphere. Thus
the possibility of overlooking an unexpected contaminant is minimized. Many
of the present-day remote sensors can sort out and identify a specified
pollutant from the "urban soup" consisting of multitudes of compounds. Such
off-the-shelf sensors as listed in Appendix 2 have demonstrated detection
sensitivities that already meet EPA needs and monitoring requirements.
There arc tvo types of remote sensors that can already meet EPA
necessities: Spectroradiometers, and Laser Radars or Lidars. Many examples
of both types of sensors are in everyday use, and some have passed the
stringent type-testing requirements of the U.S. Army and Navy. They are not
"ivory tower" instruments or future goals of dreamy-eyed researchers.
3
-------
There are several functions that remote sensors can provide without the
necessity of physically entering private, restricted or hazardous areas:
* Monitoring emissions from point sources such as stacks;
Defining the amount of pollutant that crosses a defined fenceline, for
example, one that surrounds a refinery;
* Measuring ambient concentrations of pollutants over path lengths as
short as a fev meters to as long as many kilometers.
Present-day lidar remote sensors (described more completely in Appendix
A) are able to detect, identify, and range-resolve toxic pollutants in the
parts-per-million (ppm) range. Spectroradiometers (Appendix 3) can detect,
identify and quantify line integral averages of toxic pollutants in the parts-
per-billion (ppb) range. Of the Spectroradiometers, the' Fourier-transform,
infrared devices (FTIR's) have the unique ability to identify quickly those
pollutants present by performing real-time transforms and matching the
detected spectra against stored libraries of pollutant spectra. Hultidetector
ultraviolet (UV), dispersive instruments used with xenon-arc sources also
show much promise although they have not been widely used to date. Laser
sources that may be tuned to desired absorption or "window" lines may also be
used, eliminating the need for spectral discrimination or dispersion in the
radiometer. All of these sensors are available today - no basic research is
required in order to apply their proven capabilities. Costs of the spectrora-
diometers rang* fro« $35R for the more simple, hand-held devices to perhaps
$100K for versatile, mobile units. Lidars run in the $100K to $200K price
range, being somewhat more complex instruments.
-------
It is now possible to deploy in the field, as hand-held, mobile or air-
lorne instruments, either singly or in suites, such lidars and spectroradio-
leters that will complement each other's performance. FTIR output data can
e used to instruct a tunable laser source, in a differential absorption lidar
DIAL), what wavelengths to use to range resolve the pollutant for better
efinition of its spatial extent. These tvo instruments, especially when used
ogether, seem to hold the most promise for field-deployed hydrocarbon and
oxic pollutant remote sensors, now, into the 1990's, and beyond.
Even though present-day remote sensors have demonstrated versatile and
eliable performance, improvements are possible. Over the next ten years,
lie detection sensitivity of spectroradiometers will be improved by a factor
E ten (10 dB) as detailed in Appendix 6. Increases in computer speed,
emory, and software sophistication will also significantly improve FTIR
jrformance. Similar enhancements in lidar performance by 10 dB in sensi-
Lvity and in rapid tuning and pollutant identification accomplishments will
» possible, even with no "breakthrough" developments in components. Remote
jnsors will be able to provide the monitoring sensitivities necessary even as
Lean air requirements become more stringent in the future. Future remote
jnsor costs should not radically increase (in 1989 dollars).
-------
2. Objectives of the Report
Air pollution by ozone-precursor hydrocarbons or other toxic compounds
is a problem on the local, regional, and global scale. Many EPA concerns,
such as TSCA, EHAP, RCRA, Alternate Fuels, and the more general matters of
acid rain and acid aerosols, stratospheric ozone depletion, global climate
change, and the sick building syndrome are affected by hydrocarbons and
toxics. The Clean Air Act requires that pollutant nonattainment areas be
characterized quantitatively and strategies be developed for attainment of the
relevant air quality standards. There is clearly a need for effective, real-
time monitoring on all geographic scales. The objectives of this report are
to demonstrate why and hov remote sensing technology is now ready to help
fulfill this need.
In addition to the day-to-day ambient pollution in urban areas, hydro-
carbon and toxic spills or the escape of such vapors and gases have become so
common as to be hardly newsworthy. Almost every train derailment seems to
produce such an event. Industrial accidents and fires force the evacuation of
neighborhoods, even whole cities because toxic fumes or smoke threaten safety
over vide areas. Are all such evacuations necessary? Are the areas where
exposure will b« dangerous alerted sufficiently in advance or at all? When
health problem* arise in the aftermath of such occurrences, is it possible to
determine if the toxic exposure was to blame? What chemical transformations
occur as such pollutants are transported in plumes or air masses across local
and regional boundaries?
-------
Ideally, public officials would be able to predict for the day-to-day
ambient conditions or during the course of an accident when, where, and at
what level the harmful constituents will exist, and be able later to say
exactly where and at what level some potentially dangerous pollutant did
exist. There is the additional problem of measuring the even lower concen-
tration but more persistent hydrocarbon or toxic emissions from landfills.
Such a service to the public cannot now be provided. In this report we
concentrate on existing remote sensing technology that shows promise for
supporting such a service in the future. Ve focus on remote sensing because
we need measurement techniques that can cover large volumes in space in a
short time, techniques that can determine where a plume was and where it was
not and that do not require entry into areas already impacted by the problem.
As an example, the Superfund mandate requires assessment, characteriza-
tion, and cleanup of hazardous waste sites. An active or passive FTIR can be
used for rapid characterization by measuring absorption spectra along multiple
paths that crisscross the site. Just five minutes of measurement and analysis
time for each path provides the data to produce rapidly a map of the site with
isopleths showing contaminated regions. This procedure would pinpoint areas
where more expensive contract laboratory procedures (CLP) might be needed,
eliminating the large costs and time required for CLP where there was no
contamination.
Hazardous vaste incinerator emissions can be monitored by FTIR's in
under two minutes compared to weeks required for gas chromatograph (GC) or
-------
mass spectrometer (MS) procedures. Quantitative agreements of PTIR's with GC
and MS were recently demonstrated (McLaren et al., 1989).
A laser SAR (source-augmented radiometer) using retroreflectors to
redirect the laser beam back to the instrument can be set up conveniently
along the fenceline of an area to be monitored. Pollutants that advect with
the wind out of the site cross the laser beam. The laser wavelengths are
selected so that one is absorbed and one is not, provided a differential
detection signature. This method has been proven during numerous experiments,
and is now used routinely by major chemical plants as a safety monitor.
Ve hope to demonstrate in the remainder of the report that such remote
sensing technology, discussed briefly above, offers a viable method for the
U.S. EPA to utilize in enhancing the performance of its mission.
-------
3. Re»ote Sensing Advantages and Characteristics
Ve define remote sensing as the measurement of a target constituent by
means of wave energy interaction with the target, using an instrument at some
distance from the target. For hydrocarbons or toxic pollutants the waves will
most likely be electromagnetic, for acoustic waves do not have a strong dis-
crimination ability for most gases or vapors. The electro-magnetic remote
sensor exploits differences in absorption, emission, fluorescence, or scat-
tering between the target constituent and the surrounding atmosphere to
identify the pollutant and to quantify its concentration.
The advantages that remote sensors have over in situ sensors are the
following:
* More representative sampling because of volume averaging.
* One sensor can sample a large volume, over many different lines-of-
sight, and at the speed of light.
One sensor can be used to sample for multiple pollutants by adjusting
its operating frequency or through selective data analysis whereas
most current EPA methods are pollutant specific with each requiring a
different method.
* Measurement results can be available in real time, in the field if
desired.
* Mapping for hot-spots of pollutants is possible quickly.
Stand-off Measurements are possible - no need to enter hazardous or
restricted sites.
Can be on-line 24 hours per day.
There is a proven base of technology, paid for largely by the DOD and
by NASA.
-------
Costs for the complete series of measurements to characterize a site
can be kept low through the expedient of scanning, volume sampling,
and real-time data analysis.
Can show the absence of known materials in the beam.
Can be used as real-time monitors to tell when in situ sensors have a
chance of measuring other than zero concentrations.
Reactive or unstable gas constituents can be measured without sampling
problems inherent in capture techniques.
Problems with surface adsorption and time delays avoided.
Rapid field calibrations allow quantitative measurements.
Details on how many of these various advantages are accomplished are
given in Appendices 2 - 5 in this report.
Before presenting more information on the characteristics of appropriate
remote sensors, it is helpful to define sensor performance. There are many
ways to describe the detection sensitivity of sensors - in parts-per-million
or -per-billion by volume or by mass, in mass of pollutant per mass or volume
of air, or by path integrated concentrations, particularly for those sensors
which intrinsically average over their light path. For path-integrated con-
centrations, one part-per-million in a 1-m path produces the same absorption
as one part-per-billion in a 1-km path, or 1 ppm-m « 1 ppb-km. Note that 1
ppm for an ideal gas is 1 ppmv(by volume or by molecules). This mixing ratio
is independent of pressure, temperature, and molecular weight.
At a temperature of 273°K and a pressure of one atmosphere, a gram-
molecule of gas or mol occupies 22.4 1 and contains 6.02 x 1023 molecules
(Avogadro's number), so the number of molecules, no, in 1 ml is
10
-------
no = 6.02 x 1023 / 22,400 - 2.69 x 10" .
At 298°K this number is 2.5 x 1019 molecules per ml, so 1 ppm « 2.5 x 1013
molecules/ml path integrated, so (since 1 ml = 1 cm3)
1 ppm-ra = 2.5 x 1013 x 100 cm = 2.5 x 1015 molecules/cm2.
Taking for example benzene, molecular weight (H.V.) of 78,
2.5 x 1015 x 78
1 ppra-ra » 1 ppb-km » » 3.3 x 10~7 gm cm"2
6.0 x 1023
or for any temperature T(K) and pressure P(Atm), the conversion is
M.tf. x ID'4 273
1 ppm-m = x x P(Atm) gm cm"2.
22,400 T(K)
It is hoped that these conversion relationships, provided by Donald Stedman,
will be found useful when comparing sensing instruments whose performance may
be described using different measures of sensitivity.
Remote sensors for airborne hydrocarbons or toxics fall naturally into
two classes - spcctroradiometers or lidars. A short description is provided
here of each class of sensor with advantages of each. More detailed infor-
mation is given in Appendices 2-7.
11
-------
A spectroradiometer is a radiation-measuring device (a radiometer) that
is able to identify a pollutant by discriminating the spectrum of the
radiation that it absorbs, scatters, or emits. The radiation that is absorbed
may come from the naturally occurring background of sunlight or thermal
emission, in which case the instrument is a "passive sensor," or it may come
from a supplied source of radiation such as a lamp or glover. The spectral
discrimination may be provided by an absorption or interference filter, it may
be supplied by a dispersing element (a prism or a diffraction grating), or a
scanning interferometer may be used.
Spectroradiometers may be packaged into small, rugged units - even hand-
held devices are possible. The technology is veil developed and proven in
the field. Costs can be in the $35K to $100K range. Operation can be quite
simple. They can monitor fencelines, or can be used to scan an area, provid-
ing an integrated measure of the target pollutant(s) along the line-of-sight.
Appendix 3 provides more details on spectroradiometers.
Lidars use laser sources to irradiate an air volume; the backseattered
or retroreflected radiation is then analyzed to detect and identify pollutant.
A Differential Absorption Lidar (DIAL) utilizes at least tvo vavelengths from
the laser, at least one of which is absorbed by the pollutant, and one of
which falls in a "window" in the pollutant absorption spectrum. Frequency
modulated (FM) lidars modulate the laser so that the modulation frequency
appears as tvo sidebands on the laser line frequency. If the laser is
operated near the edge of a pollutant absorption line, the sidebands will be
differentially affected. Raman lidars utilize the Raman shift in frequency of
12
-------
the backseattered radiation that is a characteristic of the pollutant being
measured.
Lidars tend to be more complex instruments than spectroradiometers.
However, they can provide range-resolved data on the concentration of pollu-
tants without requiring retroreflectors. They can therefore be used to
document plume dispersion models when the lidar is advantageously placed to be
able to scan large volumes of the plume. Airborne lidars are ideal
instruments for investigating long-range transport of aerosols in plumes.
Range resolution as fine as several meters is available with many instruments.
Mobile lidars in small vans have been used for years in air pollution studies,
so there is a substantial foundation of experience to draw upon. Lidar costs
are in the $100K and up range, depending on the versatility designed into the
instrument, but one lidar may be able to replace many in situ instruments.
The cost of the entire measurement needs to be considered, and not simply the
single instrument cost. More details on lidars are given in Appendix 4.
13
-------
4. Conclusions
A number of remote sensing techniques are now available for monitoring
hydrocarbons and toxic airborne pollutants. Filter and dispersive spectro-
radiometers with restricted versatility are sold commercially. The more
versatile FTIR's and DIAL'S are now becoming commercially available. Auto-
matic, unattended field operation has been successfully demonstrated for
several types of such adaptable remote sensing systems. FTIR's and DIAL'S
complement each other in identifying a pollutant's presence and its spatial
extent; there is potential for improving their performance by a factor of ten
or more in the next ten years.
Much of the promise for FTIR's lies in their reliance on computer soft-
ware and hardware to perform the frequency analyses and to access the stored
libraries of reference spectra. In both of these areas we can expect signi-
ficant advances and cost cuts in the next ten years, judging from the past
ten. Gains in optical hardware and radiation sources will also enhance FTIR
performance. No "breakthrough" component development is required to enhance
their present quite adequate capabilities.
DIAL performance can also be enhanced over the next ten years by about
one order of Mfnitude vith no breakthroughs. If avalanche photodiodes for
the thermal infrared can be made feasible for field work, there is a potential
for an additional order of magnitude or more of performance improvement.
Rapid line-tuning will be easier as the development of optical modulators for
the infrared progresses.
14
-------
5. Recommendations
We recommend that EPA solicit proposals for a systems analysis of the
time and costs of monitoring several problem sites or cities using existing:
in situ instruments; a suite of remote sensors, and a combination of remote
and in situ sensors. In this way we will obtain an objective assessment of
the total time and costs to determine if remote sensing can effectively
supplement environmental monitoring techniques now in use.
There are obvious improvements and modifications possible in the remote
sensors described in more detail in Appendix 2. The EPA should solicit
proposals for improvements in spectroradiometers and in DIALs to obtain
instruments that are as perfectly matched to the Agency's requirements as
technology permits. Side-by-side sensor comparison tests should continue to
be run at selected industrial and toxic waste disposal sites to document the
utility and capabilities of existing instruments, and of nevly developed
spectroradiometers and laser sensors.
Vhen the Agency has seen demonstrated the utility of remote sensors, and
it decides that remote sensors are indeed useful in accomplishing its mission,
it should consider procuring a mobile spectroradiometer/DIAL system. The
system vould b« used for continuing screening experiments, monitoring, and
emergency response. As remote sensing becomes an accepted, reference tech-
nique, the use of remote sensors should be added to accepted compliance
measurement methods.
15
-------
Appendix 1
Glossary of Acronyms and Remote Sensing Terms
CLP - contract laboratory procedures
DIAL - Differential Absorption Lidar
DOD - U.S. Department of Defense
DOE - U.S. Department of Energy
EPA - The United States Environmental Protection Agency
EMAP - Ecological Monitoring Assessment Program
FLIR - forward-looking, infrared
FTIR - Fourier transform, infrared (spectroradiometer)
GC - gas chromatograph
IR - infrared; wavelengths from 0.75 to 15 urn in this report
laser - light amplification by stimulation of emission
lidar - light detection and ranging, usually with a laser source
MS - massspectrograph
NASA - U.S. National Aeronautics and Space Administration
NOAA - U.S. National Oceanic and Atmospheric Administration
NRC - National Research Council
ppm - parts per million by volume
ppb - parts per billion by volume
RCRA - Resource Conservation and Recovery Act
SAR - Source-Augmented Radiometer
S/N - signal*to-noise ratio
TEA - transverse-excited, atmospheric (pressure); a C02 laser
TDL - tunable diode laser
TSCA - Toxic Substance Control Act
UV - ultraviolet (wavelengths 200-400 nm or 0.2-0.4 urn here)
16
-------
Appendix 2
Reacte Sensors Capabilities List
This appendix provides in a concise form a representative list of air
pollution remote sensors that exist today, their capabilities and charac-
teristics, and vhere they are located. Readers of the report are encouraged
to contact any of the people listed for more information. These data sheets
were developed by attendees at the workshop; the list is not necessarily all-
inclusive.
Each of the data sheets was provided by one of the workshop attendees.
They have not necessarily been checked for accuracy, but each one reports on a
different instrument with unique capabilities. Host of these instruments have
already been utilized in one or more field campaigns. Information is
generally available not only on the capabilities of the instrument but also
on the experience of the operating crev.
The instruments are listed in order of complexity. First come the
passive radioMters; next are the SAR's; and last, the laser DIAL'S. Note
that instrument spectral resolution is often given in "wavenumbers" or cm'1,
sometimes also termed "inverse centimeters."
17
-------
Type of Instrument:
Applications:
Contact person:
Organization:
Address:
Telephone:
Pax:
Measuring principle:
Gases to be Measured:
Resolution/tile
resolution:
Spectral range:
:
Operating crev:
Special logistics:
Potential improvements:
Remote Sensing Passive PTIR spectrometer
Site characterization, remote target component
monitoring
Jack Demirgian
Argonne National Laboratory
9700 South Cass Avenue
Argonne, IL 60439
(312) 972-6807
(312) 972-5287
Infrared emission of gases
Volatile organics
2 wavenumbers, 35 scans/sec
3500 - 2500 cm'1 (3 - 4 um);& -
1250 - 800 cm'1 (8 - 12.5 urn)
One chemist
Need temperature differential of 0.01°C
between target and background.
Enhance qualitative identification and
add quantitative capabilities.
18
-------
Type of InstruHent:
Applications:
Contact person:
Organization:
Address:
Telephone:
Measuring principle:
Gases to be Measured:
Resolution:
Weight:
Dimensions:
Pover required:
Sensitivity:
Special logistics:
Nuaber of units:
Remote Sensing Passive FTIR spectrometer
(non-military version)
Fenceline monitoring, detection of hazardous
atmospheric materials
Robert Kroutil
U.S. Army Chemical Research, Development and
Engineering Center
Commander, US Army CRDEC
Attn: SMCCR-RSL/ Robert Kroutil
Aberdeen Proving Ground, MD 21010-5423
(301) 671-3021
Passive infrared emission of gases
All of which absorb in the 8-12 urn region
2 wave numbers
< 30 Ib
0.6 ft3
< 10 V for interferometer @ 24 V, additional power
required for portable PC
1-30 ppm-m in the passive mode
Liquid N2 needed for detector
2 exist (cost .$35 K each). The optical head is
commercially available from MIDAC Corp., Costa Mesa,
CA. Data processing module and software available
through CRDEC.
19
-------
Type of Instrument:
Applications:
Contact person:
Organization:
Address:
Telephone:
Measuring principle:
Gases to be Measured:
Resolution:
Scan time:
Weight:
Dimensions:
Pover required:
Sensitivity:
Special logistics:
Number of units:
Additional notes:
Remote Sensing Passive FTIR spectrometer
(military version)
Fence line monitoring, detection of hazardous
atmospheric materials
Robert Kroutil
U.S. Army Chemical Research, Development and
Engineering Center
Commander, US Army CRDEC
Attn: SMCCR-RSL/ Robert Kroutil
Aberdeen Proving Ground, MD 21010-5423
(301) 671-3021
Passive infrared emission of gases
All of which absorb in the 8-12 urn region
4 vavenumbers
5 scans per sec
< 50 Ib
1.4 ft3
< 80 V
military design goal < 150 mg/m2
None; liquid N2 for detector supplied by a Magnavox
rotary cooler.
. 30 units exist
This military version is called the XM-21 passive
interferometer; it is designed for the automatic
detection and alarm for chemical vapors; ZnSe
optics; flex-pivot interferometer.
20
-------
Type of Instrument:
Applications:
Contact person:
Organization:
Address:
Telephone:
Measuring principle:
Range:
Remote Sensing FTIR spectrometer
Site characterization, remote target component
monitoring
Bill Herget
Nicolet Analytical Instruments
5225 Verona Road
Madison, WI 53711
(608) 271-3333 Fax: (608) 273-5046
Active - longpath IR absorption
Passive - IR emission
Active - 0.7 km with 14.5" telescopes
2 km with 24" telescopes
Passive - depends on source emission
Gases to be Measured: Volatile organics
Spectral resolution
and scan ti»e:
Spectral range:
Detector:
Number of instruments:
Weight:
Dimensions:
Pover:
Operating crev:
Calibration:
Accuracy:
Safety considerations:
Potential improvements:
2.0 to 0.3 cm"1 (vavenumbers),..
"60 sec
600 to 6000 cm'1 (1.7 - 1 urn)
HgCdTe, LN2 cooled (others available)
5 active; 10 - 15 passive
150 to 200 Ib (light-weight prototype being tested)
34"x26"xl5" for main optical bench
20 amps @ 110 volts
Tvo people needed to align active system, one person
for operation; one person for passive system
Use standard gas mixtures with internal cell or use
existing reference spectra
2 - 10X
No safety problems
Total automation of data analysis procedure
(preliminary software for this being tested)
21
-------
Type of Instrument:
Suggested applications:
Contact person:
Organization:
Address:
Telephone:
Fax:
Measuring principle:
Gases to be measured:
Resolution/time
resolution:
Spectral range:
Operating crev:
Special logistics:
Potential improvements:
Calibration method/
accuracy:
FTIR - Active sensing SAR
Characterization of hazardous waste site, detect
target contaminants in solids such as soils and
sludge; continuous emission monitoring.
Jack Demirgian
Argonne National Laboratory
9700 South Cass Avenue
Argonne, IL 60439
(312) 972-6807
(312) 972-5287
Source-augmented infrared spectro adiometer
Volatiles and semivolatile organics
0.5-2 vavenumbers, 10 scans/sec @ 2 vavenumbers
4000 - 700 cm'1, 2.5-14.3 urn
One technician
Liquid N2 for the HgCdTe detector
Identification of non-target components
Calibrated against a standard gas /
accuracy ±2X; precision ±20X.
22
-------
Type of Instrument:
Applications:
Contact person:
Organization:
Address:
Telephone:
Fax:
Measuring principle:
Gases to be measured:
Resolution:
Time per spectrum:
Weight:
Size:
Pover:
Features:
Field experience:
Special logistics:
Potential improvements:
Calibration method:
Automatic scanning FTIR spectrometer, with and
without multiple remote sources
Plant fenceline and operating area monitoring,
emergency gas release monitoring and tracking, toxic
storage and handling
R.L. Sandridge, R.N. Hunt
Mobay Corporation, Research Department
New Martinsville, WV 26155
(304) 455-4400 Ext. 2207
(304) 455-4400 Ask for. FAX terminal 2438
FTIR absorption spectra in the 8-14 urn region;
computerized turret aims at remote IR sources
(quantitative) or ambient background(qualitative).
Emission spectra of hot plumes may also be obtained.
All vhich absorb in 8-14 um region
< 1 vavenumber
2-4 sec
110 Ib plus remote computer terminal
48" high, 12"-14" diameter
110 V, 60 Hz, 500 V (220 V, 50 Hz also)
Automatic operation, spectrum averaging, background
correction, simultaneous video display of area being
monitored, computer assisted calibration,scale
expansion, alarm and printout functions, rugged
construction.
Extensive field tests in industrial setting;
permanent installations in U.S.A. and Vest Germany.
Electrical power only; can run on portable field
generator.
3-5 um operation available in near future
Flow-through gas cell using standard gases; computer
assisted against multiple remote sources.
23
-------
Type of Instrument:
Applications:
Contact person:
Organization:
Address:
Telephone:
Measuring principle:
Gases to be measured:
Resolution:
Weight:
Pover:
Sensitivity:
Field experience:
Special logistics:
Potential improvements:
Calibration method:
Number of instruments:
Safety:
SAR, Long-path UV absorption spectrometer, with no
moving parts, 75-tf xenon arc lamp retroreflector,
prism, and diode array
Plant fenceline and operating area monitoring, flux
determination,collaborate with long-path IR SAR
Donald H. Stedman
Dept. of Chemistry, University of Denver
University Park
Denver, CO 80208-0179
(303) 871-3530
Single beam gas-phase absorption
Aromatic hydrocarbons and substituted aromatics,
chlorine and aldehydes, N02, S02, and other gases
with UV absorption
±2 to ±20 Angstroms, selectable
90 Ib plus personal computer
110 V, 10 A
Noise is ~103 absorbance units (base 10) which is <1
ppb km of phenol, "10 ppb km of benzene and toluene
Extensive field tests at seven hazardous waste
disposal sites in NY and NJ
Dry air or N2
Add long-path IR to same beam; lover power & weight,
better software & library
With added standard gas cells; once calibrated, the
system can not lose calibration since all measure-
ments are I/I0 ratios.
One exists; ** 2 months required to build
Eye safe at 6 ft from xenon arc; high voltage
starter for arc lamp needs careful shielding.
24
-------
Type of Instrument:
Applications:
Contact person:
Organization:
Address:
Telephone:
Measuring principle:
Gases to be Measured:
Resolution:
Weight:
Pover:
Sensitivity:
Distance:
Response time:
Field experience:
Calibration Method:
Number of instruments:
Safety:
Laser-SAR, IR absorption radiometer, using terrain or
retroreflector target
Plant fenceline and operating area monitoring, flux
determination
Orman Simpson
TECAN Remote (ELS)
3000 Northvoods Pkwy., #185
Atlanta, GA 30071
(404) 242-0977
Single beam gas-phase absorption, beam may be scanned
through 16° x 16°
All those that absorb at C02 laser lines
Width of C02 laser line
Approximately 300 Ib
110 V, 20 A
1 to 400 ppm-m, sensitivity is gas dependent
100 m to 5 km, depending on laser mode (CW or
pulsed), and target used
Less than 1 sec in most cases
Extensive field tests at seven hazardous waste
disposal sites in NT and NJ, plus permanent
installations at industrial sites
With added standard gas cells
Several have been installed, mobile van available for
rapid response measurements
Bye safe at telescope aperture
25
-------
Type of Instrument:
Suggested applications:
Contact person:
Organization:
Address:
Phone:
Fax:
Measuring principle:
Gases to be measured:
Resolution/Time
Resolution:
Spectral range:
Veight:
Dimensions:
Pover:
Operating Crev:
Special Logistics:
Potential improvements:
Safety considerationst
Frequency agile C02 TEA lidar
Plant monitoring
Volker Klein
Battelle Institute
Am Roemerhof 35
D-6000 Frankfurt 90
Federal Republic of Germany
49-69-7908-2859
49-69-7908-80
C02 DIAL measurement
All which absorb in the 9-11 um range
C02 lines/~l second
9R40 line to 10P40 line, 9-10 um region
160 Ib
40nx40nxl5"
1.5 kW
1 engineer
Liquid N} for the HgCdTe detector
(filling can be automated)
Fully automated operation
Eye-safe laser is used
26
-------
Type of Instruaent:
Suggested applications:
Contact person:
Organization:
Address:
Phone:
Fax:
Measuring principle:
Gases to be measured:
Spatial resolution:
Spectral resolution:
Measurement time:
Dimensions:
Pover:
Operating Crev:
Special Logistics:
Safety considerationsi
Status:
Coherent-detection, dual TEA C02 lidar, MAPM (for
Mobile Atmospheric Pollutant Happing)
Emission monitoring from factories
Steven Alejandro
Air Force Geophysical Laboratory
AFGL/OPA
Hanscom AFB, MA 01731
(617) 377-4774 or -3695
(617) 377-4498
C02 DIAL measurement
All which absorb in the 9.2-10.7 urn range including
ethylene, raethanol, ozone, organic solvent vapors,
SFg, vinyl chloride, hydrazines
50 m to 1 km depend ng on range, r, which can be 200
m < r < 3 km
C02 lines
10 s to 1 rain, depending on range; set up time for
instrument is " 1 hr
Housed in a 35x8x13% ft semi-trailer which requires
an air-ride tractor
10 to 20kV
1 scientist, 1 technician
Liquid N. for the HgCdTe detector, He, N2, and C0}
gas bottles for laser
Eye-safe laser is used
MAPM is a research instrument developed by the Jet
Propulsion Laboratory for the U.S. Air Force. It was
assembled using essentially off-the-shelf components
to demonstrate the feasibility if such an approach.
Ref.: (V.B. Grant,1989)
27
-------
Type of Instrument:
Suggested applications:
Contact person:
Organization:
Address:
Phone:
Fax:
Measuring principle:
Gases to be measured:
Resolutions:
Spectral range:
Veight:
Diaensions:
Pover:
Operating Crev:
Special Logistics:
Number of instruments:
Safety considerations:
CO 2 TEA lidar w/heterodyne receiver
Plume monitoring, wind measurements
R. Michael Hardesty
NOAA Wave Propagation Laboratory
325 Broadway
Boulder, CO 80303
(303) 497-6568
(303) 497-6750
Doppler & CO DIAL measurement using atmospheric
aerosols as backseatter target
All which absorb in the 9-11 ym range
Range - 300m increments to 20 km range
horizontal, ~5km vertical
Wind speed - 0.5 m/s
Time -"Is
9.25 - 10.6 um
Housed in 20,000 lb semi-trailer
32x12x8 ft
-5 kW
2 experienced personnel
Liquid N2 for the HgCdTe detector
One in operation
Eye-safe laser is used
28
-------
Appendix 3
Spectroradioaeters
Introduction
Traditionally, remote sensors fall into tvo classes, either active or
passive. An active sensor employs a transmitter to emit radiation that
interacts with the target plume; then that radiation is collected by a
receiver and analyzed. A passive sensor uses naturally-occurring radiation,
such as incident sunlight or thermal emission from the target, terrain, or the
atmosphere. We see immediately that a passive sensor has the potential for
being a simpler device than an active sensor since it does not need its own
transmitter. In general, a passive sensor also will be lighter in weight and
less expensive. Even hand-held instruments are quite feasible. Because of
their long history of use, accurate calibration techniques are readily
available for radiometers, such as absorption cells and blackbody sources of
radiation.
However, a passive sensor is not as able to provide ranging information
on the pollutant plume unless two or more are used for triangulation, and may
suffer from other sensitivity problems (for example, passive thermal sensors
can experience thermal washout, when the background and the target are all
29
-------
at the same temperature). We will examine both active and passive spectro-
radiometers in this appendix. Since radiometers operate over a line-of-sight
determined by the optical axis of the instrument, they are ideal for monitor-
ing a fenceline.
Passive Spectroradiometers
A device that measures radiation is called a radiometer. To identify a
toxic pollutant we must be able to discriminate the spectrum of the radiation
that it absorbs, scatters, or emits by using a spectroradiometer. There are
several types of Spectroradiometers that ve need to consider. These are
classified by the technique used to produce or identify the spectrum: filter,
dispersive, gas-absorption, and interferometer instruments. Some sources of
hydrocarbons and toxics such as gas flares or the warm plumes from sevage
plants are by their very nature ideal targets for passive Spectroradiometers.
There is a wealth of experience with such devices; for example, thirty-seven
different designs have been developed and tested at the U.S. Army's Edgewood
Arsenal/Aberdeen Proving Ground since 1945.
Filter Radiometers
The arMd forces have been forced to face the problem of toxic plume
detection even longer than civilian agencies, for chemical warfare had been a
military threat since World War I. Filter Spectroradiometers have been
developed and field tested under the extreme environmental conditions that
military operations face. Although some details of these devices are
30
-------
classified, the general approach to the detection, identification, and
mapping of chemical warfare plumes is veil known.
A severe threat comes from the nerve agents which all contain phosphorus.
The P-O-C bond in the nerve agent molecules has an absorption band within the
8-13 urn atmospheric window. The U. S. Navy has developed the AN/KAS-1
Chemical Warfare Directional Detector (CVDD) (NRC, 1984), a manually operated,
filter radiometer that can monitor high concentrations of chemical nerve
agents immediately after munitions have burst. These concentrations are far
in excess of the amounts of toxic pollutants that the U.S. EPA will wish to
detect, but the experience in developing and deploying the AN/KAS-1 (and the
Forward-Looking Infrared or FLIR detectors on which it is based) is directly
applicable to our problem. The three infrared filters that are sequentially
inserted into the optical path of the instrument produce different contrast
scenes to the observer who then determines subjectively if a nerve agent was
released. Although this data processing technique is primitive, the con-
siderable problems of deploying a cryogenic-detector, spectroradiometer in a
shipboard, noisy, corrosive environment have been solved.
There are more advanced versions of such radiometers under development by
the DOD. However, filter radiometers can only identify those pollutants for
which an appropriate filter is available. Thus they are of limited appli-
cability for plumes of unknown constituents. The ability to image the toxic
plume is a great advantage in determining its spatial extent and to provide
insight on whither it is advecting with the wind.
31
-------
Infrared spectral regions are not unique in possessing absorption signa-
tures of toxic pollutants; there are ultraviolet (UV) absorption character-
istics as veil in many noxious gases and vapors. The energetic UV photons
allow a variety of sensors to be used, such as silicon diodes, photomulti-
pliers and vidicons. Commercial available imaging systems can detect and
display pollutant plumes using scattered UV in sunlight (Saeger et al., 1988),
but such instruments may have only a limited number of pollutants that can be
detected and only if the proper filters are ready to use.
Dispersive Instruments
Dispersive instruments use either a prism or diffraction grating to
disperse the frequency spectrum that irradiates the instrument into a spatial
separation of frequencies. Classical spectroscopy used either photographic
plates to record the spectrum, or it vas scanned across a single detector for
detection and recording. With our improved technology in multi-element
detectors, multiplexed, parallel channels for data are feasible, thus
improving the throughput of the instrument in proportion to the number of
these channels. Examples of such infrared devices have been produced by the
Canadian companies, Moniteq and Barringer (MONITEQ, 1986). With linear arrays
of 1,000 detectors, or vith spectrally coded masks used over photo-multiplier
cathodes in th« UV, the spectral resolution and, thus, the pollutant discrimi-
nation of most such spectroradiometers will not be competitive vith interfero-
metric instruments but may still be quite adequate for many pollutants. The
simplicity of such instruments and relatively low cost of about $20K
32
-------
indicates that they should not be ruled out as having a role to play for
hydrocarbon and toxic pollutant remote sensing.
Gas-Filter Correlation Instruments
These instruments use a reference cell of the gas to be sensed to provide
a reference spectrum against which the spectrum from the monitored scene is to
be compared. They are therefore limited to sensing only those gases for which
a reference cell is available. For those specialized applications their
simplicity has much to offer for their sensitivity can reach fractions of ppm
levels (MONITEQ, 1986). Their incorporation into a generalized remote sensing
suite of instruments cannot be recommended.
Interferometers
The U. S. Army's XM-21 Remote Sensing Chemical Agent Alarm (RSCAAL)
spectroradiometer (NRC, 1984) has been through years of development and field
testing, mostly against DHMP (dimethyl methyl phosphonate), a compound that
closely duplicates the spectral absorption characteristics of nerve agents.
RSCAAL operates in the 8-13 urn vindov using the thermal emission from the
background and the toxic target. When a low-lying (thus warm) cloud of nerve
agent is viewed against a clear (thus cold) sky, the contrast of the target is
enhanced and detection is optimized. If there is terrain in the field-of-
view, and the terrain is at the same effective radiation temperature as the
cloud, then detection is not possible. However, this condition is not a
33
-------
significant problem for most interferometers because the temperature
difference between the cloud and the environment for a militarily-important
target is less than 0.1° C. Studies of atmospheric background temperatures
have indicated that a typical vapor cloud has at least a 10° C temperature
difference between it and its surroundings over 99X of the measured time over
all atmospheric conditions (Sutherland, 1982).
To recover the spectral information from an interferometer, the data can
be processed from either the time domain or the frequency domain. The
frequency domain processing generally requires a stable background emission
profile that does not contain absorption features of the species to be
measured. In many cases, this can be a difficult operational requirement.
Modern signal processing methodology has been used to extract the signal of
interest and eliminate many of the background spectral contributions (Small et
al., 1988). This methodology involves the use of finite impulse response or
infinite impulse response digital filters coupled with some form of automated
pattern recognition technique. This signal processing methodology has been
tested in several field tests in which an interferometer vas mounted on a U.S.
Army UH-1 helicopter and flown past a source of either DHNP or SF6.
Future generations of passive interferometers will use advanced signal
processing components. The digital signal processing microprocessors are a
family of specialized computer chips which have gained much attention in the
past several years. Their architecture has been optimized to perform
instructions in a single computer clock cycle and generally concentrate on
sum-of-product computations. These processors are ideal for many of the
34
-------
commonly used algorithms such as impulse response filters, fast Fourier
transformations, and matrix manipulations. Current performance of the small,
low-cost hardware indicates that a 100-fold improvement in processor speed is
available over the state-of-the-art, general microprocessors.
A number of instruments have been developed commercially (MONITEQ,1986),
some for the express purpose of field detection of toxic pollutants. In the
passive mode they have demonstrated the capability of ppm-m sensitivity for a
number of toxic pollutants.
Source-Augmented Radiometers or SAR's
The washout problems that passive spectroradiometers may experience can
be eliminated by providing such instruments with an active source. The
resulting instruments can be called "source-augmented radiometers" or SAR's.
Sources of radiation in the appropriate wavelength band may be located on the
far side of an area that needs to be monitored, or the source and radiometers
may be co-located, with only a mirror on the far side to reflect the radiation
back to the receiver. Such instruments provide only a line integral average
of the pollutant, but if desired (and if feasible), multiple mirrors may be
located at various distances along the path to provide some range discrimina-
tion. A rigid support, such as a sturdy tripod with a pan-tilt head can
serves as a mirror mount. For thermal infrared instruments, such as FTIR's,
the source is usually a globar (a ceramic heater rod) equipped with a
telescope collimator. Tunable laser sources can also serve for broad-band
35
-------
radiometers, eliminating the need for spectral discrimination in the receiver,
an instrument we may call a laser-SAR.
A source-augmented FTIR has recently been used at seven hazardous vaste
disposal sites in Nev Jersey and New York for preremedial investigations
(McLaren et al., 1989). A number of compounds were identified in real-time to
the ppb concentration levels. A minicomputer is required to provide the
computational power to provide real-time spectra, spectrum matching, and
identification. Even with present-day computers this requirement does not
preclude a mobile system in a small van, and in the future much more powerful
yet smaller data processors will be available. Costs for such units are in
the $50K to $100K range. Future improvements and additions may double this
price, but volume production will decrease it.
Industrialized versions of FTIR SAR's are already installed at several
U.S.A. locations and in Europe; improved units are under development. These
sensors perform automatically to detect, identify, alarm and record accidental
gas releases in operating plants or along fencelines.
Augmentation can also be used in the UV, where high-pressure, xenon-
arclamps provide intense sources. The other advantage to UV instruments is
that the energetic photons allow the use of room-temperature, linear arrays
of diode detectors. Field use of a 1024-diode silicon array in a long-path,
xenon source, retroreflector instrument showed ppb sensitivity at the pre-
viously mentioned hazardous waste disposal sites (McLaren et al., 1989).
36
-------
The UV electronic-transition spectra of target toxic pollutants, namely
aromatic hydrocarbons, are well known and have been used for 60 years for
laboratory, quantitative measurements, usually in the liquid phase; phase
makes little difference for these species at moderate spectral resolution. Up
to ten or more species in the same air volume may be separable. More complex
mixtures may lead to some shielding of strongly absorbing species by others.
37
-------
Appendix 4
Laser Re*ote Sensors
Introduction
Laser remote sensors or lidars for environmental monitoring have been in
use since 1963, shortly after the laser was invented. There are several
texts and reviews, available in most technical libraries that describe their
capabilities (Gauger and Hall, 1970; Hinkley, 1976; Svain and Davis, 1978;
Killinger and Mooradian, 1983; Measures, 1984; Kobayashi, 1987; Measures,
1988).
Current measurement capabilities in spectral regions that span the
ultraviolet, visible, and infrared wavelengths include plume opacity, aerosol
mass loading, 03, N02, and S02 monitoring, even aerosol spectrometric
chemical analysis by means of dielectric breakdown in a focused laser beam.
In this appendix we provide more details on the status and capabilities of
laser remote sensors.
38
-------
DIAL
A lidar that operates vith a laser at a wavelength that is absorbed by a
target species and one that operates at a nearby wavelength that is not so
absorbed is called a Differential Absorption Lidar or DIAL. Strictly
speaking, DIALs operate using range-resolved, aerosol backscatter; terrain or
mirrored return of the laser beam is also possible for improved, path-
integrated sensitivity in a technique known as laser long path monitoring or
more appropriately, laser SAR.
The DIAL concept was invented by Schotland (1966), and there is a wealth
of experience (and of caveats) in DIAL use. To be able to map a toxic pollu-
tant plume over distances of several kilometers with spatial resolution of a
few tens of meters requires lasers of significant energy with short pulses and
with mean power measured in tens of watts. Safety considerations will almost
certainly demand that plume mapping DIAL'S operate at "eye-safe" wavelengths
where the cornea of the eye is opaque. This rules out the visible and near-
IR regions.
Because many toxic pollutants have absorption bands in the wave length
region 9-11 Mm, a region covered by C02 lasers, and because C02 laser tech-
nology is quit* advanced, it is predicted that C02 DIAL'S and C02 laser long-
path will be the active instruments of choice in the 1990's. Present line
selection techniques, such as motor-driven micrometers or piezo-electric
transducers for diffraction gratings provide for selection of about 70 lines
at frequencies of a few hertz using "garden variety" 12C1602- This frequency
39
-------
of tuning is adequate for most DIAL applications. Nev developments of
thallium arsenic selenide (TAS) acousto-optic, intracavity, tunable filters
(Denes et al., 1988) promise to extend the tuning frequencies to 10 kHz as may
be required for rapidly changing concentrations over short ranges, where high
S/N ratios are available for each laser pulse. Rapid advances are also being
made in catalytic converters for C02 lasers, so that isotopic species such as
13C1602 and 12C1802 will be economic, and triple the number of lines available
for C02 DIAL'S (Gibson et al., 1979).
CO lasers operating in the 5 urn region are also feasible. The large
effort by the DOD on HP and DP lasers operating in the 3-5 urn region is
resulting in commercial units that can be safely operated at these shorter
wavelengths, but not yet at energies or powers suitable for DIAL. Alterna-
tively, C02 laser frequencies may be doubled (Henyuk et al., 1976) or even
tripled (Menyuk and Iseler, 1979) with efficiencies of from SZ to 10X into
this spectral region. Using such a crystal in a DIAL system, Peichtneret
et al. (1986) have detected ethane at a wavelength of 3.4106 urn. Optical
quality CdGeAs2 crystals used for this harmonic generation have been difficult
to produce, however. More development support and time is required before we
can predict if this DIAL approach will be a serious competitor for 9-11 urn
instruments. Recently the Soviet Institute of Atmospheric Optics at Tomsk has
claimed success in producing CdGeAs2 crystals for frequency doubling, as well
as ZnGePj and Te3AsSe}. If the Soviet claims check out, this may point the
way to greater utilization of the 3-5 urn region for DIAL or laser SAR's.
40
-------
Tunable laser sources for the near and intermediate IR have also been
investigated in research instruments. Menyuk and Killinger (1987) incor-
porated a continuously tunable Co:MgF2 laser into a DIAL and demonstrated
H20, HC1 and CH4 sensing at 40 p'pm at a distance of 200 m through the
atmosphere. Tuning was continuous over 1.5-2.3 urn, but the work has been
halted at least temporarily. Another approach is to pump an optical
parametric oscillator (OPO) with a short-wavelength laser. Eckardt et al.
(1986) have demonstrated that a AgGaSe2 crystal pumped by a Ho:YLF laser can
produce continuously tunable energy in the 3.3-5.6 urn with pulse energies of
O.SmJ, not enough for DIAL, but pointing to technology that should be
monitored during the 1990's.
There is already more than ten year's of field experience with C02 SAR's
and DIAL'S (Menzies and Shumate, 1976). Early work was usually done with
continuous wave (CV) lasers that integrated the amount of target species
between the transmitter and a mirror or terrain target, a laser SAR. Such CV
lasers operate at a lov plasma tube pressure, only a fraction of atmospheric.
However, the significant development effort by both DOD and DOE in high-power
TEA (transverse-excited, atmospheric pressure) lasers made available such
units for DIAL work in the early 1980's by Killinger et al. (1983) who also
demonstrated the improved signal available with heterodyne detection.
There is note averaging required with a heterodyne DIAL because the
single frequency, coherent beam leads to speckle noise in the return
(Hardesty, 1984), and the lidar system is also much more complex because of
the requirement for diffraction-limited alignment of the transmitter and
41
-------
local oscillator lasers. However, the apparent complexity of heterodyne
lidars should not rule out their being reduced to reliable practice. NASA is
now contracting for the LAVS (Laser Atmospheric Wind Sounder), a heterodyne
lidar to fly unattended in the 1990's on the Polar Platform. The lessons to
be learned in building this instrument will benefit those who wish to apply
heterodyne technology for other problems, such as toxic pollutant sensing. A
DIAL system that can also measure the vind in the toxic plume, an inherent
ability of coherent lidars, has certain advantages in air pollution studies
(Eberhard et al., 1989).
Airborne C02 DIAL'S have been used to map SFg plumes (Uthe, 1986), and
commercially available DIAL'S or laser SAR's can be rented or custom units
contracted for (Simpson, 1987). Note, however, that a DIAL must know what C02
line to tune to, and what line(s) vill be at a non-contaminated reference
wavelength. Present lasers can be tuned to 60 or 70 emission lines in the
9-11 ym region. If the exact constituents of the toxic plume are not known in
advance, some other intelligence must be available on these constituents
before a DIAL can be useful! When the constituents are known, C0} DIAL'S can
detect a wide variety of organic compounds and other toxic agents.
PM Laser SAR's
Frequency modulation (FH) spectroscopy using lasers is a technique where
the laser is modulated at a high frequency so that the modulation frequency
shows up as two sidebands on the laser line frequency. This is a laser SAR
technique. If the laser is operating near the edge of an absorption line of a
42
-------
target species, then the two sidebands will be differentially affected - the
sideband toward the line center will be more strongly absorbed, and the one
away from the line will be more weakly absorbed (Iran et al., 1984). Although
the technique has been primarily used in the laboratory to date as a laser
SAR, there is no reason that it cannot be extended to field, lidar use. The
Electric Power Research Institute (EPRI) is now concentrating their research
efforts for point measurements on this laser SAR technique (Hansen, 1988).
To use this technique for a variety of target species would require that a
narrow-line laser be tuned to the appropriate absorption line edge, and then
frequency modulated. Demonstrated laboratory sensitivities in the parts per
trillion range make this a technique to monitor closely as development
progresses.
This technique may be viewed as a development over ten years old based on
tunable diode lasers (TDL's) (Hinkley, 1976). Application of pressure to lead
salt semiconductor lasers provides a wide tuning range. With the continual
improvement in TDL's over the past few years, this technique certainly merits
close tracking for application to EPA remote sensing needs.
Raman Lidar
Raman lidar detects atmospheric species by detecting the Raman-shifted
backscatter (Cooney, 1970; Melfi and Vhiteman, 1985). Although this molecular
scattering mechanism is enhanced as the fourth-power of the frequency, the
Raman cross-section remains about three orders of magnitude below Rayleigh
(molecular) scattering. High power excimer lasers have recently made the
43
-------
technique attractive for abundant species in the atmosphere such as vater
vapor (Petri et al., 1982), but it is predicted that Raman techniques will not
be competitive for ppm sensing, even into the 1990's because of the extremely
low backseatter coefficients and interference from aerosol fluorescence.
Fluorescence Lidar
A fluorescence or resonance scattering lidar is one that employs the
absorption and subsequent spontaneous emission of the irradiating laser
photons to produce the returned signal. This technique has been used with
great success in probing atmospheric constituents in the upper atmosphere
(Hegie, 1988). However, this technique does not work well in the lower
atmosphere because the higher pressure almost always leads to quenching
collisions with the excited molecule before it has a chance to emit its
characteristic radiation. The technique may have some limited application in
the detection of aerosols, dusts, and films, but there is little chance that
it has application for airborne, hydrocarbon or toxic pollutant plumes,
except when fluorescent aerosols are intentionally injected into the plume as
a sensible tracer.
-------
Appendix 5
Examples of Remote Sensor Successes
During early autumn of 1988, measurements of airborne toxic pollutants
and hydrocarbons vere conducted at seven potential Superfund sites in New York
and New Jersey (Minnich et al., 1989; McLaren et al., 1989). The FTIR SAR
used during these field measurements detected eight different compounds, all
in the lov ppm-m range, listed in Appendix 7. Path lengths up to 400 m to the
folding mirror vere used. Set up time was less than one hour after arriving
at each site. A standard Nicolet Model 730 FTIR system was used, supplemented
by 37 cm aperture Cassegrain telescopes for both the radiometer and the globar
source. Signal processing electronics included an 18-bit analog/digital
converter and a Nicolet 620 data station that is built around a minicomputer
with 50 m byte hard disk.
The success obtained in identifying pollutants during these field tests
indicates that FTIR remote sensing techniques are at a stage of maturity that
enable significant results to be quickly obtained for concentration levels of
importance in Monitoring toxic waste sites. In addition, visible and near-IR
lidars have long been used for studying pollutant transport and diffusion in
plumes and air masses (e.g., Eberhard and McNice, 1986; Vakimoto and McElroy,
1986; McElroy and Smith, 1986; McElroy, 1987) or fluorescent tracers in such
45
-------
plumes (Uthe, et al., 1985). This substantiates the statement made earlier,
that remote sensing is not an "ivory tower" concept, but one that is now ready
to help solve EPA hydrocarbon and toxic pollutant sensing problems.
The NOAA Wave Propagation Laboratory's pulsed Doppler lidar uses a
coherent C02 laser at 10.6 urn wavelength. The lidar measures the back-
scattered radiance to determine turbidity from large aerosols and also deter-
mines the wind field. This is accomplished by measuring the Doppler frequency
shift caused by the scattering aerosols' motion as they advect with the wind.
A range-resolved component of the wind parallel to the laser beam can be
displayed in real time at slant ranges of 20 km or more in the lower atmos-
phere. A number of successful experiments have been accomplished since the
lidar was placed in operation in 1981 (Hall et al., 1987). IR DIAL measure-
ments are accomplished by time sequencing the laser between absorbing and
window lines.
This lidar participated during December 1987 as part of the Denver Brown
Cloud Study, documenting the meteorology of aerosol and CO polluting events
during wintertime in a high-elevation metropolitan region (Eberhard et al.,
1989). The lidar mapped the aerosol cloud out to distances of typically 20 km
in north-east Denver. Simultaneous vind data from the lidar revealed much
about the associated mesoscale meteorology. The lidar confirmed the existence
of a weak drainage flov toward the northeast, followed by a return flow driven
by typical pressure gradients that carry polluted air back to Denver. It also
showed how stagnant, polluted air could cover part of the metropolitan area
-------
while clean air was present in other zones where strong mountain winds
extended to the surface.
-------
Appendix 6
Future Improvements Possible in Remote Sensor Performance
In this appendix ve examine the performance equations of the promising
remote sensors discussed in Appendices 3 and A to understand where improve-
ments in performance may be expected in such instruments over the next five
to ten years.
Spectroradiometers
There are fundamental, physics-imposed limitations to radiometer
performance. These can be best studied by examining the S/N (signal-to-noise)
equations that apply to radiometers. First, an equation that applies to any
type of radiometer is presented. The nomenclature of Silva (1978) is used to
show that the noise equivalent pover (NEP) of a photon-detector radiometer
is,
( A *8F )«
NEP
D*
where A is the area of the detector, SF is the electronic bandwidth of the
radiometer amplifier, and D* is the detectivity of the detector. The value of
D* is, of course, a function of wavelength, but we will not list this
subscript to avoid clutter in the equations. Detector area, A, is related to
48
-------
the field of view (FOV 9) of the radiometer, and the effective optics focal
length, f by,
9 * /Dt
since the detector is the limiting field stop, and the FOV is usually small.
The effective focal length is used because most radiometers do not form an
image of the source onto the detector but rather use field lenses or condens-
ing optics to increase the "speed", E, or f/ratio of the optics. E D/f,
where D is the entrance aperture diameter.
The input-signal power, *, to the audiometer detector over the spectral
increment employed is,
where r is the atmospheric transmittance from the source to the radiometer, L
is the radiance of the source, and e is the optical efficiency of the
radiometer. In order to maximize
n>LDE0D*
S/N - «/NEP » 57
the numerator tens must be made as large as possible and the denominator as
small as possible. Only the numerator variables can be improved, for the
electronic bandwidth is determined by the "scan" time, which is in turn
dictated by inherent atmospheric variability in the pollutant volume or by the
49
-------
time available to measure the plume from different aspects, and these times
will probably not change over the next ten years.
Taking the numerator terms in order, first, the wavelength increment in
the atmosphere that has maximum transmittance must be used, but since the
8-13 urn vindov is already used in many devices, and the transmittance is
close to 1.0 over path lengths of a fev kilometers, there is not much opti-
mizing left to do vith r . With passive radiometers, there is not much that
can be done vith the source radiance, L, since this is most often the
Planckian thermal scene radiance. Vith source-augmented radiometers, however,
there are improvements that can be made. The globar sources commonly used
for FTIR's operate at a maximum temperature of about 1400K. As pointed out by
Smith et al. (1957) specially constructed tungsten, v-grooved filament lamps
can operate at more than twice this temperature. Experimental tests shov an
8-13 um radiance tvice that of globar. Vith the improvements in IR vindov
materials in the past 20 years, only a modest amount of development effort
should be needed to improve sources for augmented radiometers. Increasing L
by a factor of 2 provides 3 dB more signal.
Some improvement in D* can be expected, but many detectors are already
very close to the quantum noise limits. Improvements in cryogenic enclosures
and radiation shields can offer some improvement for thermal-IR detectors, and
these may improve D* by a factor of 2, or 3 dB in signal strength.
Note that if a radiometer could be devised vhich employed multiple detec-
tors for examining the pollutant volume, the S/N improves as the square-root
50
-------
of the number of detectors, because the signal in each of the detectors is
correlated but the noise is not. One technology that has seen dramatic
improvement in the past fev years, and that promises to continue improving in
the future is in the ability to produce multi-element, solid state detectors
in linear or areal arrays. It seems reasonable to forecast that in five to
ten years arrays of ten to thirty times greater complexity than present
designs will be available. What is not obvious is hov to advantageously
employ multiple element detectors in FTIR's, although they could be used to
average-out speckle noise in coherent DIAL'S; multiple detectors have already
proven their advantage in dispersive instruments.
Another improvement may come from the optics aperture. Light-veight
optics are another area where significant recent improvements have been
shown, with further advances most probable. In field instruments it is often
the weight of the device that limits its utility, so larger optics diameters
need not impact on an instrument's mobility. An increase of 0 by a factor of
2 is an easy way to achieve another 3 dB of signal. So if we add together
all of these improvements, we find that 9 dB (or a factor of nearly 10) can
be expected over the next fev years in radiometer sensitivity enhancement.
Since FTIR performance is strongly determined by digital computing speed
and by software sophistication, it is predicted that a future FTIR will be
improved overall in three categories: (1) In the number of pollutants that it
can store in memory for real-time identification, (2) In the speed with which
these identifications can be made, and (3) In improved sensitivity into the
fraction of ppb range - the 9 dB, or factor of nearly ten, enhancement
51
-------
predicted above. There are no critical component weaknesses to exploit in
order to achieve greater improvements. Costs of mobile FTIR's with their
real-time computers will be in the range of $50K to $200K, 1988 dollars. UV
dispersive instruments should be somewhat less expensive.
Laser Remote Sensors
The S/N equation for a lidar is not too different than that for a
radiometer since the lidar receiver is really just a radiometer adapted for
the high frequency response required to achieve adequate range resolution.
Since the laser radiation backseat tered from atmospheric or topographic tar-
gets decreases as the inverse-square of the range, R, to the scatterer, the
energy, E, returned to the detector is,
E = E. T R-2(nD2/4)(3(ct/2)
it a
where E, is the energy in the laser pulse of duration t, 3 is the backseat ter
coefficient of the target, c is the velocity of light, and the other symbols
have the same meaning as on pgs. 48-49. This returned energy is converted to
a signal current, i , according to the expression,
s
is - *r
where the signal pover is simply * * E/t and T is the responsivity of the
detector and amplifier (usually given in amperes per watt).
52
-------
The predominant noise term appropriate for the lidar equation will vary
according to the spectral region and thus the type of detector that will be
used (Hall, 1974). In general, the noise can be attributed to a sum of these
noise currents,
in - [2eSF(is
where e is the electronic charge, and the subscripts b and d stand for back-
ground and dark respectively.
Vhen looking for those improvements in lidar (DIAL) system performance
that may be expected over the next ten years, it is expected that improved
light weight optics will make an increase in D possible without impacting the
system mobility. An increase in optics area by a factor of 2 for a 3 dB gain
in signal will be feasible. The energy available from line-tunable or
continuously tunable sources will also increase as laser efficiencies and
materials are improved, both for gaseous and solid state units. An improve-
ment in energy by a factor of 3, or 5 dB is frugal.
The potential for the greatest improvement breakthrough exists with
infrared detectors. Present detectors, such as the ubiquitous HgCdTe
operated at 77°K have an extremely low responsivity, and therefore require
high-gain preanplifiers before the signals can be operated upon. But to
achieve the frequency response required for a pulsed DIAL, the fundamental
Johnson noise in the preamp swamps the meager detector signal. This is the
principal reason why heterodyne systems have been developed, to amplify
53
-------
optically the signal to a level where the preamp noise is negligible.
Avalanche photodiodes are just becoming available in the near infrared (Brown
and Ridley, 1986), and liquid helium-cooled devices have been demonstrated for
the thermal infrared (Stapelbroek, 1987). If a practical, field-deployable,
avalanche detector can be developed for C02 DIALs, the sensitivity of a
heterodyne DIAL may be approached without the speckle limitations or the com-
plications of interferometer alignment. Sensitivity could then be increased
by a factor of 10 to 100 (or equivalently, 10 to 20 dB).
Overall lidar or DIAL sensitivity can be expected to improve by a factor
of six (or 8 dB) in the next ten years without detector improvements. If
avalanche detectors become available, the total improvement could be factors
of 60 to 160 (or 18 to 28 dB). These 1990's DIALs will also be much more
versatile, reliable, and sophisticated compared to today's units.
Laser remote sensor overall performance can thus be improved in three
categories also: (1) In the number of lines that can be tuned to quickly, from
the present 70 or so to perhaps 200 for C02 instruments (by using isotopes of
C02; prototypes of such lasers have been built) and over wider increments in
the UV, visible, and near IR using nev solid state lasers, (2) In the speed
with which this tuning can be accomplished, and (3) In sensitivity into the
fraction of ppa range for sensing ranges of several kilometers and into the
ppb range for shorter (* 10 m to 100 m)ranges. C02 laser technology is more
advanced than competing sources for DIAL, and will likely remain so into the
1990's. Avalanche detectors present a critical component development oppor-
tunity for thermal IR DIAL'S, and much work remains on improving tunable diode
54
-------
and tunable solid state lasers. Costs for mobile C02 OIALs will be in the
S150K to S500K range in 1989 dollars.
55
-------
Appendix 7
Reacte Sensing Detection Levels for Hydrocarbons and Toxics
The table that follows lists examples of compounds that have been
detected using existing sensors under field conditions.
Pollutant
Phosgene
CFC-11
CFC-12
NH3
Ethylene
oxide
Propylene
oxide
o-xylene
Toluene
Chloroform
Methanol
Acetone
p-xylene
1,1,1-tri-
chloroe thane. .
1,3,5-tri-
methylbenzene
Ethylene
Trichloro-
ethylene
Chloro-
benzene
Absorption
Spectrum
Band, ym
11.8
10.4, 10.7
10.7-11.9
11.4-12.2
7.1-14.3
n
n
n
n
n
N
If
II .
10P(14),(12)
10P(20),(10)
9R(26),9P(12)
Spectro- Laser (DIAL)
radiometer Sensor
Detection Detection
Limit Limit Reference
"1 ppm.m Mobay Corporation
(see p. 20)
"1 ppm.m "
"1 ppm.m "
"1 ppm.m "
"30 ppm.m "
n
"30 ppm.m "
n
"33 ppm.m Hinnich et al.
(1989)
"10 ppm.m "
< 4 ppm.m "
"1.4 ppm.m "
"23 ppm.m "
<30 ppm.m "
"3 ppm.m "
<3 ppm.m "
0.3 ppm.m Grant (1989)
0.9 ppm.m "
5.3 ppm.m "
56
-------
Appendix 8
Attendees at the EPA Workshop on Remote Sensing of Hydrocarbons and
Toxic Pollutants, Las Vegas, Nevada, 6-7 April 1989.
Bath, Raymond
Behar, Joe
Carsvell, Allan
Daubner, Ludo
Demirgian, Jack
Diebel, Dorothee
Emery, Silvio
Fovler, Robert
NUS Corporation
1090 King George Post Road
Edison, NJ 08837
(201) 225-6160
U.S. EPA-EMSL-LV
P.O. Box 9378,
Las Vegas, NV 89193-3478
(702) 798-2216
Physics Dept., York University
4700 Keele St.
N. York, Ontario, Canada M3J P3
416) 736-2100
x7755
Barringer Research
304 Carlingview Dr.
Rexdale, Ontario, Canada M9V3H5
(416)675-3870
Argonne National Lab.
9700 Cass Ave.,
Argonne, IL 60439
(312) 9 2-6807
UNLV/ERC
4505 Maryland Parkway
Las Vegas, NV 89154
(702) 798-2435
U.S. Army CRDEC
SMCCR-RSP,
Aberdeen Proving Gnd., HD 21010
(301) 671-3518
Hughes Aircraft Co.
E55/G223, P.O. Box 902
El Segundo, CA 90245
(213) 373-2925
57
-------
Gillespie, James
Grant, William
Hall, Freeman
Herget, Bill
Karl, Robert
Kert, John
Kibby, Hal
Killinger, Dennis
Klein, Volker
Knapp, Ken
U.S. Array
Atmospheric Sciences Lab.
SLCAS-AR-P
WSMR, NM 88002
(505) 678-6609
NASA Langley Research Center
MS 401A
Hampton, VA 23665-5225
(804) 864-5846
UNLV/ERC
202 Ocean Street
Solana Beach, CA 92075
(619) 259-2721
Nicolet Instrument Corp.
5225 Verona Rd.
Madison, VI 53719
(608) 271-3333
Los Alamos Natl. Lab.
MS-J-567, P.O. Box 1663
Los Alamos, NM 87545
(505) 667-4702
Hughes Aircraft Co.
P.O. Box 902
El Segundo, CA 90245
(213) 616-0008
U.S. EPA EMSL-LV
P.O. Box 93478,
Las Vegas, NV 89193-3478
(702) 798-2522
Dept. of Physics
Univ. of South Florida
Tafflpa, FL 33620
(813) 974-3995
Battelle Institute
Am Roemerhof 35
D-6000 Frankfurt 90, F.R.G.
49-69-7908-2859
U.S. EPA-AREAL
Research Triangle Park, NC 27711
(919) 541-2454
58
-------
Kroutil, Robert
Leonard, Donald
Leonelli, Joe
Lewis, Harry
U.S. Army CRDEC
SMCCR-RSL
Aberdee Proving Gnd.t MD 21010
(301) 671-3021
GTE Govt. Sys. Corp.
P.O. Box 7188
Mountain View, CA 94039
(415) 966-3707
SRI International
333 Ravenswood
Menlo Park, CA 94025
(415) 859-2326
Contraves Corp.
Block Engr. Grp.
28 Travis St.
Boston, HA 02134
(617) 254-4401
Aerospace Corp.
M2-253, P.O. Box 92957
Los Angeles, CA 90009
(213) 336-7418
U.S. EPA EMSL-LV
P.O. Box 93478
Las Vegas, NV 89193-3478
(702) 798-2361
ERC/UNLV
4505 Maryland Parkway
Las Vegas, NV 89154
(702) 798-2359
Battelle Northvest
P.O. Box 999
Richland, VA 99352
(509) 375-2044
CIRES
Box 0216
University of Colorado
Boulder, CO 80309
(303) 497-6509
Sandridge, Robert Research Dept. Consultant
Mobay Corp.
New Martinsville, WV 26155
(304) 455-4400
x2207
Loper, Gary
McElroy, James
McGown, Hike
HcVeety, Bruce
Rye, Barry
59
-------
Shrieves, Van
Simpson, Orman
Spear, Richard
Spellicy, Robt.
Stedman, Donald
Uthe, Ed
Visvanathan,
Ramesh
Vilkerson, Tom
Williams, Llcv
Zimmermann,
Rainer
U.S. EPA-Region 4
345 Courtland St.
Atlanta, GA 30312
(404) 347-2864
TECAN Remote (ELS)
3000 Northwoods Pky. #185
Atlanta, GA 30071
(404) 242-0977
U.S. EPA-Region 2
Voodbridge Ave.
Edison, NJ 08837
(201) 321-6686
Optimetrics, Inc.
106 E. Idaho
Las Cruces, NM 88001
(505) 523-4986
Chemistry Dept.
Univ. of Denver
University Pk
Denver, CO 80208
(303) 871-3530
SRI International
333 Ravensvood
Menlo Park, CA 94025
(415) 859-4667
ERC/UNL
4505 Maryland Parkway
Las Vegas, NV 89154
(702) 798-2268
Univ. of Maryland
Inst. for Phys. Sci. & Tech.
College Park, MD 20742-2431
(301) 454-5401
U.S. EPA/EMSL-LV
P.O. Box 93478
Las Vegas, NV 89193-3478
(702) 798-2109
UNLV/ERC
4505 Maryland Parkvay
Las Vegas, NV 89154
(702) 798-2435
60
-------
References
Brovn, R.G.V. and K.D. Ridley, 1986. Miniature solid-state photo detectors
for photon correlation spectroscopy and laser anemometry, Technical
Digest, CLEO '86, Opt. Soc. Amer./IEEE, San Francisco, CA, p. 348.
Cooney, J., 1970. Laser Raman probing of the atmosphere, in Laser Applica-
tions in the Geosciences, Edited by J. Gauger & F.F. Hall, Jr., Western
Periodicals, N. Hollywood, CA, pp. 51-68.
Denes, L.J., M. Gottlieb, N.B. Singh, D.R. Suhre, H. Buhay, and J.J. Conroy,
1988. Rapid tuning mechanism for C02 lasers, Society of Photo-Optical
Instr. Engrs., 0-E/LASE '88 paper.
Eberhard, V.L. and G.T. McNice, 1986. Versatile lidar for atmospheric
studies, including plume dispersion, clouds, and stratospheric aerosol.
J. Atmos. & Oceanic Tech. 3, 614-622.
Eberhard, V.L., R.E. Cupp, and V.D. Neff, 1989. Vind fields and aerosol
distributions in the Denver Brovn Cloud observed by C02 Doppler lidar.
Preprint Volume, 6th Joint Conference on Applications of Air Pollution
Meteorology, Jan. 30 - Feb. 3, Anaheim, Amer. Meteor. Soc., 117-119.
Eckardt, R.C., Y.X. Fan, and R.L. Byer, 1986. Tunable IR optical parametric
oscillator using silver gallium selenide, Tech. Digest, CLEO '86, Opt.
Soc. Amer./IEEE, San Francisco, CA, p.52
Feichtner, J.D., J.G. Havley, A. Rosengreen, and J.E. van der Laan, 1986.
Remote detection of ethane using C02 laser radiation upconverted to 2932
cm'1, Technical Digest, CLEO '86, Opt. Soc. Amer./IEEE,San Francisco, CA,
p. 338.
Gauger, J. and F.F. Hall, Jr., 1970. Laser Applications in the Geosciences,
Vestern Periodicals, N. Hollywood, CA, 283 pp.
Gibson, R.B., K. Boger, and A. Javan, 1979. Mixed isotope multi-atmosphere
C02 laser, IEEE J. Quan. Elec. QE-15, 1224-1227.
Grant, V.B., 1989. The Mobile Atmospheric Pollutant Mapping (MAPM) system: A
coherent C02 DIAL system. Proc. SPIE 1062, paper 29.
Hall, F.F., Jr., 1974. Laser systems for monitoring the environment, in Laser
ApplicatJena, Vol. 2, Edited by M. Ross, Academic Press, New York, pp.
161-225.
Hall, F.F., Jr., R.E. Cupp, R.M. Hardesty, T.R. Lawrence, M.J. Post, R.A.
Richter, and B.F. Weber, 1987. Six years of pulsed-Doppler lidar field
experiments at NOAA/WPL. Preprint Volume, 6th Symposium on Meteoro-
logical Observations and Instrumentation, Jan. 12-16, New Orleans, Amer.
Meteor. Soc., 11-14.
61
-------
Hansen, O.A., 1988. Personal communication from EPRI.
Hardesty, R.M., 1984. Measurement of range-resolved water vapor concentration
by coherent C02 differential absorption lidar, NOAA Tech. Memo. ERL-
VPL-118, Boulder, CO, 263 pp.
Hinkley, E.D., Editor, 1976, Laser Monitoring of the Atmosphere, Springer-
Verlag, New York, 380 pp.
Ingalls, M.N., 1989. On-road vehicle emission factors from measurements in a
Los Angeles area tunnel. Paper 89-137.3, Air & Waste Management Assn.,
Annual Mtg., Anaheim, Calif., June 25-30.
Killinger, O.K. and A. Mooradian, Editors, 1983. Optical and Laser Remote
Sensing, Springer-Verlag, Nev York, 383 pp.
Killinger, O.K., N. Menyuk, and V.E. DeFeo, 1983. Experimental comparison of
heterodyne and direct detection for pulsed differential absorption CO.
lidar, Appl. Opt. 22, 683-689.
Kobayashi, T., 1987. Techniques for laser remote sensing of the Environment.
Remote Sensing Reviews, 3, 1-56.
McElroy, J.L. and T.B. Smith, 1986. Vertical pollutant'distributions and
boundary layer structure observed by airborne lidar near the complex
Southern California coastline. Atmos. Environ. 20, 1555-1566.
McElroy, J.L., 1987. Estimation of pollutant transport and concentration
distributions in complex terrain of Southern California using airborne
lidar. J. Air Pollution Control Assn., 37, 1046-1051.
McLaren, S.E., D.H. Stedman, G.A. Bishop, M.R. Burkhardt, and C.P. DiGuardia,
1989: Remote sensing of aromatic hydrocarbons using long path ultra-
violet spectroscopy, Paper presented at June Air & Vaste Management
Assn., Annual Meeting, Anaheim, California.
Measures, R.M., 1984. Laser Remote Sensing, John Viley & Sons, NY.
Measures, R.M., Editor, 1988. Laser Remote Chemical Analysis, John Viley &
Sons, Nev York, 546 pp.
Megie, G., 1988. Laser measurements of atmospheric trace constituents,
Chapter 5 in Laser Remote Chemical Analysis, edited by R.M. Measures,
John Wiley & Sons, NY. 333-408.
Melfi, S.H. and D. Vhiteman, 1985. Observation of lower-atmosphere moisture
structure and its evolution using a Raman 1 dar. Bull. Amer. Meteor.
Soc. 66, 1288-1293.
Menyuk, N., G.W. Iseler, and A. Mooradian, 1976. High-efficiency high-average
power second harmonic generation with CdGeAs2, Appl. Phys. Lett. 29,
422-424.
62
-------
Menyuk, N. and G.W. Iseler, 1979. Efficient frequency tripling of CO -laser
radiation in tandem CdGeAs2 crystals, Opt. Lett. 4, 55-57.
Menyuk, N. and O.K. Killinger, 1987. Atmospheric remote sensing of water
vapor, HC1 and CH using a continuously tunable Co:MgF, laser, Appl. Opt.
26, 3061-3065.
Menzies, R.T. and M.S. Shumate, 1976. Remote measurements of ambient air
pollutants with a bistatic laser system, Appl. Opt. 15, 2080-2090.
Minnich, T.R., R.D. Spear, O.A. Simpson, J. Faust, W.F. Herget, D.H. Stedman,
S.E. McLaren, and W.M. Vaughan, 1989. Remote sensing of air toxics for
pre-remedial hazardous waste site investigations, Paper presented at
June 25-30, Air & Waste Management Assn., Annual Meeting, Anaheim,
California.
MONITEQ, Ltd., 1986. Applications of remote sensing systems for monitoring
airborne pollutants emitted from toxic chemical spills and hazardous
waste disposal sites, Report on EPA Contract No. 68-03-3245, MONITEQ,
Ltd., Concord, Ontario, Canada L4K. 2H7, 63 pp.
National Research Council, 1984. Assessment of Chemical and Biological Sensor
Technologies, National Academy Press, Washington, O.C., 110 pp.
Petri, K., A. Salik, and J. Cooney, 1982. Variable-wavelength solar-blind
Raman lidar for remote measurement of atmospheric water-vapor
concentration and temperature. Appl. Opt. 21, 1212-1218.
Saeger, M.L., C.K. Sokol, S.J. Coffey, R.S. Wright, W.E. Farthing, and
K. Baughman, 1988. A review of methods for remote sensing of atmospheric
emissions from stationary sources. Report on EPA Project Number
688-02-4442, Research Triangle Institute, Research Triangle Park, NC,
p. 89.
Schotland, R.M., 1966. Some observations of the vertical profile water vapor
by means of a laser optical radar. Proc. Fourth Symposium, Remote
Sensing of Environment, U. Mich., Ann Arbor, MI, pp. 273-283.
Silva, L.F., 1978. Radiation and instrumentation in remote sensing, Chap. 2
in Remote Sensing; The Quantitative Approach, edited by P.H. Swain,
McGraw-Hill International Book Co., pp 21-135.
Simpson, O.A., 1987. Remote sensing technologies for hazardous gas detection,
Sensors (July), Helmers Publishing, Inc., Peterborough, NH, p. 3.
Small, G.W., R.T. Kroutil, J.T. Oitillo, and W.R. Loerop, 1988. Detection of
atmospheric pollutants by direct analysis of passive Fourier transform
infrared interferograms. Anal. Chem., 60, 264-269.
Smith, R.A., F E. Jones, and R.P. Chasmar, 1957. The Detection and Measure-
ment of Infra-red Radiation, Oxford, Clarendon Press, Oxford University,
London, 458 pp.
63
-------
Stapelbroek, M.G., 1987. Photon-counting solid-state photomultiplier, J.
Opt. Soc. Amer. A, 4, P117.
Stockberger, L., K.P. Knapp, and T.G. Ellestad, 1989. Overview and analysis
of hydrocarbon samples during the Summer Southern California Air Quality
Studies. Paper 89-139.1, Air & Waste Management Assn., Annual Mtg,
Anaheim, Calif., June 25-30.
Sutherland, K., 1982. XM21 Remote Sensing Chemical Agent Alarm, Final
Scientific & Technical Report, CDRL Sequence No. AOOG, Contract No.
DAAK11-79-C-0051, Prepared for ARRADCOM, Aberdeen Proving Ground, MD by
Honeywell Aerospace & Defense Group, Clearwater, PL, p. 324.
Swain, P.H. and S.M. Davis, Editors, 1978. Remote Sensing; The Quantitative
Approach, McGraw-Hill, Inc., New York, 396 pp. Tran, N.H., R. Kachru, P.
Fillet, H.B. van Linden van den Heuvell, T.F. Gallagher, and J.P. Vatjen,
1984. Frequency-modulation spectroscopy with a pulsed dye laser:
experimental investigations of sensitivity and useful features. Appl.
Opt. 23, 1353-1360.
Uthe, E.E., 1986. Airborne C02 DIAL measurement of atmospheric tracer gas
concentration distributions, Appl. Opt. 25, 2492-98.
Uthe, E.E., V. Viezee, B.M. Morley, and J.K.S. Ching, 1985. Airborne lidar
tracking of fluorescent tracers for atmospheric transport and diffusion
studies. Bull. Amer. Meteor. Soc., 66, 1255-1262.
Vakimoto, R.M. and J.L. McElroy, 1986. Lidar observation of elevated
pollution layers over Los Angeles. J. Climate & Appl. Meteor., 25,
1583-1599.
64
------- |