PROCEEDINGS
SECOND CONFERENCE
ON
ENVIRONMENTAL QUALITY SENSORS
National Environmental Research Center
Las Vegas, Nevada
October 10-11, 1973
Environmental Protection Agency
Office of Research and Development
Office of Monitoring Systems
Washington, D.C. 20460
-------
DECEMBER 10, 1973
PROCEEDINGS
SECOND CONFERENCE
ON
ENVIRONMENTAL QUALITY SENSORS
National Environmental Research Center
Las Vegas, Nevada
October 10-11, 1973
Environmental Protection Agency
Office of Research And Development
Office of Monitoring Systems
Washington, B.C. 20460
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.0.20402 - Prlca $8.25
-------
FOREWORD
The Second Conference on Environmental Quality Sensors
provided an interchange of information between the Environ-
mental Protection Agency's (EPA) research scientists who are
developing and testing improved ways of monitoring the quality
of the environment and the Agency's ten Regional Offices. In
order to achieve the desired communications and response from
the participants, the attendance was limited to those in EPA
who are utilizing, or are planning to employ remote sensing
technology in their planning and regulatory function.
Participating with EPA in the conference were other Federal
and State agencies, universities and several industrial
research organizations. Complete texts of the technical
papers are included and constitute the major portion of
this document. Presentations by regional and program offices
focused on how remote sensing could be effectively utilized
in their respective geographical areas. Numerous applications
and program requirements were presented and are being reviewed
with the Regions.
The program and agenda were prepared by the Steering
Committee consisting of sensor experts within the agency.
They are as follows:
Murray Felsher - Office of Enforcement and
General Counsel
Robert F. Holmes - Office of Monitoring Systems
Donald R. Jones - Office of Hazardous Materials
Office of Air and Water Programs
John D. Koutsandreas - Office of Monitoring Systems
Anthony F. Mentink - NERC, Cincinnati
John W. Scotton - Office of Monitoring Systems
Donald T. Wruble - NERC, Las Vegas
The technical presentations are arranged in the order
that they appear in the Conference Program Agenda.
cr^'* 1) - Xc j7t~SV-vi .UO
John D. Koutsandreas
Conference Chairman
December 10, 1973
-------
TABLE OF CONTENTS
Page
Preliminary Report Of The Second Environmental
Quality Sensor Conference
Conference Agenda ........................................ ax
Conference Key Note Address
Willis B. Foster .......................................
Session I - NERC-Las Vegas Programs
Integrated Remote Sensing System
Leslie M. Dunn and Albert E. Pressman ................ I - 1
Aerial Air Pollution Sensing Techniques
R. B . Evans .......................................... I - 32
Lidar For Remote Monitoring
S. H. Melfi .......................................... I- 39
Evaluation of the Zeeman Atomic Absorption
Spectrometer
E . w. Bretthauer ..................................... I - 44
Session II — NASA-Langley Sensor Programs
Multiwavelength Lidar For Remote Sensing of
Chlorophyll a In Phytoplankton
Peter B. Mumola, Olin Jarrett, Jr., and
Clarence A. Brown, Jr .............................. .. II - 1
Remote Detection of Water Pollution With MOCS:
An Imaging Multispectral Scanner
Gary W. Grew .................................. . ...... II - 17
Summary of NASA Langley Research Center Remote
Sensing Activities Under The Environmental
Protection Agency Interagency Agreements :
Potential For Regional Applications
James L. Raper .................. . .................... II - 40
The Use of Near- Infrared Reflected Sunlight
For Biodegradable Pollution Monitoring
Walter E. Bressett and Donald E. Lear, Jr ............ II - 69
-------
Session III - Air Quality Sensor Developments
St. Louis Regional Air Monitoring System
James A. Reagan Ill _ i
Long Path Optical Measurement of Atmospheric
Pollutants
Andrew E. O'Keeffe Ill _ 17
A Review of Available Techniques For Coupling
Continuous Gaseous. Pollutant Monitors to
Emission Sources
James B. Homolya Ill _ 22
Session IV — Water Quality Sensor Development
Insitu Sensor Systems For Water Quality
Measurement
K. H. Mancy IV - 1
Continuous Monitoring By Ion Selective Electrodes
Julian B. Andelman • IV - 27
A Regional Water Quality Monitoring System
Russell H. Susag IV - 38
Water Quality Monitoring In Some Eastern
European Countries
Peter A. Krenkel and Vladimir Novotny IV - 58
An Airborne Laser Fluorosensor For The Detection
Of Oil On Water
H. H. Kim and G. D. Hickman IV - 85
Session V- Environmental Thematic Mapping
The U.S. Geological Survey and Land Use Mapping
James R. Anderson V - 1
Remote Sensing Data, A Basis For Monitoring
Systems Design
James V. Taranik and Samuel J. Tuthill V - 17
Techniques And Procedures For Quantitative
Water Surface Temperature Surveys Using
Airborne Sensor
E. Lee Tilton, III and Kenneth R. Daughtrey V - 32
Remote Sensor Imagery Analysis For Environmental
Impact Assessment
C. P. Weatherspoon, J. N. Rinker, R. E. Frost
and T. E. Eastler V - 59
Aerial Spill Prevention Surveillance During
Sub-Optimum Weather
P. M. Maughan, R. I. Welch and A. D. Marmelstein .... V - 73
-------
Session VI - Oil And Hazardous Materials Sensors
Single Wavelength Fluorescence Excitation For
On-Site Oil Spill Identification
J. Richard Jadamec VI - 1
A Remote Detection And Identification Of
Surface Oil Spills
Herbert R. Gram ^ VI - 21
Environmental Keys For Oil And Hazardous
Materials
Charles L. Rudder and Charles J. Reinheimer VI - 34
Video Detection Of Oil Spills
John P; Millard and Gerald F. Woolever VI - 47
Measurements Of The Distribution And Volume Of
Sea-Surface Oil Spills Using Multifrequency
Microwave Radiometry
J. P. Hollinger and R. A. Mennella VI - 65
The National Environmental Monitoring System
Theodore Maj or VI - 75
Session VII - Satellite Environmental Monitoring Applications
Skylab Application To Environmental Monitoring
Victor S. Whitehead VII - 1
Environmental Applications Of The Earth Resources
Technology Satellite
Dorothy T. Schultz and Charles C. Schnetzler VII - 11
ERTS-1 Detection Of Acid Mine Drainage Sources
Elliott D. DeGraff and Edward Berard VII - 27
Water Quality Analysis Using ERTS-A Data
H. Kritikos, L. Yorinks and H. Smith VII - 37
Satellite Studies Of Turbidity, Waste Disposal
Plumes And Pollution-Concentrating Water
Boundaries
V. KLemas VII - 57
Data Collection Platforms For Environmental
Monitoring
J. Earle Painter VII - 88
-------
Session VIII - Environmental Monitoring Applications
Aircraft And Satellite Monitoring Of Lake Superior
Pollution Sources
James P. Scherz and John F. Van Domelen VIII - 1
LIDAR Polarimeter Measurements Of Water
Pollution
John W. Rouse, Jr VIII - 28
Detection Of Dissolved Oxygen In Water Through
Remote Sensing Techniques
Arthur W. Dybdahl VIII - 48
A Search For Environmental Problems Of The Future
James E. Flinn, Arthur A. Levin and
James R. Hibbs VIII - 67
Session IX - Environmental Monitoring Requirements
Sensor Utilization In New England
Region I - Helen McCammon IX - 1
Aerial Monitoring Experience
Region II - F. Patrick Nixon IX - 6
Region Ill's R&D Representative Report
Region III - Edward H. Cohen IX - 10
Region IV Environmental Monitoring Requirements
Region IV- Edmond P. Lomasney IX- 15
Environmental Monitoring Needs In Region V
Region V - Clifford Risley, Jr IX - 18
Remote Monitoring In Region VI
Region VI - Ray Lozano IX- 21
Comments On Environmental Monitoring In Region VII
And Some Monitoring Plans And Needs Prepared For
Second Conference On Environmental Sensors
Region VII - Aleck Alexander IX - 23
Remarks
Region VIII - Russell W. Fitch IX - 26
Remarks
Region IX - Douglas Longwell IX - 29
-------
Session IX (Continued)
Region X Environmental Monitoring Requirements
And Applications
Region X - Ralph R. Bauer IX- 32
Scope of Research Needs
OEGC-NFIC-Denver - Arthur W. Dybdahl IX - 35
_Session X - Comments
Review Of The EPA Remote Sensing Activity
Arthur G. Anderson X - 1
Appendixes
Appendix A - Attendees A - 1
Appendix B - Photograph B - 1
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
OFFICE OF
RESEARCH AND DEVELOPMENT
Preliminary Report of the Second Environmental
Quality Sensor Conference
Chairman, Second Environmental Quality
Sensor Conference
Assistant Administrator for
Research and Development
SUBJECT:
FROM:
TO:
THRU: Acting Director /9/
Equipment and Techniques Division
Deputy Assistant Administrator foy
Monitoring Systems
The Second Environmental Quality Sensor Conference was
convened in Las Vegas on October 10-11, 1973. The purpose of
the conference, as stated by Mr. Willis B. Foster, was to
provide a forum for discussing the role of remote sensing in
the Agency's program to those who have a responsibility for
monitoring. He invited all of those present to work cooper-
atively towards applying those methods and techniques which
could be effectively utilized by the Environmental Protection
Agency (EPA) , and hold the potential of becoming standard
methods in monitoring the environment. Participation of
approximately 130 individuals included representatives from
each EPA Region, National Environmental Research Center (NERC) ,
and from several program offices. Key representatives from
other Federal agencies, universities and industry were also
present.
The conference was organized into ten sessions which in-
cluded 25 papers. A summary of the technical sessions is
included in the attached. The conference proceedings
are to be published and disseminated to all participants.
Copies of the proceedings will be available to the general
public through the Government Printing Office about January 1974,
Prior to the conclusion of the two day conference, regional
and program office presentations focused on how remote sensing
could be effectively utilized in their respective geographical
areas. Numerous applications and program requirements were
-------
suggested by the speakers, some of which are currently
underway and being coordinated by the Office of Monitoring
Systems and NERC-Las Vegas. A few of the more urgent
regional applications for monitoring by remote sensors
are listed:
1. Salinity control in agriculture
2. Water reuse in the West
3. Brown cloud over Denver
4. Agricultural runoff
5. Eutrophication
6. Strip mining
7. Pest infestation
8. Industrial outfalls
9. Everglades protection
10. Power plant thermal plumes.
11. Oil spills
12. Acid mine drainage
13. Ocean dumping
14. Solid waste management
15. SOo from copper smelters
16. Sedimentation and algae productivity
From the above, it appears that the Regions and the
program offices are prepared to utilize remote and automated
in-situ technology in numerous applications.
In response to the Conference Committee invitation,
Dr. Virginia Lee Prentice, National Academy of Sciences (NAS),
Committee on Remote Sensing Programs for Earth Resource
Surveys (CORSPERS) stated that the NAS-CORSPERS Environ-
mental Measurements Panel would prepare a formal response
to EPA on the Academy's recommendations for future research
and development for environmental quality sensors.
Dr. Prentice suggested increased use of data from the
National Aeronautics and Space Administration's Earth
Resources Survey Program. She specifically encouraged
the utilization of high resolution absorption spectroscopy
for measuring atmospheric constituents to a few parts per
billion and the use of the micrometer band for water
temperature measurements.
It is felt that the conference achieved the objective
in communicating new monitoring technology to the Regions,
NERC's and program offices. Through the participation of
the other Federal agencies we have stimulated some technology
exchange which can be applied directly to EPA monitoring
problems. Some large gaps existing between present technology
capabilities and needs of the Agency are being identified.
11
-------
During the conference the submission of the Environmental
Research Need statements pertaining to remote and automated
in-situ requirements was emphasized. A more personal relation-
ship was encouraged in the preparation of the statements to
assure that the resultant plans are responsive to the program
requirements.
Because of the EPA involvement with other Federal agencies,
more formal recognition of interagency activities and communi-
cations relevant to remote sensing should be established. It
is recognized that there are many informal contacts and exchanges
between EPA NERC's and Regions that are not recognized at EPA
Headquarters. This could lead to uncoordinated activities
which are inconsistent with strategies designed to meet the
requirements of program offices. Formal recognition and
guidance by the appropriate organizations within the Office
of Research and Development will lead to a more useful sensor
and/or data product and the timely exchange of research data
as required by the Agency.
The lack of interdisciplinary coordination among remote
sensor specialists in the Agency prompts the need for an intra-
agency remote sensing coordinating and advisory committee in
order to more fully utilize remote sensor technology. Such a
committee was suggested by many of the Agency participants.
The committee should consist of members' from all Regions, NERC's
and program offices working with remote sensors, to form a
working advisory group within EPA. Some of the functions
suggested for the committee are listed.
1. Intra-agency coordination of remote sensing
efforts and liaison responsibilities within EPA and at
interagency levels.
2. Identification, development and prioritization
of both short-term and long-term support requirements/
objectives which are amendable to a remote sensing approach.
3. Definition, planning and recommendation of a
realistic environmental remote sensing program to meet program
office requirements.
4. Review of existing and planned EPA remote sensing
programs assessing their adequacy toward meeting program office
requirements and objectives.
Additional functions and scope of this committee will be
formulated and forwarded to the appropriate offices for review,
comment and approval.
111
-------
In closing, it is felt that notable advances were made
at this conference in implementing the remote and automated
sensor technology toward the Agency's monitoring requirements.
It was the general concensus that remote sensing will not
replace many present methods of manual sampling but will be
used to point out specific locations where environmental
problems exist, so that in-situ methods may be more effectively
and efficiently employed. In this regard, I believe the
conference was a success and more than paid for the time,
effort and funds involved.
ohn D. Koutsandreas
Attachment
IV
-------
SUMMARY OF TECHNICAL SESSIONS
SECOND ENVIRONMENTAL QUALITY SENSOR CONFERENCE
I. National Environmental Research Center-Las Vegas Programs
The aerial remote sensing system being developed was
described, including aircraft imagery and data acquisition
systems, as well as ground data processing systems. A LIDAR
sensor being developed for aerial monitoring of aerosol
characteristics and mixing layer measurements was demonstrated.
The Zeeman effect atomic absorption spectrometer is being
evaluated for direct analysis for mercury in the parts-per-
billion range. This instrument was designed and built for
the National Science Foundation and the Atomic Energy
Commission by the Lawrence Berkeley Laboratory. The
application of helicopter-borne instrumentation for data
gathering in the Los Angeles Reactive Pollutant Program,
is providing valuable information.
•*••*• • National Aeronautics and Space Administration-Langley
Research Center Sensor Programs
Work performed under the Environmental Protection Agency
(EPA)/National Aeronautics and Space Administration interagency
agreements for the evaluation of remote sensors for water
pollution detection and monitoring was presented. The results
of laboratory tests, fixed height platform field tests, and
helicopter flight tests of a four frequency laser system
induces fluorescence in algae for remote monitoring. Measure-
ment results are repeatable and comparison with "ground truth"
data raise questions about which is a more accurate measure of
algae quantity. An imaging multispectral scanner was flight
tested and the data analyzed for the identification of various
pollutants. A technique based on film and filters for
quantitatively measuring the amount of chlorophyll in a water
body was discussed. Understanding what physical processes
are involved in sensing a pollutant, both from the sensor and
pollutant viewpoint, are paramount in this research. The status
of work underway on investigations dealing with passive microwave
radiometers, multispectral scanners and sewage outfall detection
was presented.
v
-------
III. Air Quality Sensor Developments
.The siting of sensor installations responsive to
programmatic needs was discussed for the St. Louis
Regional Air Pollution Study. Utilization of the
latest accepted air monitoring equipment was discussed.
The use of laser technology for long path measurement
of atmospheric pollutants is considered a promising
candidate for monitoring carbon monoxide and ethylene.
Optical techniques are being effectively utilized for
monitoring stationary sources. Instrument research
specifications for air emission controls were presented.
IV* Water Quality Sensor Development
The basic theory and applications of recent water
quality contact sensors were presented. The measurement
of total cyanides by selective ion electrodes appears
practical with some pretreatment. Automated in-situ
sensors are being utilized over 200 river miles by the
Minneapolis-St. Paul Metropolitan Sewer Board. Extended
measurements (beyond the basic property parameters) are
unique, requiring selectivity in sample preparation to
optimize sensor performance. The phase-out of systems
designed for continuous monitoring was due to improper
maintenance, poor selection of sensors, or unavailability
of preferred sensors.
V. Environmental Thematic Mapping
Land use is a key factor in determining environmental
quality in most parts of the Nation. Regular monitoring
of change in quality and use of land and water resources
is required. Thermal mapping data can provide information
on the location of sources of potential pollution from
landfills, from chemical and petroleum storage facilities,
and from sewage lagoons and feedlot operations. Techniques
of acquiring remote quantitative measurements of surface
water temperature with an infrared scanner in an aircraft
are proven. Computer produced images having gray levels
corresponding to absolute temperature (within 1/2°) and
numerical grids giving absolute temperatures at uniform
levels are shown. Remote sensor imaging can assist in
the location and characterization of pollution sources,
assessment of impact, projection of trends, location of
potential problem areas, and the assessment of preventive
measures. Recommendations were presented for an operational
system which provides data during suboptimum aerial
photographic conditions.
VI
-------
VI. Oil and Hazardous Materials Sensors
A detection and identification of specific petroleum
products on water surfaces and volumetric determination
of oil spills through remote sensing were shown to be
not only practical but also adaptable operationally for
surveillance, oil spill cleanup support, and enforcement
purposes. The use of environmental photo interpretation
keys for identification and evaluation of oil and hazardous
materials production, storage, and processing facilities
has significant potential as a spill prevention surveillance
function.
VI1' Satellite Environmental Monitoring Applications
A methodical approach, using aircraft and satellite
data, was presented for locating and monitoring polluted
mine drainage. Such factors as texture, shape, and spectral
response are analyzed and compared to previously obtained
data. Computer processing techniques have been developed
for identifying low, medium, and highly reflective types
of water which are associated with various concentrations
of suspended matter. Turbidity patterns, acid disposal
plumes and convergent water boundaries along with high
concentrations of pollutants have been detected from
satellite imagery. Space data relay systems provide for
automatic collection of data from in-situ environmental
sensors. These networks are supplying data to regulatory
agencies on a daily basis.
VIII. Environmental Monitoring Applications
Through careful preflight planning, in-depth analysis
and coordination with limited ground truth parties it is
possible to utilize aerial photography, both vertical and
oblique to monitor water quality and prepare evidence for
court trials. Quantitative data on water quality can be
obtained from remote sensor data. Some ground truth data
are preferred in order to quantify the density levels of
the imagery. From this point it is a simple matter to
establish correlation between gray-level and selected
water quality parameters. The present state-of-the-art
in remote sensing can produce useful data for monitoring
and enforcement actions for EPA. The use of the data in
operational programs can also bring about advancement of
the state-of-the-art by recognizing gaps in current
procedures.
vix
-------
ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH & DEVELOPMENT
OFFICE OF MONITORING SYSTEMS
SECOND CONFERENCE ON ENVIRONMENTAL QUALITY SENSORS
NATIONAL ENVIRONMENTAL RESEARCH CENTER
LAS VEGAS
OCTOBER 10 AND 11, 1973
— CONFERENCE AGENDA —
Tuesday - October 9
4:00 p.m. -8:30 p.m. - Registration: Royal Las Vegas
Hotel Lobby
Wednesday - October 10
8:00 a.m. - 8:30 a.m. - Late Registration - NERC-Las Vegas
Administration Building
8:30 a.m. - 9:00 a.m. - Conference Convenes
Mr. John D. Koutsandreas - Chairman
Dr. Delbert S. Barth - Welcome
Mr. Willis B. Foster - Keynote Address
Mr. John D. Koutsandreas - Conference Program Overview
I. 9:00 a.m. - 10:10 a.m. - NERC-Las Vegas Programs
Mr. Donald T. Wruble, NERC-Las Vegas
Imagery Acquisition System - Messrs. Leslie M. Dunn and
Albert E. Pressman, NERC-Las Vegas
Aerial Air Pollution Sensing Techniques - Mr. Roy B. Evans,
NERC-Las Vegas
LIDAR for Remote Monitoring - Dr. S. Harvey Melfi,
NERC-Las Vegas
Evaluation of the Zeeman Atomic Absorption Spectrometer -
Mr. Erich Bretthauer, NERC-Las Vegas
10:10 a.m. - 10:30 a.m. - Coffee Break
xx
-------
II. 10:30 a.m. - 11:40 a.m. - NASA-Lanqlev Sensor Programs
Mr. James L. Raper, NASA-Langley
Multi-Wavelength Laser Induced Fluorescence of Algae In
Vivo: - A New Remote Sensing Technique - Mr. P. B. Mumola,
Mr. Olin Jarrett, Jr. and Mr. C. A. Brown, Jr.
. Remote Detection of Water Pollution With Multichannel
Ocean Color Scanner: - An Imaging Multispectral Scanner -
Mr. Gary W. Grew
Summary of NASA-Langley Remote Sensing Activities Under
The EPA Interagency Agreements: - Potential For
Regional Applications - Mr. James L. Raper/ NASA-Langley
The Use Of Near-Infrared Photography For Biodegradable
Pollution Monitoring Of Tidal Rivers - Mr. Walter E.
Bressett, NASA-Langley and Dr. Donald E. Lear, Jr.,
EPA Annapolis Field Office
11:40 a.m. - 1:15 p.m. - Lunch
III. 1:15 p.m. - 2:10 p.m. - Air Quality Sensor Developments
Mr. Charles E. Brunot, Office of Monitoring Systems, OR&D
The St. Louis Regional Air Pollution Monitoring System -
Mr. James Reagan, NERC-RTP
Long Path Optical Measurement of Atmospheric Pollutants -
Mr. A. E. O'Keeffe, NERC-RTP
Performance Specifications for Stack Monitoring Systems -
Mr. James B. Homolya, NERC-RTP
IV. 2:10 p.m. - 3:20 p.m. - Water Quality Sensor Development
Mr. A. F. Mentink, NERC-Cincinnati
In-Situ Sensor Systems for Water Quality Measurement -
Dr. K. H. Mancy, University of Michigan
. Description of Selected Available Sensors -
Dr. Julian Andelman, University of Pittsburgh
Metropolitan Sewer Board Water Quality Monitoring Program
for the Minneapolis-St. Paul Metropolitan Area -
Dr. Russell Susag, St. Paul Metropolitan Sewer Board
Water Quality Sensing in Some Eastern European Countries -
Dr. Peter Krenkel, Vanderbilt University and
Dr. Vladimir Novotny, 'Marquette University
An Airborne Laser Fluorosensor for the Detection of
Oil on Water - Dr. H. H. Kim, NASA-Wallops Station
3:20 p.m. - 3:40 p.m. - Coffee Break
x
-------
V. 3:40 p.m. - 5:05 p.m. - Environmental Thematic Mapping
Mr. John D. Koutsandreas, Office of Monitoring Systems, OR&D
Remote Sensing and Identification of Critical Area of
Environmental Concern - Dr. James R. Anderson, USGS
Remote Sensing Data, A Basis for Monitoring System
Design - Mr. James U. Taranik, Iowa Geological Survey
Recent Development in Remote Thermal Mapping of Water
Surfaces - Mr. E. L. Tilton III, Mississippi Test Facility
Remote Sensor Imagery Analysis For Environmental Impact
Assessment-- Messrs. C. P. Weatherspoon, J. N. Rinker,
R. E. Frost and T. E. Eastler, U.S. Army Engineers,
Topographic Laboratories
Aerial Spill Prevention Surveillance During Sub-Optimum
Weather - Messrs. Robin I. Welch, Allan D. Marmelstein
and Paul M. Maughan, ERTSAT
5:05 p.m. - 5:35 p.m. - NERC Las Vegas Tours
Laser Demonstration - Dr. S. Harvey Melfi
Imagery Processing and Interpretation Lab Tour -
Mr. Albert E. Pressman and Mr. Leslie M. Dunn
Thursday - October 11
VI. 8:30 a.m. - 9:55 a.m. - Oil and Hazardous Materials Sensors
Mr. Donald R. Jones, Division of Oil & Hazardous Materials, OAWP
Single Wave Length Fluorescence Excitation for On-Site
Oil Spill Detection - Mr. J. Jadamec, U.S. Coast Guard
The Remote Detection and Identification of Surface Oil
Spills - Mr. Herbert R. Gram, Spectrogram Corporation
Environmental Keys for Oil and Hazardous Materials
Detection - Messrs. Charles L. Rudder and Charles J.
Reinheimer, McDonnell Aircraft Company
Video Detection of Oil Spills - Mr. John P. Millard,
NASA-Ames and Lieutenant Commander Gerald F. Woolever,
U.S. Coast Guard
. Measurements of the Distribution and Volume of Sea-Surface
Oil Spills Using Multifrequency Microwave Radiometry -
Dr. James P. Hollinger and Mr. R. A. Mennella,
Naval Research Laboratory
The National Environmental Monitoring System -
Mr. Theodore Major, The Magnavox Company
9:55 a.m. - 10:15 a.m. - Coffee Break
XI
-------
VII. 10:15 a.m. - 12:00 Noon - Satellite Environmental Monitoring
Applications
Mr. John D. Koutsandreas, Office of-Monitoring Systems, OR&D
Skylab Application to Environmental Monitoring -
Dr. Victor S. Whitehead, NASA Johnson Space Center
Environmental Applications of the Earth Resources
Technology Satellite, D. T. Schultz, General Electric
Co., Space Division, Charles C. Schnetzler,
Goddard Space Flight Center
ERTS-1 Detection of Acid Mine Drainage Sources -
Messrs. Elliott D. DeGraff and Edward Berard,
Ambionics, Inc..
Water Quality Analysis Using ERTS-A Data -
Dr. Harry Kritikos, University of Pennsylvania and
Mr. Hubert Smith, EPA, Region III
Application of ERTS-1 to Coastal Environmental Problems -
Dr. V. Klemas, University of Delaware
Data Collection Platforms for Environmental Monitoring -
Mr. J. Earle Painter, NASA-Goddard
12:00 Noon - 1:30 p.m. Lunch
VIII. 1:30 p.m. - 2:40 p.m. - Environmental Monitoring Applications
Mr. Robert F. Holmest Office of Monitoring Systems, OR&D
Aircraft Monitoring of Lake Superior Pollution Sources -
Dr. J. P. Sherz and Mr. J. F. Van Domelen,
University of Wisconsin
LIDAR Polarimeter Measurements of Water Pollution -
Dr. J. W. Rouse, Jr., Texas A&M University
Detection of Dissolved Oxygen in Water Through
Remote Sensing Techniques -
Mr. Arthur W. Dybdahl, OEGC, National Field
Investigation Center-Denver
A Search for Environmental Problems of the Future -
Messrs. James E. Flynn, Arthur A. Levin, Battelle
Memorial Institute and Dr. James R. Hibbs, EPA, OR&D
2:40 p.m. - 3:00 p.m. Coffee Break
XII
-------
IX.
X.
3:00 p.m. - 4:45 p.m. - Environmental Monitoring Requirements
Mr. C. E. James, Office of Monitoring Systems, OR&D
Plans, Programs and Requirements
Region I - Dr
Region II - Mr
Region III - Mr
Region IV - Mr
Region V - Mr
Region VI - Mr
Region VII - Mr
Region VIII - Mr
Region IX
Region X
NFIC- Denver
Helen McCammon, OR&D
Francis P. Nixon, S&A Division
Edward Cohen, OR&D
Edmond T. Lomasney, OR&D
Clifford Risley, Jr., OR&D
Ray Lozano, S&A Division
Alex Alexander., OR&D
Russell W. Fitch, OR&D
Dr. Douglas Longwell, S&A Division
Mr. Ralph R. Bauer, S&D Division
Mr. Arthur W. Dybdahl, OEGC, Denver
4:45 p.m. - 5:15 p.m. - Comments
Mr. John D. Koutsandreas, Office of Monitoring Systems, OR&D
National Academy of Sciences - CORSPERS, Environmental
Measurement Overview and Assessments, Panel Conference
Comments - Dr. Virginia L. 'Printice
Closing Remarks - Mr. John D. Koutsandreas
5:15 p.m. - 5:45 p.m. - NERC-Las Vegas Tours
Laser Demonstration - Dr. S. Harvey Melfi
Imagery Processing and Interpretation Lab Tour -
Mr. Albert E. Pressman and Mr. Leslie M. Dunn
Xlli
-------
CONFERENCE KEY NOTE ADDRESS
BY
WILLIS B. FOSTER
DEPUTY ASSISTANT ADMINISTRATOR FOR
MONITORING SYSTEMS
ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
xv
-------
It has been about two years since I last addressed many of you
in this same auditorium. At that time, we endeavored to gain a better
understanding of how sensing technology, both remote and in-situ,
could be directed toward meeting the needs of the EPA Regional Offices,
the State environmental agencies, and the environmental agencies'
municipalities. Although the resources have been insufficient to allow
us to realize many of our objectives, through the efforts of many here in
this auditorium, -we have been able to a degree to demonstrate the
potential of this valuable technology. It is anticipated that through
demonstration and operational application, remote sensors will continue
to play an increasingly significant role in the monitoring program of our
agency-- as well as at the State and local level.
It is encouraging to learn that recently the program offices
within EPA are increasing the application of remote sensor data for
their requirements. One specific example, worthy of mention, is the
Reserve Mining case where remote sensor data were successfully
introduced by both sides and accepted by the court as evidence. Here
the Office of Enforcement and General Counsel acquired imagery from
high altitude aircraft and the ERTS satellite. These data were
effectively utilized in a court exhibit to show the sediment outfall from
Reserve Mining operation, and its transport across the west end of
Lake Superior. I understand you will hear more of this effort during
the conference.
xvi i
-------
Other Federal agencies have also been helpful in promoting
remote sensor technology in the Environmental Protection Agency, as
is evidenced by the scientists and engineers from other agencies on the
agenda and in the audience today. However, an all-government team
is not enough. The universities and industry must also play an important
part in research and development. It will be refreshing and a privilege
to hear some o£ their viewpoints during this conference.
Although research and development is essential and is being
promoted by this Agency, we are often under the gun to show immediate
results and operational applications. This stems from the regulatory
nature of the Agency where the primary emphasis is on enforcement of
environmental quality standards. We often take action using techniques
that are neither thoroughly proven nor standardized. Here again we
look to the universities, industry, and other Federal agencies for
sensor techniques under development which might be applied to our
needs. Thus, the primary theme for our conference this year is one
of "applications." I understand that in the next two days the speakers
will share with us those methods and techniques which could be
effectively utilized right now, and which are potential standard
methods.
The promulgation of standards and regulations is an important
aspect of the EPA program. It has been identified as a specific
management objective for FY 1974 and, therefore, a major focus
xviii
-------
of the Administrator's personal attention. The magnitude of our task
in monitoring the nation's environmental quality is tremendous -- it's
going to require something more than the proven techniques. To meet
our requirements to enforce standards and regulations, we must use
new remote monitoring instrumentation. There are a total of 135 known
standards, regulations and regional reports to Congress which the
Agency is currently developing. We are pursuing automated in-situ or
remote sensing techniques to supplement the older, accepted, time-
consuming laboratory techniques.
Over 32, 000 major air pollution point sources must be brought
into compliance with final emission limitations as prescribed in approved
state implementation plans. Achievement of this objective will require
a vigorous program of air quality monitoring and enforcement to insure
performance by uncooperative polluters. Here again, remote sensing
is expected to play a prominent role.
Under authority of the new environmental laws, EPA has a
mandate to encourage the modification of certain kinds of land use which
aggravate air and water pollution. The nation needs a comprehensive
land use act that will embrace the land aspects of all environmental
problems -- air and water pollution, noise abatement, waste disposal,
management of toxic substances, outdoor recreation, urban planning,
population dispersal and control of population growth itself. Unlike
air and water pollution, the results of land misuse are often irreversible;
we must live with it for generations and in some cases forever. Overhead
xix
-------
monitoring can provide the regions and states with an accurate inventory
of their land assets so that they are in a better position to assure that
growth does not destroy cultural, historical and aesthetic values or
unique ecological systems.
Through permits, we are responsible for controlling the discharge
of pollutants from point sources into the nation's waterways, under the
Federal Water Pollution Control Act of 1972. This requires regulating
40, 000 of the nation's 300, 000 industrial water users and nearly 13, 000
municipal sewage treatment plants. The periodic monitoring and
reporting of discharges calls for efficient, accurate and timely moni-
toring -- automated in-situ and remote monitoring technology may offer
a practicable sDilution. But that's enough for pollution as such.
With less than six percent of the world's population, the United
States consumed one-third of the globe's energy production in 1972.
Our energy demand is projected to more than double by the year 2,000.
In the meantime, the United States is approaching the end of its supply
of cheap fossil fuel reserves, with little prospect of new power sources
for the next fifteen years. As you are aware, the crisis was caused by
several complex and interacting factors, including increased environ-
mental consciousness, which forced cutbacks in the use of certain
high pollution fuels and stymied efforts to construct a number of nuclear
power plants. You are all well aware of the delays resulting from fear
xx
-------
of damage to the environment resulting from construction of a pipeline
carrying 170° oil across tundra and permafrost regions. Energy is a
vital component of environmental rehabilitation as well as America's
prosperity. The problem is whether reasonable energy demands can
be met without harming the environment. Among the EPA energy
objectives are included the promotion of efficiency and conservation;
and in the production of energy -- we're responsible for environmental
assessments of the entire energy chain -- extraction, processing, trans-
portation and utilization. Remote sensing can and should play an
instrumental role in monitoring this entire energy chain.
Successful nations are those that are able to respond to challenge
and to change when circumstances and opportunities require change.
The demands on our natural resources have grown commensurate with
our growth in population, industrial capacity and increased prosperity.
All these indices reflect the constant upward thrust in the American
standard of living. But, in order to maintain our standard of living
and maintain a pleasant, clean environment, a high degree of technolog-
ical sophistication will be required.
We look to you scientists and engineers gathered at this Second
Environmental Quality Sensor Conference to help provide some of the
advanced technology for monitoring the quality of the environment and to
provide the capability for assuring its protection and enhancement -- I
wish you success in this meeting.
xxi
-------
SESSION 1
NERC-LAS VEGAS mOGRAMS
CHAIRMAN
MR. DONALD T. WRUBLE
NERC-LAS VEGAS
-------
INTEGRATED REMOTE SENSING SYSTEM
by
Leslie M. Dunn
and
Albert E. Pressman
Monitoring Operations Laboratory
Imagery Acquisition and Interpretation Branch
NATIONAL ENVIRONMENTAL RESEARCH CENTER
U.S. ENVIRONMENTAL PROTECTION AGENCY
LAS VEGAS, NEVADA 89114
I - 1 NERC-LAS VEGAS
-------
5322
CONTENTS
Figures 1-3
Abstract I->4
Sections
I Introiduction 1-5
II Airborne Data Acquisition 1—17
III Data Processing 1-19
IV Aircraft 1-23
V Appendices -O
± _ 2 NERC-LAS VEGAS
-------
FIGURES
No. Page
1 Quick Response Operational Remote Sensing System 1-12
2 Santa Monica Plume 1-13
3 Canyons entering Santa Monica Bay 1-14
4 Construction activity in canyon entering into I*. 15
Santa Monica Bay
5 Outfall discharges in Monterey Bay 1-16
6 Airborne Data Acquisition System 1-18
7 Ground Data Interpretation Station 1-20
8 Contour Plotting System for IR Scanner Data 1-22
_ 3 NERC-LAS VEGAS
-------
222
INTEGRATED REMOTE SENSING SYSTEM
Leslie M. Dunn
and
Albert E. Pressman
Monitoring Operations Laboratory
National Environmental Research Center
U. S, Environmental Protection Agency
P.O. Box 15027
Las Vegas, Nevada 89114
ABSTRACT
An aerial remote sensing system is being implemented to help meet
the research, surveillance and enforcement requirements of the various
offices of the U.S. Environmental Protection Agency (EPA). An integral
part of the system is a computerized data acquisition system which
allows for partial automatic data processing of aerial thermal mapping
and multispectral measurements. Simultaneous photographic coverage is
time-related to radiometric measurement for detailed analysis, A video
system which records analog data on magnetic tape provides the near
real-time interpretation necessary to respond to emergency situations.
This integrated remote sensing system is an attempt to meet specific
monitoring requirements and provide the EPA with a fully operational
capability. Anticipated use of this data collection system to meet
specific EPA objectives in oil and hazardous material spill emergencies,
outfall detection and inventory, runoff models development for agricul-
tural pesticides, and other airborne monitoring investigations is
discussed. Types of data collected and final results of typical projects
are shown.
1-4 NERC-LAS VEGAS
-------
SECTION I
INTRODUCTION
The Environmental Protection Agency (EPA) was established by the
President as a regulatory agency to safeguard our nation's air, water,
and land resources for present and futur$ generations. In order for
the EPA to accomplish this task, it must establish a comprehensive
and effective program for monitoring the environment. EPA has the
dual responsibility for developing techniques for collecting envi-
ronmental quality data and the utilization of that data for planning
and regulatory purposes. In accomplishing these objectives, EPA
utilizes new developments from other agencies such as NASA.
The National Environmental Research Center in Las Vegas, Nevada,
(NERC-Las Vegas) has been charged with the responsibility for develop-
ing applicable monitoring techniques and programs and providing demon-
stration studies to assess their effectiveness. To meet this
responsibility the Monitoring Operations Laboratory, Imagery
Acquisition and Interpretation Branch, is developing an integrated
system for the collection, processing, interpretation and reporting
of remote sensing data. While remote sensing techniques have limita-
tions, the synoptic nature of these measurements and the ability to
provide needed information rapidly over large areas with a minimum
number of ground monitoring stations can have great value in augment-
ing Regional monitoring programs.
1-5 NERC-LAS VEGAS
-------
55SZ
Aircraft, sensors, processing and reduction equipment, and data
analysis methods have been selected to best serve a wide range of
environmental problem solving tasks. The appendices provide detailed
information on major equipment and instrumentation which comprise the
integrated remote sensing systems. Listed below are five areas we are
pursuing that address the needs submitted by the EPA regional offices,
Office of Enforcement and General Counsel, and the Division of Oil and
Hazardous Materials.
1. Outfall Detection, Location, and Analysis
This program will develop a catalogue of water outfall character-
istics (relative to aerial monitoring) for the more significant
categories of outfalls recommended by all EPA Regions. Based on
this first phase, optimum aerial remote sensing techniques
from existing state-of-the-art capabilities at NERC-Las Vegas
will be field tested at selected sites throughout the country.
Data will be analyzed and a manual prepared showing ground and
airborne coverage of the outfalls. Emphasis will be placed on
identification of unique characteristics of outfalls with ex-
isting techniques and defining future research and developmental
needs in this area. Table 1 is a list of outfalls which were
submitted for detailed study by the EPA regional offices.
2. Oil Spill Damage Assessment and Documentation
This program involves the management of aerial surveillance
responses to major incidents of oil or hazardous materials
spills. Six contractors have entered into basic ordering
I _ 6 NERC-LAS VEGAS
-------
32Z
Table 1. Outfall Categories for Investigation by Remote Sensing
STANDARD INDUSTRIAL CLASSIFICATION (SIC) SIC #
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Feed Lots
Metal Mining
Bituminous Coal and Lignite Mining
Crude Petroleum and Natural Gas (Extraction)
Beet Sugar
Pulp Mills
Paper Mills
Chemical and Allied Products
Industrial Inorganic Chemicals
Plastics Materials, Synthetic Resins and Nonvulcanizable
Elastomers
Petroleum Refining
Blast Furnaces (incl. coke ovens) Steel Works and
Rolling Mills
Primary Metal Industries and Fabricated Metal Products
Telephone and Telegraph Apparatus
Electric Services
Municipal Outfalls
Ocean Dumping
0211
1011
1211
1311
2063
2421
2600
2800
2810
2821
2911
3312
3352
3661
4910
4952
4953
1-7 NERC-LAS VEGAS
-------
agreements (BOA) with NERC-Las Vegas to provide remote sensing
services for emergency spills too distant for a timely response
from NERC-Las Vegas. Figure 1 describes the organization
interactions and information channels for a BOA activation.
Subsequent efforts will include the investigation of systems de-
signed for increased resolution and accuracy and the implemen-
tation of better methods of aerial surveillance reaction to
emergency spills.
1-8 NERC-LAS VEGAS
-------
S3S
3, Agricultural Chemical Run-Off Model
Efforts in the remote sensing portion of this program will be to
provide certain quantitative terrain parameters which can be
inserted into mathematical models of river basins. The parameters
of immediate study will be identification of crop type, delineation
of land cover, drainage patterns, surface radipmetric temperature,
topographic slope, soil type, and conservation practices. Follow-
ing initial flights over small instrumented test plots, a large
scale survey of an entire river basin may be undertaken.
4-- Non-Point Source Water Pollution Monitoring Approaches and Techniques
Identification and description of the most effective approaches
for characterization of each category of non-point source pollution
will be accomplished through intensive literature review and
contact with Agency personnel engaged in assessing such sources.
A report describing applicable non-point source water pollution
monitoring approaches and techniques, (both contact and remote)
based on the current state-of-the-art, will be generated. Anti-
cipated coordination with the Pollutant Fate Research Program at
the Southeast Environmental Research Laboratory will aid in the
selection of procedures to calibrate and validate mathematical
models for estimating pollution loads from non-point sources.
5. Technical Assistance Program
This program consists of relatively short response projects to
meet immediate Regional needs. In practice, NERC-Las Vegas is
1-9 NERC-LAS VEGAS
-------
requested to collect and analyze aerial remote sensor data to be
used by the Region in support of an operational objective.
Examples of these include outfall inventories and documentation
of spillage from waste pits. Results are delivered to the Regions
in the form of annotated data displays, raw data with location
indexes, isothermal contour maps, or final interpretation reports.
The goal of NERC-Las Vegas under this program is to provide Regions
with useful information within 45 days of notification; 15 days to
collect the data and 30 days for analysis and reporting. We are
getting closer to accomplishing this goal as the program obtains the
necessary manpower and other resources.
Figures 2 through 5 illustrate a typical set of display boards;
these were prepared for Region IX. Figure 2 is a photomosaic
showing a sediment plume offshore at Santa Monica, California.
Figure 3 was made in an attempt to track the sediment inland
to its source. Figure 4 spotlights the construction activity
which was believed to be the cause of this sediment entering the
Bay. Figure 5 is an oblique view facing south along the California
coastline at Moss Landing approximately 25 miles south of Santa Cruz.
Shown are two facilities of interest to EPA in our routine surveil-
lance activities under the national discharge permit program. Two
power plants are located at 'B1 and a facility engaged in the ex-
traction of magnesium from the sea is shown at 'A1. In connection
with this program stack emissions such as that evident at (B), and
outfalls into streams seen at (3) and into bays (2) are monitored and
I - 10
NERC-LAS VEGAS
-------
documented. Number (1) is believed to be the surface manifesta-
tion of a newly buried outfall extending offshore. It will be
investigated further.
NERC-LAS VEGAS
-------
FIGURE 1
QUICK RESPONSE OPERATIONAL REMOTE SENSING SYSTEM
Request
AERIAL AND/OR
SURFACE DATA
ACQUISITIO;
OK
vV
KjjB
HEAR REAL
TIME ASSESS-
MENT. VISUAL
OBSERVATIONS,
VIDEO TV, ETC.
PROCESS, REDUCE
ENHANCE, ETC.
CURSORY INTERP.
COMPREHENSIVE
INTERPRETATION,
ANALYSIS,
CORRELATION
FINAL REPORT
GENERATION
Report
NERC-LV i j Briefmg/Keporti7
OPERATIONS'
uoorcn nation
Report
Reoort
DIRECT SUPPORT
OF CLEAN-UP
DAMAGE ASSESS.
EXTENT OF
POLLUTANT
LEGAL RESPON-
SIBILITY, ETC.
*Division of Oil and Hazardous Materials
1-12
-------
SANTA MONICA CANYON
SANTA MONICA BAY.
DATE: 8/8/73
TIME:~1330 HRS
SANTA MONICA. CALIFORNIA
I NERC-LV
120O 60O O 12OO
SCALE IN FEET
24OO
Figure 2 - Santa Monica i'lume
-------
H
I
CONSTRUCTION ACTIVITY
CONSTRUCTION ACTIVITY
FACING NORTHEAST
DATE: 8/8/73
TIME:~1330 MRS
SANTA MONICA CALIFORNIA
NCRC-LV
Figure 3 - Canyons entering Santa Monica Bay
-------
I
M
01
SANTA MONICA
CONSTRUCTION ACTIVITY
FACING SOUTHEAST
DATE: 8/8/73
TIMci~1330 HRS
SANTA MONICA CALIFORNIA
NCRC-LV
Figure 4 - Construction activity in canvon entering
Santa Monica bay.
-------
•«?;ZZJ5>i> #
^~-*»-.£^ ,, r. f
'^••*Ktm
DATE 7-4-73
TIME -1300 HRS
A) KAISER REFRACTORIES
B) PACIFIC GAS & ELECTRIC
\) OUTFALLS
•
..SS2.
NtRC-LV.
Figure 5 - Outfall discharges in Monterey Bay
-------
SECTION II
AIRBORNE DATA ACQUISITION
The basic design for our data acquisition system is oriented
toward operational remote sensing techniques development. Sensors
selected for incorporation into our data acquisition system consist
primarily of off-the-shelf instrumentation that requires only inter-
facing to produce an operationally integrated system. This system,
shown in Figure 6, is designed to minimize the manual data processing
required to produce final results. This has been accomplished by
cross correlating all data with a common time base; even the non-
digital instrumentation such as the camera systems are annotated with
a digital time code so that the photographs can easily be cross
referenced with information collected by other sensors.
An integral part of the airborne data acquisition system is a
computer which formats all flight parameters for automatic data
processing. Time, position, heading, ground speed, drift, roll,
pitch, airspeed, and altitude are digitally recorded and multi-
plexed on 7-track computer-compatible tape. Camera exposure pulses
and spectrometer and IR radiometer(PRT-5) signals are also recorded
on the 7-track tape. The IR scanner data are recorded on an analog
14-track wide-band tape recorder. All analog information collected
is displayed in the aircraft in real time on a strip chart recorder
to insure that all sensors are functioning properly. The strip
chart also serves as an aid in the screening of data for initial
processing.
i _ 17 NERC-LAS VEGAS
-------
Figure 6
I
M
CD
AIRBORNE DATA ACQUISITION SYSTEM
FLIGHT PARAMETERS
POSITION
HEADING
GROUND SPEED
DRIFT
ROLL & PITCH
AIRSPEED
ALTITUDE
SENSORS
CAMERAS
IR SCANNER
SPECTROMETER
IR RADIOMETER
HOUSEKEEPING
TIME
DATE
LABEL
EVENTS
MUX
S/D
&
A/D
SEQ
&
COMPUTER
CONTROL
ANNOT
S7CC
MAGNETIC
TAPE
/O 14 TRACK
, ANALOG
"QUICK
LOOK"
-------
532
SECTION III
DATA PROCESSING
After the information is collected and recorded, the system is
designed for preprocessing at a ground data interpretation station
which consists of a 14-track wide-band tape playback recorder, a high
speed analog-to-digital converter, a computer, a 5-inch film recorder,
an image enhancement system, a scan converter, and a 9-track computer-
compatible tape recorder. Figure 7 is a schematic diagram of the
ground interpretation station.
The IR scanner output recorded on the 14-track tape is digitized
and stored in the buffer memory of the computer. A computer program
then thins the data by predetermined programming into elementary data
points. The information is then transferred to a 9-track computer-
compatible tape.
For applications requiring high spatial resolution, such as
detection of small outfalls, the thermal infrared imagery is recorded
on 5-inch black and white film from the original 14-track tape re-
cording. No data thinning is undertaken. The scan conversion system
allows the operator to view the tape on a TV monitor and select those
portions which contain anomalies of interest. An additional analysis
option is to display thermal levels on the infrared imagery as discrete
colors for quantification of temperature zone's in thermal plumes. The
7- and 9-track tapes then go to the central computer facility where the
data are processed in the GDC 6400. From the information on the 9-track
tape an atmospheric effects table is generated which corrects all data
1-19
NERC-LAS VEGAS
-------
Figure 7
GROUND DATA INTERPRETATION STATION
14 TRACK
TIME
H
I
to
o
VIDEO \
SYNC/'
ANALOG
TAPE
A/D &
CONTROL
I S !
LEVEL
SLICER
SEQUENCER
&
COMPUTER
CONTROL
9 TRACK
COMPUTER
TAPE
5 INCH
FILM
RCDR
_„ _.„..,__..,/ r~ 1 1 fi K f
1 FILM /
TAPE TO TAPE/TAPE TO FILM
-------
points to ground datum. The IR scanner has been d.c. restored and
equipped with detectors having increased sensitivities and black
body reference sources (so that quantitative temperature measurements
can be made). We hope to achieve an overall system accuracy of about
1/2° C after atmospheric effects are considered. Because of the scan
angle geometry, it is also necessary to make a rectilinear correction
of each scan line. The geometrically and radiometrically corrected
data are then plotted as an isothermal contour map on a 25-inch drum
plotter. A schematic of the data collecting, processing, and contour
plotting system is illustrated in Figure 8.
1-21
NERC-LAS VEGAS
-------
Figure 8
CONTOUR PLOTTING SYSTEM FOR I R SCANNER DATA
I
(O
B-26 AIRCRAFT | GROUND STATION
DOPPLER TIME
RADAR COD
1 1
NOVA 1200
COMPUTER
/7-TR
«{
V
DATA
CALCU
PROC
/DA
ACK\
"\
j: PAOAM
<
POINT /
1 ATION •» Cl
SRAM \
77NT(x,y)
f POINT )
V TAPE / CA
X^ I p
i
NUMERICAL
APPROX.
PROGRAM
i
E I
E f
X~^Xv(t) S' Nv^-
/A \ A/D NOVA 840 [ 9-TRACK |
S~^^<\ ]_ CONVERTER COMPUTER \^ TAPE J^
(l R ) ^ _] R h|te ^ ~]
\ 1 | O Ullb
^/SCANNER\ ! [
__.. o._~ " — — - — — — - — --- — - — — - — — — — — — - — — — ---- — -i— 1
COMPUTER PROCESSING
ETERS | | [^PARAMETERS
, T T
^ "X /- — \ ATMOSPHERfC
UC 640U\ TEMPERATURE / LuSo ^ LQSS TAB|_E
">K^niiTrnl» — - pnnnrpfinN ~* TARI r «
DMPUTCRI- CORRLCTION 1 TABLC 1- GENERATION
TADC / PROORAM \ TAPF / •» .
mPb^ I HOUHAM N^AI h Z^ PROGRAM
1 fpARAMFTFRS
1 AHI 1-
UKII) L
XULATION
ROGRAM - — (^PARAMETERS
/* ^ Tfx,y)
, / oKID \ pi O-TTITQ I r^HMTni IP /
f rii p i . A \
,1 MLL ) rnNTnnq|Nr; *\ PLO! \
\ TAPE Z_ / /
-------
SEZ
SECTION IV
AIRCRAFT
A Monarch B-26 is used by NERC-Las Vegas to fly the missions
previously described. This aircraft^ in our estimation, represents
one of the best trade-offs of financial and operational factors
that was commercially available. This aircraft satisfied the require-
ment for access to all equipment in flight to make equipment adjust-
ments, change film and filters, and monitor acquisition of all data
at a central console. The aircraft also has the speed and altitude
capabilities which are dictated by our need to respond to emergency
oil spills for the Division of Oil and Hazardous Material. Additional
characteristics and specifications of the B-26 are found in Appendix A.
1-23 NERC-LAS VEGAS
-------
322
SECTION V
APPENDICES
Page
A. Aircraft 21
B. Data Collection System 22
C. Data Processing 26
D. Photographic Processing 27
1-24 NERC-LAS VEGAS
-------
53S
APPENDIX A
AIRCRAFT
B-26 SPECIFICATIONS
Wing Span
Length
Height
Maximum Gross Weight
Empty Weight
Landing Weight
Engines - two
Fuel
Cruise Speed
Landing Speed
Altitude
VFR Range (with reserve)
Flight Crew
Sensor Operators
Navigation Equipment
Communicat ions Equipment
Control Equipment
70 feet
54 feet
18 feet
34,000 pounds
21,139 pounds
31,000 pounds
R-2800-71 ZOOO HP
80.0 gallons
240 knots
100 knots
Up to 25,000 feet
1200 nm
2
2
Includes DOPPLER, VOR, ADF, DME
Includes VHP, UHF, and ADF
Includes Flight Director System
and Autopilot
1-25
NERC-LAS VEGAS
-------
APPENDIX B
DATA COLLECTION SYSTEM
1. Infrared Scanner (RS-310)
Field of View (Degrees)
Resolution (milliradians)
V/H Range (Radians/Second)
Scanner RPM
Noise Equivalent Temperature,
Large Target (Degrees C Refer-
enced to 300 K Background)
Wavelength Region (Microns)
2
Collecting Aperture (inches )
Number of Units in System
Size and Weights of Units
Scanner/Recorder (Unit 1)
Control Panel
(Unit 2)
90
1.5
.03 to 0.3
3000 (200 Scans/Sec)
0.20 (InSb)
0.12 (GeHg)
.3 - 14 (Optics Capability)
6.35
5
16.50" L x 14.50" W x 15.00" H
55 Lbs. (without magazine
5.?5" L x 5,32" W x 3.00" H
2.7 Ibs.
Power Supply
Film Magazine
(Unit 3)
(Unit 4)
Compressor (Unit 5)
(when closed cycle cooling)
13.00" L x 8.50" W x 3-75" H
11.4 Ibs.
6.40" L x 5,925" W x 8.45" H
9.0 Ibs (without film)
10.50" L x 8.25" W x 10.50" H
19.0 Ibs.
1-26
NERC-LAS VEGAS
-------
APPENDIX B (Coat.)
Charac teri stic
Ambient Temperature Limitations
Altitude Limitations
Input Power Requirements:
AC Voltage
AC Power
DC Voltage
DC Power
Noise Equivalent Temperature (NET)
8-14 pm
1.0 mrad
3.0 mrad
10.0 mrad
3 - 5 pm
1.5 mrad
3.0 mrad
Number of Dectectors
Types of Dectectors
8 - 14 pm
3-5 pm
Detector Cooling
Stabilization
Recording Format
Specification
0° - 100° F
0 - 10,000 feet
117 Vac, 400 Hz, 3 Phase
45 VA
28 Vdc
135 Watts Run
160 Watts Start
0.15° C
0.07° c
0.02° C
0.20° C
0.10° C
Mercury-Cadmium Telluride (HgCdTe)
Indium Antimonide (InSb)
Open-cycle system (Dewar) operating
at 77° K
Roll 4_10
2.30 in.
1-27
NERC-LAS VEGAS
-------
APPENDIX B (Gont.)
2. Aerial Camera (KA-76)
Primary Use
Film Information
Format (in.)
Width (in.)
Length (ft.)
Frames (per roll)
Magazine
FMG Rate (ips)
Optics
Focal Length (in.)
Max Apt
Angular Coverage ( )
Resolution (L/mm) (TC 1000:1)
Shutter Speed
Shutter Type
Remarks
Day/Night Reconnaissance
4.5 x 4.5
5
250/100
600/250
Cassette
0.15 - 10.8
1.75
5.6
105°
4.5
, o
218
,o
1.5
o
74" 41" 41
12
3-5
21°
35 pan-X
40 plus-X
25 IR
40 pan-X
1/60 - 1/3000
Focal Plane
Automatic Exposure Control
3. Scanning Spectrometer
Component
Rotating Filter Wheel
Filters
Visible Region (4000-7000)A°
Nominal Half-band width
Center Wavelengths (3400-7000)AC
Center Wavelengths (7000-9800)AC
Function
Selectable Discrete Bandpass Analysis
3400 - 9800 A°
Circular variable
250 A°
@ 300 A° intervals
400 A intervals
'I - 28
NERC-LAS VEGAS
-------
APPENDIX B (Cont.)
Component
Detector, Silicor Photodiode,
Extended UV Coverage
Telescope, 3-inch, f-3 Newtonian
Control Electronics
4. TV System
Component
RCA SIC Camera PK 501
Shibaden Vidicon Camera HV-155
Video Monitor RCA PM 9C
Video Monitor Shibaden 9 inch
Video Tape Recorder Shibaden SV-510U
Special Effects Gen SE-101S
SYNC Generator SG 105-L
TV Zoom Lens V5 x 20
Function
Photovoltaic
Reflective, 2° f.o.v.
SYNC and Output Selection
Function
Flight Path Surveillance
Display Monitor Camera
Equipment Operator Monitor
Pilot Monitor
IR Video Recording
Mix 2 Camera Video
Video SYNC
Zoom Capability
1-29
NERC-LAS VEGAS
-------
APPENDIX C
DATA PROCESSING
Major Equipment
Data General Nova 840
Computer, L6K Memory
Phoenix 8BIT A to D Converter
Sangam 14 Channel Analog
Record/Reproduce and Audio
Shibaden Video Tape Record/
Playback
ISI Model VP-8 Image Analyzer
Function
Processing and Control
Digitize IR Scanner Video Signal
IRIG Intermediate and Wide Band
Group II Record and Reproduce
and Annotation.
Quick Look Viewing and Data
Screening and Selection
Assigned Color Imaging System
and Level Slicing to Represent
IR Imagery Thermal Levels
1-30
NERC-LAS VEGAS
-------
APPENDIX D
PHOTOGRAPHIC PROCESSING
Log-Etronics MK II R5B
Pako Processor 33
Kodak Versamat 1411 RT
Log-Etronic SP 10/70 B
EN 6A2
Log-Etronic Enlarger E-10
Robinson Copy Camera
Miller Holzworth EN 46A Enlarger
Step and repeat contact printer
with two times enlarging capability,
equipped with automatic exposure
and even tone printing.
Modified to 3 step chemistry for
processing color prints up to 16"
wide at 72" per minute.
Quick change kit was installed for
processing both color positive and
color negatives.
Continous contact strip printing with
automatic exposure and even tone print-
ing. Has capability of printing -60
f.p.m.
Prints color and black/white film
in rolls from 700 mm to 9%" width
(mostly used for color film trans-
parency) .
Enlarger with automatic exposure
control and even tone printing.
Scaled reproduction of 10,600"
accuracy, for interpretation,
photomosaic, and reports.
Fixed ratio enlarger for enlarging
70 mm film four times.
1-31
NERC-LAS VEGAS
-------
AERIAL AIR POLLUTION SENSING TECHNIQUES
R. B. Evans
Monitoring Operations Laboratory
National Environmental Research Center
U.S. Environmental Protection Agency
P. 0. Box 15027
Las Vegas, Nevada 89114
ABSTRACT
A capability for aerial monitoring of air pollutants has been
developed at the National Environmental Research Center at Las Vegas
(NERC-LV). A comprehensive monitoring system incorporating measurement
instrumentation for NO, NOX, 03, CO, non-methane hydrocarbons, sulfur
gases, particle size, light scattering, temperature, dewpoint, and
fluorescent particle tracer concentration has been developed and installed
in a twin-engine helicopter for measurement of ambient levels over
metropolitan Los Angeles. The system records its data on magnetic tape.
Available instrumentation suitable for aerial monitoring was surveyed,
and components of the system were selected on the basis of seven criteria:
resistance to vibration and stress, stability under variations of temperature
and altitude, sensitivity, fast response time, low power consumption, size
and weight, in order of decreasing importance. Commercially available
analyzers have evolved sufficiently that satisfactory instruments for most
of the listed parameters are now available. System components were tested
for stability under varying altitude and temperature in an environmental
chamber. Various navigational aids were evaluated as possible components
of the system.
Mathematical techniques to compensate for instrument response time
are available. The convolution integral technique appears to be the
simplest in application.
1-32
-------
AERIAL AIR POLLUTION SENSING TECHNIQUES
By Roy B. Evans
The Los Angeles Reactive Pollutant Program (LARPP), currently in
progress, and the Regional Air Pollution Program (RAPS) both
include important roles for airborne contact sensing of air
pollutants. The LARPP is designed to compile a data base suitable
for developing and validating mathematical models of air pollution
photochemistry. The RAPS has the more ambitious goal of developing
regional models which will deal with both photochemistry and dis-
persion of pollutants.
The LARPP is concerned with photochemical reactions within a parcel
of air as it moves downwind. Air parcels are isolated and followed
to measure pollutant concentrations within the parcels (NO, N02,
CO., 03, non-methane hydrocarbons, and particulates) from both air-
borne and ground-based sampling platforms. The LARPP aircraft are
also measuring temperature, dewpoint and fluorescent particle con-
centrations. Ultra-violet radiation is being measured by ground-
based stations throughout the Los Angeles basin. The product of
this effort is a time history of vertical profiles of each of the
pollutants being measured within the parcel.
The aerial monitoring requirements for the RAPS have not yet been
defined in detail, but vertical profiles and horizontal cross-
sections at relatively low altitudes above the RAPS St. Louis
ground station network will be necessary. Sulfur gases will be of
interest in St. Louis, in addition to those parameters presently
being measured by the LARPP aircraft.
MONITORING PLATFORMS
The type of aircraft selected for monitoring platforms depends on
several factors: the geographical area over which monitoring is to
be performed, the required payload, electrical power required for
monitoring instrumentation, the necessity for a crew to operate
instrumentation, aircraft range and the desired monitoring airspeed.
The limiting consideration for both the LARPP and RAPS programs is
the requirement to operate at low altitudes over urban areas. In
the case of the LARPP, the air parcels to be sampled are identified
with clusters of three tetroons, ballasted to attain a nominal
altitude of about half of the inversion height (typically, a nominal
ballast altitude of 500 feet). The sampling vehicle then executes
square patterns about the centroid of the tetroon cluster at altitudes
1-33
-------
ranging from 200 feet up to the base of the inversion. The RAPS
program is also expected to require operation at altitudes as low
as 200 feet over urban St. Louis. Safety considerations and
Federal Aviation Administration (FAA) restrictions essentially
dictate the use of either twin-engine helicopters or some kind of
lighter-than-air vehicle.
The instrument payload in the LARPP and SAPS aircraft amounts to
approximately 1500 pounds, and the electrical power load is
approximately 4 kilowatts. Two instrument men and a navigator are
required in addition to the pilot.
A flight range of at least two hours is required. Both helicopters
and lighter-than-air vehicles were investigated initially, but the
lighter-than-air vehicles were quickly discarded. The payload of
instruments and crew exceeds the capacity of available blimps, and
blimps possess limited maneuverability. Operation of blimps at
altitudes as low as 200 feet appears to be of questionable safety,
and rental rates for blimps appear to be on the order of $2000/day.
There are two types of twin-engine helicopters available in the
United States which satisfy the LARPP and RAPS requirements: The
Bell 212 (military designation UH-1N) and the Sikorsky 58 (military
H-34) modified to twin-turbine capability (58T). Both are capable
of sustained flight with one engine, and both can carry useful loads
of more than 2500 pounds. Both can provide up to 10 kilowatts of
electrical power. The Bell 212 has a useful range of approximately
1 hour and 45 minutes, while the twin-turbine Sikorsky 58T has a
range of approximately 2 hours and 15 minues. Rental rates for the
Bell 212 are on the order of $20,000 per month plus $200 per hour,
including fuel and pilot. Comparable rates presently quoted for the
Sikorsky 58T are $14,000 per month and $150 per hour. The Sikorsky
58T has greater cabin space and is probably a more convenient
flying laboratory. Two leased Bell 212 helicopters are presently
in use in the LARPP, and their performance has been satisfactory,
although the present instrumentation load uses nearly all of the
available space.
INSTRUMENTATION
The most important considerations in selecting instruments for aerial
monitoring of air pollutants at the present state of instrument
development are stability under flight operating conditions and response
time. Power requirements and weight are of less importance because
suitable sampling platforms with more than enough electrical power
and payload are available. The total altitude range in sampling for
the LARPP is approximately 5000 feet (sea level to 5000 feet MSL),
1-34
-------
and the same range is expected to be adequate in St. Louis. Ambient
temperatures at which the instruments are to function range from
0°F. to 120°F. Vibrational stress in the helicopter environment
is severe, although no measurements of stress were made in the Bell
212's presently in use. The constant sampling airspeed used in the
LARPP is 60 knots, and an aircraft operating at this speed covers
approximately one-fifth of a mile in 10 seconds. This means that
an instrument having a 10-second lag time and a 10-second response
time to 90% of full-scale will require a total of approximately
0.4 miles to respond to a step-function increase in concentration.
Table 1 lists the instrumentation selected for the LARPP/RAPS
helicopters. All of these instruments, with the exception of the
sulfur gas analyzer and the particle size analyzer, are presently
in use in the LARPP.
Chemiluminescence analyzers were selected for measurement of NO
and NOX. These instruments depend on the Chemiluminescence reaction
of NO with ozone :
NO + 03 >• N02 + 02 + hv
The emitted light is measured with a thermo-electrically-cooled
photoraultiplier tube. Total oxides of nitrogen (NOX) are measured
by first converting N02 to NO by passing the sample gas over heated
(450°C.) molybdenum. N02 is obtained by subtraction: (NOX) - (NO).
There are several chemiluminescent NO-NOjj analyzers which are
commercially available, but the TECO 14B (Thermo-Electron Corporation,
Waltham, MA) was selected because the instrument uses a low-pressure
(300-torr) reaction chamber and the sample flow is controlled with
a critical orifice, making the instrument relatively altitude-
insensitive.
Ozone is measured with an instrument which is dependent upon the
chemiluminescent reaction of ozone with ethylene. Again, the light
emitted by the reaction is measured with a photomultiplier tube.
Of the several ozone analyzers commercially available, the REM
Model 612 was selected because it utilizes critical orifice flow
control and is stable under varying altitudes. The instrument
also differs from others in utilizing electronic temperature drift
compensation rather than thermo-electric cooling of the phototube.
Of the several non-dispersive infrared analyzers available, only the
Andros is small enough to be conveniently used in the aircraft as a
CO monitor. This instrument utilizes dual-isotope fluorescence and
has better sensitivity and stability than conventional NDIR. It is
subject to significant zero drift while in flight, however. The
1-35
-------
Mine Safety Appliances (MSA) Model 11-2 non-methane hydrocarbon
analyzer utilizes two parallel flame ionization detectors (FID),
one of which is preceded by a catalytic srubber to remove hydro-
carbons heavier than methane. One FID thus measures total hydro-
carbons, the other methane, and electronic subtraction gives the
difference. There was relatively little choice in the selection
of this instrument, even though the two in the LARPP helicopters
are essentially prototypes. This measurement technique is the
only one available which yields continuous measurements of non-
methane hydrocarbons, and the MSA instrument is the only one
currently packaged in a form suitable for field use. The alter-
native was the use of a chromatograph-type instrument, such as the
Bendix 8201., which yields measurements of grab samples collected
about three minutes apart.
DATA RECORDING
Because of the number of parameter inputs and the total data volume
expected for both the LARPP and the RAPS., recording of data either
by hand or with strip-chart recorders would be too laborious to
be depended upon for the data archive. A digital data acquisition
system which records the data inputs directly in standard BCD
code on magnetic tape was incorporated into the system. Both
manual observations and strip charts are taken, but these serve
primarily for debugging and quick-look purposes. A Monitor Labs
Model 7200 Data Acquisition System was selected to record directly
onto a Cipher Model 70 Magnetic Tape Deck. These components have
been generally satisfactory^ but the tape deck is sensitive to
overheating and will not function at temperatures above 105°F.
ENVIRONMENTAL CHAMBER TESTING
One instrument of each type was tested in an environmental chamber
to determine its response to varying temperatures and altitudes.
Instrument span and zero drift were measured as functions of
temperature and altitude (barometric pressure). The data obtained
in these tests must receive further verification before publication,
but the instruments selected seem to perform well under varying
ambient conditions. Span drifts are less than 10% over the
operating ranges of temperature and altitude. Zero drifts as
functions of altitude are negligible, and zero drifts as functions
of temperature depend on the instrument type: the 03 instrument
appears to experience a zero shift of as much as 2 pphm out of
200 pphm full-scale; the NO and NO exhibit drifts of as much as
0.1 ppm at the upper end of the ambient temperature range, around
95°F ambient. At the same ambient temperature, CO drift of about
1 to 2 ppm out of 20 ppm full-scale is experienced.
1-36
-------
CALIBRATION
NO, NOX, 0
-------
#
Minimum J C
Heasureaer.t Instiument Instrument Detectable Lag Time & M
Parameter Method Kanufacturer/Hod«l Ranges Concentration Response Tlmn
Particles > . 3 ftn
Visibility
K>
Nu
X
CO
so2
ar.d
TV.
T-»
few Point
Altitude
Individual
Particle
Countet
light scattering
Cheni luminescent
Chemilumine scent
Chcni luminescent
ND1R
/Fluorescence)
Flame
Photometric
no
Therwoc lac trie
Thermoelectric
Pressure
Kee Industrles/ll
Royco/220
KRI/1550B
Thermo Electron
Corp./l4»
Thermo Electron
Corp./UB
DEH/6128
Andros 7000
Meloy SA 160
HSA U-2
Cambridge
Systems Mdl
137-Cl-SIA-TII
Cambridge Systems
Hdl 137-CI-SlA-Tli
Computer
Instrument Corp.
Hdl 8000
Continuous,
I, 2. 5, and
10 second
intervals
10 channels
Three Ranges
. 05 , , 1 , . 25 ,
5.0, 10 ppm
full scale
. 05 , . 1 , . 25 ,
.5, 1.0, 2.5,
5.0, 10 pin
lull scale
U-200 pphm
0-20pphra
0- 2pplua
20,50,100,
200 ppm
lOppm
0-5ppra
0-20ppm
0-30,000 ft
bseat • -1
. 0005 ppm
.0005 ppm
.01 pphm
.Ippm
, OOSppm S0_
SOppb
440' to
20,000'
+0.4Z above
20,000'
250 ma RT per particle
Variable R.T.
.1 - 200 sec. -*
<2 sec
with modifications
(manual mode)
<2 sec
with modifications
(manual mode)
-------
LIDAR FOR REMOTE MONITORING
S. H. Melfi
U.S. Environmental Protection Agency
National Environmental Research Center, Las Vegas
P. 0. Box 15027
Las Vegas, Nevada 89114
ABSTRACT
In the past the Environmental Protection Agency (EPA) has underempha-
sized remote sensing in its overall monitoring program. However, remote
monitoring can play a valuable role in providing a new perspective in sensing
the environment. As an adjunct to contact sensing, remote monitoring pro-
vides synoptic information over a wide geographical area which is important
in determining pollution dispersion and predicting pollution levels and epi-
sodes. Recently, increased emphasis is being placed on remote sensing of the
environment by EPA through its National Environmental Research Center located
at Las Vegas, Nevada (NEUC-LV). This paper will discuss tue resaatxh pj.ogiavn
being initiated at NERC-LV. The areas for which remote monitoring research
will be conducted include air, water and possibly terrestrial pollution,
The remote or; it or Ing technique to be discussed in this paper is LIMR,
an acronym for Light Detection And Ranging. LIDAR is similar to Radar in
that it provides a range resolved measurement of scattered electro-magnetic
radiation, but utilizes a pulsed laser as the source and an optical telescope
as the receiver. LIDAR has applications for the remote monitoring of aerosol
characteristics, sensing the height of the mixing layer and remote measure-
ment of visibility. These applications will be discussed in detail along
with a description of the mobile LIDAH unit presently under construction.
1-39
-------
LIDAR FOR REMOTE MONITORING
S. H. Melfi
Monitoring Systems Research and Development Laboratory
National Environmental Research Center
U.S. Environmental Protection Agency
P. 0. Box 15027
Las Vegas, Nevada 89114
In the past the U.S. Environmental Protection Agency (EPA) has
underemphasized remote sensing in its overall monitoring program. How-
ever, remote monitoring can play a valuable role in providing a new per-
spective in sensing the environment. As an adjunct to contact sensing,
remote monitoring provides synoptic information over a wide geographical
area which is important in determining pollution dispersion and predicting
both pollution levels and episodes. Recently, increased emphasis is be-
ing placed on remote sensing of the environment by EPA through its National
Environmental Research Center located at Las Vegas, Nevada (NERC-LV). The
remote monitoring technique to be emphasized in this paper is LIDAR, an
acronym for Light Detection And Ranging. LIDAR is similar to Radar in
that it provides a range-resolved measurement of scattered electro-magnetic
radiation, but utilizes a pulsed laser as the source and an optical tele-
scope as the receiver. LIDAR has applications for the remote monitoring
of aerosol characteristics, sensing the height of the mixing layer and re-
mote measurement of visibility.
Before discussing the LIDAR program, it is useful to review the role
remote monitoring will play in the EPA monitoring program. Figure 1 is a
block diagram of the integrated approach for monitoring systems. As is
shown in the figure, an integrated monitoring system concept is developed
addressing a specific monitoring need, and both remote and contact sensors
contribute their unique advantages which result in an integrated system
that provides the solution. The close coupling of the two techniques is
also shown in the figure. Remote monitoring provides information on the
optimum placement of contact sensors whereas the contact sensors provide
1-40
-------
calibration "ground truth" information for the remote sensors. Be-
cause of the wide geographical coverage inherent in remote sensing, the
technique provides input to models and "quick looks" at environmental
quality violations. Contact sensors also provide data for model input
and can be utilized to verify violations detected remotely.
As part of an integrated system, LIDAR is being developed to assess
meteorological and topographical effects on the dispersion of pollutants.
Figure 2 shows an artist's concept of a LIDAR system mounted in an air-
craft. The pulse of laser energy interacts with molecules and aerosols
as it propagates down from the aircraft. Some of the energy is scattered
back toward the telescope. Analysing the signal as a function of time
provides a range-resolved indication of aerosol or particulate concentration.
This range-resolved data is presented to a TV monitor resulting in a two-
dimensional display of the aerosol mixing as shown in the insert of the
figure. A detailed diagram of the LIDAR system is shown in Figure 3.
In addition to the TV display capability, the analog data are digitized
by the ADC and stored on magnetic tape for later detailed analysis.
This is only one technique to provide much needed regional monitoring
data. In general, the advantages of remote monitoring are:
1. Wide geographical coverage
2. Measurement above ground
3. Synoptic information
4. Input to models
5. Cost effective
6. Dispersion of pollutants
7. Placement of contact sensors
8. "Quick look" at violations
Making use of these advantages will insure that remote monitoring will
have a significant role to play in the EPA's program to monitor the qual-
ity of the environment.
1-41
-------
ROLE OF REMOTE MONITORING
NEED
REMOTE
t
|;r.!\son I
PLACEMENT 1
ICAL
i
BRATlON
j
CONTACT
Figure 1
T.V. MONITOR
Figure 2
1-42
-------
AIRBORNE LIDAR SYSTEM
Figure 3
1-43
-------
EVALUATION OF THE ZEEMAN ATOMIC ABSORPTION SPECTROMETER
E. W. Bretthauer
Monitoring Systems Research and Development Laboratory
National Environmental Research Center
U.S. Environmental Protection Agency
P. 0. Box 15027
Las Vegas, Nevada 89114
ABSTRACT
The Lawrence Berkeley Laboratory (LBL), under a grant from the
National Science Foundation and the U.S. Atomic Energy Commission,
has designed and constructed an instrument which may be applicable
for the direct analysis of mercury in the parts-per-billion range
for most types of environmental samples. The U.S. Environmental
Protection Agency's National Environmental Research Center in Las
Vegas, Nevada, was selected to evaluate the instrument for mercury
analysis in a wide variety of environmental media.
The instrument is nearly identical to the usual atomic absorption
spectrometer with one major exception. Two close-lying optical lines
(VL cm apart) are used: one to monitor the atomic mercury vapor
from the host material and the other to monitor the molecular vapor
alone. The difference in optical absorption of these two close-lying
optical lines is proportional to the atomic vapor. The lines are
generated using the so-called Zeeman effect, i.e., the splitting of
an optical line into two close-lying optical lines using a magnetic
field. The instrument is capable of direct analysis of most types
of samples in approximately 30 seconds.
Three instruments have thus far been constructed for mercury
analysis. Efforts are underway at LBL to develop appropriate
modifications so that the instrument can also be used for cadmium
and lead analyses. Preliminary results of EPA's evaluation of one
of the instruments for mercury analysis are discussed.
1-44
-------
PRELIMINARY EVALUATION OF THE ISOTOPE ZEEMAN ATOMIC ABSORPTION SPECTROMETER
By
E. Bretthauer, W. Beckert, and S. Snyder
INTRODUCTION
The NERC-Las Vegas' Monitoring Systems Research and Development
Laboratory is conducting projects concerned with the development and/or
evaluation of instruments for the measurement of mercury in various media.
On request of the Equipment and Techniques Division of the Office of
Monitoring Systems Office of Research and Development, EPA, on April,
1973, the NERC-Las Vegas began a program in late June to evaluate a
newly developed prototype instrument designed for the direct analysis
of mercury in environmental samples. The instrument had been developed
by Dr. T. Hadeishi and others at Lawrence Berkeley Laboratory under
grants from the National Science Foundation and the U.S. Atomic
Energy Commission.
This paper is a report on the progress of the evaluation to date.
The remaining laboratory work is expected to be completed in early
1974, at which time a final report will be promptly initiated.
PRINCIPLE OF OPERATION
In the Isotope-shifted Zeeman Atomic Absorption Spectrometer
(IZAA), the sample is thermally decomposed and.oxidized in a furnace
maintained at a temperature of about 900° C. Under these conditions,
mercury is supposedly stable only in its elemental form. The gaseous
sample is then swept into a heated absorption tube by a stream of
oxygen. Here it is probed with a light beam consisting of two
constituents; one has a wavelength centered on the absorption profile
of natural mercury in air, while the other is slightly displaced
(less than 1 cm ) from the mercury absorption line. Absorption
of the centered constituent in the absorption tube is caused by mercury
vapor as well as by nonmercury decomposition products - smoke, and any
thermally stable molecular species present; the absorption of the dis-
placed constituent is due only to the nonmercury products. In the
1-45
-------
vicinity of the 2537 R mercury line, this background absorption does
not change significantly over 1 cm . Consequently, by taking the
difference in the absorption of the two constituents, one determines
the absorption due to mercury alone.
The heart of the technique lies in the mode in which the probe
and the reference constituents are generated, and the method used to
distinguish them from each other. In this instrument (see Figure 1)
x 204
both constituents are supplied by a single Hg lamp operated in a
15-kilogauss magnetic field. When such a lamp is viewed perpendicular
to the applied magnetic field, the Zeeman effect splits the 2537 A
line into three components: a a component shifted to longer wavelength,
a a component shifted to a shorter wavelength, and an unshifted ir
component.
Any mercury present in the absorption tube consists of the
naturally-occurring mixture of several stable isotopes. Since the
absorption tube is operated at one atmosphere, the absorption lines of
each isotope are pressure-broadened, and shifted slightly towards longer
wavelength. In Figure 2 the resulting total absorption profile due to
naturally-occurring mercury in one atmosphere of N? is plotted; super-
imposed upon this profile is the Zeeman-split emission spectrum of the
204
Hg lamp. Note that the TF component corresponds accurately to the
peak of the absorption profile of natural mercury, while the 0 components
are both well off on the wings of the profile. Consequently we may use
the differential absorption of the TT and a components as a measure of
the quantity of mercury present in the absorption tube (the ir component
becomes the probe beam, and the o components taken together become the
reference beam).
Another feature of the Zeeman effect provides a convenient means of
separating the ir and the a components. At right angles to the magnetic
field, both a components are linearly polarized perpendicular to the
field, whereas the ir component is polarized parallel to the field.
Consequently, either component may be viewed independent of the other
with a properly aligned linear polarizer.
1-46
-------
1.0
CV1
z
c _c
J*
«*-
0 .C
• 2 "o/ve
-^ a 0.5
o
«- 6
o
§2
c
o
a.o
204Hg t~ (15KQ
204Hg
-------
H
l
£>
00
PHOTOMULTIPLIER
INTERFERENCER FILTER
FOR 2537A
LINEAR \ SAMPLE 1 f
POLARIZER PORT \JK
VARIABLE PHASE
RETARDATION PLATE?
LAMP MOUNTED IN 15 KG
f MAGNETIC FIELD
'ABSORPTION
TUBE, 800- 1000'C
AUDIO
AMPLIFIER
POLARIZATION
COMPENSATOR
2
VALVE
Figure 2
-------
In order to optimally utilize the IZAA technique, one must employ some
method of alternately allowing the probe and the reference beams to fall upon
a detector after being transmitted through the absorption region. The variable
phase-retardation plate accomplishes this beam switching; it exploits the
optical properties of fused quartz. When fused quartz is stressed, it becomes
birefringent - that is, light polarized along the stress axis propagates
through the quartz at a different velocity than light polarized perpendicular
to the stress axis. If a plate of fused quartz is oriented so that the stress
axis forms an angle of 45 with the plane of polarization of incoming light,
the birefringence of the quartz introduces a phase shift proportional to the
applied stress between the two perpendicular components of the light. By
appropriately choosing the stress, the quartz can be made to function as a
half-wave plate.
In the present instrument, such a plate is oriented at 45 with
respect to the magnetic field applied to the light source. The quartz
is mounted within a C-frame pulse-transformer core on which is wound a
driver coil. Since the length of the quartz plate is chosen to leave an
air gap of 0.5 mm on one side of the split core, varying the current in the
driver coil varies the stress on the quartz plate; we have, in effect, a
magnetic clamp.
After traversing the variable phase-retardation plate described above,
the light passes through the absorption tube, and falls upon a linear polarizer
f
oriented parallel to the light-source magnetic field. When the current applied
to the magnetic clamp is zero, the polarizer passes only the IT, or mercury
probe component of the light; when the current is adjusted so that the quartz
is a half-wave plate, the quartz rotates the plane of polarization of both the
TT and the 0 components by 90 , so now the polarizer passes only the a,
or reference components.
The light which passes the linear polarizer next encounters an inter-
ference filter which passes all components of the 2537 A line equally well,
but which discards spurious light. After leaving the filter, the light finally
reaches the photomultiplier, where it generates an electrical signal propor-
tional to its intensity. If no mercury is present in the absorption tube, the
probe and reference beams are absorbed and scattered identically by the non-
1-49
-------
mercury background. Hence, as they alternately fall upon the photomultiplier,
the light intensity does not change, and the photomultiplier output signal
remains constant. In the presence of mercury, however, the probe component
will be more strongly absorbed than the reference component, and so the photo-
multiplier output will vary at the audio frequency at which the switching from
one beam to the other takes place. This audio component of the phototube
output is extracted and amplified by a lock-in amplifier.
In practice, in order that the lock-in amplifier output be proportional
to the quantity of mercury present in the absorption tube, two additional
devices are necessary. The first is an amplifier with electronically-
controlled gain following the photomultiplier; its gain is automatically
adjusted to compensate for the attenuation of the transmitted light by the
non-mercury constituents, and for variations in the intensity of the light
source. The second is a quartz polarization-compensator plate - simply a
piece of quartz in the light path oriented at an angle with the light beam.
By varying the angle, one can compensate for accidental differences in the
intensity of the probe and reference components which mimic the presence of
mercury even in the absence of any sample.
EVALUATION
To evaluate the applicability of the IZAA for the determination of sub-
microgram quantities of mercury in various types of environmental samples,
experiments were initiated to explore the following parameters:
1. Effect of sample size on signal for various types of samples.
2. Reproducibility of signal.
3. Effect of speciation on signal with emphasis on those chemical forms
of mercury likely to be encountered in environmental samples.
4. Detection and sensitivity levels.
5. Interference effects (especially those likely to be encountered in
environmental samples).
6. Economics of operation (man-hour/analysis, maintenance requirements,
down-time, etc.).
This paper will report on experiment types 1-4 concerned with aqueous
media only. Laboratory work on experiment types 1-4 concerned with other
sample types as well as experiment'types 5-6 will be reported subsequently.
1-50
-------
EFFECT OF SAMPLE SIZE ON SIGNAL
Any instrument proposed for the direct analysis of mercury in environ-
mental samples should produce a signal that is independent of sample size
over some specified range. An operating range of 40 microliters was considered
adequate for aqueous sample types.
Mercuric chloride standard solutions containing 0.4 to 5 ppm of -mercury
in 0.2N HC1 were prepared and 10-, 20-, and 50-microliter aliquots analyzed with
the IZAA. The results are shown in Table I. Since the standard deviations
of the signal sizes overlap, the signals appear to be independent of the
concentrations in aqueous systems when the sample sizes are 10-50 microliters.
Similar experiments have been initiated using alfalfa, bovine tissue, blood,
benzene, and ethanol samples.
REPRODUCIBILITY OF THE SIGNAL
The reproducibility of a measurement is perhaps the most important factor
in selecting any analytical instrument. TTarious concentrations of mere-uric
chloride standard aqueous solutions were prepared and 10 consecutive analyses
performed at each concentration. The results are shown in Table II. The
relative standard deviations of the signals ranged from 2~7% over a con-
centration range of 0.1-5 ppm. Experiments to determine the reproducibility
over lower concentration ranges for aqueous, alfalfa, bovine tissue, and blood
sample types are in progress.
EFFECT OF SPECIATION ON SIGNAL
Mercury occurs in the environment in a number of chemical forms. Any
instrument designed to quantitate mercury in environmental samples must provide
the same signal regardless of the chemical form of the mercury. To evaluate
the effect of speciation on signal, an inorganic mercury compound (mercuric
chloride) and an organic mercury compound (thimerosal) were used. The primary
reason for choosing thimerosal as a representative for organic mercury
compounds was its relatively low volatility. Aqueous standard solutions
of each of the compounds were prepared containing 8, 16, and 28 ng of
mercury, respectively, per sample. Analyses of the samples using the IZAA
spectrometer gave the results listed in Table III. Since the standard
deviations of the signals for given quantities of mercury overlap, the
values determined by the IZAA method appear to be independent of the
1-51
-------
TABLE I
Effect of Sample Size on Signal
Mercury
Standard
ng/ sample
50
50
40
40
30
30
20
20
20
Sample
Size
microliters
10
50
20
50
20
50
10
20
50
Mercury
Standard
ppm
5.0
1.0
2.0
0.8
1.5
0.6
2.0
1.0
0.4
Signal
Average
of 10 readings
1.870
1.944
1.668
1.520
1.431
1.269
0.879
0.940
0.942
Standard
Deviation
0.062
0.051
0.058
0.032
0.087
0.020
0.039
0.064
0.045
Standard
Error
0.021
0.017
0.016
0.010
0.025
0.006
0.012
0.020
0.014
1-52
-------
TABLE II
Reproducibility of Signal
Mercury
Standard
ng/ sample
50
25
20
15
10
8
6
40
30
20
16
12
8
4
50
40
20
10
5
Sample
Size
Microliters
10
10
10
10
10
10
10
20
20
20
20
20
20
20
50
50
50
50
50
Mercury
Standard
ppm
5.0
2.5
2.0
1.5
1.0
0.8
0.6
2.0
1.5
1.0
0.8
0.6
0.4
0.2
1.0
0.8
0.4
0.2
0.1
Signal
Average
of 10 readings
1.870
0.974
0.879
0.726
0.513
0.449
0.359
1.668
1.431
0.940
0.838
0.713
0.491
0.336
1.944
1.520
0.942
0.606
0.430
Standard
Deviation
0.062
0.049
0.039
0.029
0.015
0.015
0.025
0.058
0.087
0.064
0.025
0.027
0.022
0.027
0.051
0.032
0.045
0.026
0.023
1-53
-------
TABLE III
Effect of Speciation on Signal
Mercury
Standard
ng/sample
Microliters
Sample
Size
Chemical Signal Standard Standard
Form Average of Deviation Error
10 readings
16
28
20
20
40
40
70
70
Mercuric
Chloride
Thimerosal
Mercuric
Chloride
Thimerosal
Mercuric
Chloride
Thimerosal
0.513
0.471
0.766
0.741
1.089
1.081
0.024
0.019
0.022
0.011
0.020
0.028
0.011
0.008
0.010
0.005
0.009
0.013
1-54
-------
chemical form of the mercury.
An experiment where the different chemical forms were combined into
a single sample and analyzed indicated that the shape of the absorption
curve was not always consistent. A set of experiments was therefore initiated
to determine if signals from different chemical species combined in the same
sample were additive.
Aqueous mercuric chloride and thimerosal standard solutions each
containing 16 and 28 ng, respectively, of mercury per sample were analyzed
separately. Then aliquots of the solutions containing the organic and
inorganic forms, respectively, of mercury were combined and analyzed. As
can be seen from the results listed in Table IV, a 20-27% signal increase
over the calculated additive values was observed, when the two different
chemical forms were combined and analyzed.
Preliminary experiments using a similar approach where thimerosal was
replaced by phenylmercuric chloride and methylmercuric chloride, respectively,
indicated a signal decrease as compared to the calculated additive values.
Experiments to determine the background of organic solvents revealed
that most of the organic solvents tested, especially aromatics such as benzene,
toluene, and xylene, showed a positive signal. This indicates the presence
of products in the light path which show a higher absorption in the ir region
than in the a region.
As these data are of extreme importance in ultimately adapting the
instrument to routine environmental analysis, especially for biological
samples, efforts are being made to obtain information on the origin of
this non-linearity of the signal when mixtures of various chemical forms
of mercury are analyzed, and on the influence of organic species on the
observed analytical values.
DETERMINATION OF DETECTION LIMIT AND SENSITIVITY
The detection limit can be defined as the concentration required to
give a signal-to-noise ratio of 2. Using this definition, the minimum
detection level for mercury in an aqueous sample was found to be 0.4 ng.
The sensitivity of mercury using the IZAA (1% absorption) was found to
be 4 ng.
1-55
-------
TABLE IV
Effect of Combining Different
Chemical Forms on Signal
*
1
2
3
4
5
6
7
Chemical
Form
Mercuric
Chloride
Thimerosal
Mercuric
Chloride
Thimerosal
Mercuric
Chloride
Thimerosal
Mercuric
Chloride
Thimerosal
Mercuric
Chloride
Thimerosal
Mercury
Standard
ng/ sample
16
16
28
28
8 )
8 )
8 )
20 )
20 )
8 )
Signal
Average
10 readings
0.766
0.741
1.089
1.081
0.957
1.341
1.306
Calculated %
Signal Difference
0.754 +27%
1.085 +242
1.085 +20%
1-56
-------
CONCLUSION
These preliminary results indicate acceptable reproducibility in the
concentration range 0.1 - 5.0 ppm and a reasonable independence of signal
from concentration (0.4 - 5.0 ppm) and sample size (10-50 yl), for aqueous
solutions. The signal is reasonably independent of the chemical form of
mercury; however, combinations of different chemical forms of mercury in
the same aqueous sample produced signals which were non-additive.
1-57
-------
Q.
h-
cn
CJ
1.0
C
CD
•2 ctQ.5
2 E
o
§2
E
V)
C
o
0.0
204,
204
Hg0-+ (15 KG)
-54.0 -36.0 -18.0
0 18.0 36.0
XBL 731-105A
Fig. 3
1-58
-------
H
Ui
O
-PHOTOMULTIPUER
r-INTERFERENCE FILTER
\_FOR 2537A
t
LINEAR \ SAMPLE
POLARIZER-^ PORT-
VARIABLE PHASE
RETARDATION PLATE
L ABSORPTION
TUBE,800-IOOO*C
v-LAMP MOUNTED IN 15 KG
\ MAGNETIC FIELD
AUDIO AMPLIFIER
^'
POLARIZATION
COMPENSATOR
02
VALVE
XBL 736-857
i. 2
-------
SESSION II
NASA-LANGLEY SENSOR PROGRAMS
CHAIRMAN
MR. JAMES L. RAPER
NASA-LANGLEY
-------
MULTIWAVELENGTH LIDAR FOR REMOTE SENSING OF
CHLOROPHYLL a IN PHYTOPLABKTON
Peter B. Mumola, Olin Jarrett, Jr.,
and Clarence A. Brown, Jr.
NASA Langley Research Center
Hampton, Virginia
Presented at the Second Environmental Quality Sensor Conference
Las Vegas, Nevada
October 10-12, 1975
II - 1
-------
MULTWAVEEENGTH LIDAR FOR REMOTE SENSING OF
CHLOROPHYLL a IN PHYTOPLANKTON
Peter B. Mumola
Olin Jarrett, Jr., and Clarence A. Brown, Jr.
NASA Langley Research Center
Hampton, Virginia, 2366.5
BIOGRAPHICAL SKETCH
Dr. Mumola received his B.S.E.E. from Polytechnic Institute of Brooklyn
in 1965, his M. S.E.E. from New York University in 1966, and his Ph. D.
in E.E. from the University of Texas in 1969. He served in the U.S.
Army, Corps of Engineers, from June 19^9 until May 1971? during which
time he was assigned to NASA's Electronic Laboratory in Boston to con-
duct research on lasers, dye lasers, and associated electronic
equipment.
From May 1971 until September 1973> Dr. Mumola was employed at NASA's
Langley Research Center to continue his work on dye lasers. During
this period, Dr. Mumola developed a four-wavelength dye laser pumped by
a single common flashlamp. Dr. Mumola1s 7 years' experience in lasers
have been in both atmospheric and water, and he has become a leader in
the field of dye lasers.
Presently, Dr. Mumola is employed by the Perkin-Elmer Corporation of
Norwalk, Connecticut.
Mr. Jarrett received his B.S. in Mechanical Engineering (Aero option)
from North Carolina State University in 1962 and his M.S. in Engineer-
ing Mechanics from Virginia Polytechnic Institute and State University
in 1968. Mr. Jarrett has been employed at NASA's Langley Research
Center since 1962. During his employment, Mr. Jarrett has conducted
research in plasma physics and infrared laser systems. Mr. Jarrett is
currently engaged in measuring water quality utilizing a laser system.
Mr. Brown is a Senior Aerospace Engineer with 24 years' experience in
aerospace research. Mr. Brown received his B.S.A.E. from Auburn
University in 1948 and his B.S.M.E. from Auburn University in 1949.
Since 19^9* Mr. Brown has been employed at NASA's Langley Research
Center and has conducted research in stability and control of aircraft
and missiles, reentry physics, meteor simulation, and project engineer-
ing and management. In the past 2 years, Mr. Brown has applied his
research skills and background knowledge to applications of remote
sensing of water using LIDAR systems.
II - 2
-------
MULTIWAVELENGTH LIDAR FOR REMOTE SENSING OF
CHLOROPHYLL a_ IN PHYTOPLANKTON
Peter B. Mumola
Olin Jarrett, Jr., and Clarence A. Brown, Jr.
NASA Langley Research Center
Hampton, Virginia 23665
ABSTRACT
A theoretical and experimental analysis of laser induced fluorescence
for remote detection of chlorophyll a_ in living algae and phytoplankton
is presented. The fluorescent properties of various species of algae
representative of the different color groups are described. Laboratory
measurements of fluorescent scattering cross sections is discussed and
quantitative data presented. A "scattering matrix" model is developed
to demonstrate the essential requirement of multiwavelength laser
excitation in order to make accurate quantitative measurements of
chlorophyll si concentration when more than one color group of algae is
present in the water (the typical case). A practical airborne laser
fluorosensor design is considered and analysis of field data discussed.
Successful operation of the Langley ALOPE (Airborne LIDAR Oceanographic
Probing Experiment) system is described and field measurements pre-
sented. Accurate knowledge of a, the optical attenuation coefficient
of the water, is shown to be essential for quantitative analysis of
chlorophyll a_ concentration. The feasibility of remotely measuring a
by laser radar is discussed.
INTRODUCTION
The application of laser radar (LIDAR) technology to the remote detec-
tion of fluorescent materials, notably oil and chlorophyll a_, in natural
waters has been actively pursued by several groups in the United States
and Canada. Thus a common appreciation for the value of remote sensing
to the oceanographic community and to those responsible for water qual-
ity management is assumed. NASA Langley Research Center has initiated
a program to develop an airborne fluorosensor with a multiwavelength
laser for detection, of chlorophyll a_ in living algae where more than
one color group may be present.
LABORATORY SPECTRAL STUDIES
Since chlorophyll a_ is insoluble in water, this molecule is found in a
host material, namely, algae and phytoplankton. The optical properties
of the host material alter the fluorescence excitation and emission
New address - Perkin-Elmer Corporation, Main Avenue, Norwalk, Connecticut 06856.
L-9215
II - 3
-------
spectra of the chlorophyll a molecule. Therefore, knowledge of the
optical properties of the algae as it is found in nature, rather than
in acetone extract solution, is required for remote detection
application.
During the past year LRC personnel have measured the fluorescence
excitation and emission spectra of over 45 different species of algae
representative of the four major color groups (blue-green, green,
golden-brown, and red). Using Rhodamine B as the fluorescence standard,
the effective fluorescence cross section has been computed as a function
of excitation wavelength for each species. The apparatus used in these
studies is shown in Figure 1. A fluorescence spectrophotometer, Hitachi
Perkin-Elmer model MPF-2A, was modified to improve its red sensitivity.
The spectra were recorded on both a strip chart and magnetic tape, the
latter being used for input to computer programs for cross-section com-
putation. Excitation spectra were measured by setting the emission
monochromator to 685 nm, the chlorophyll a_ fluorescence peak, and scan-
ning the excitation wavelength from 3^0 nm upward through the visible
spectrum. Both monochromators were set to 5 nm slit widths to obtain
usable signal levels without destroying the spectral resolution. Emis-
sion spectra were then recorded by setting the excitation monochromator
to the optimum excitation wavelength (determined above) and scanning
the emission monochromator from that wavelength upward to 800 nm. Simi-
lar spectra were measured using a 10" 7 molar solution of Rhodamane B in
ethanol. Cross sections were then computed using these spectra and
accounting for instrumental effects such as monochromator transmittance
and lamp spectral intensity. Figure 2 shows typical results for algae
of each color group. Note in the emission spectra that each color group
emits strongly at 685 nm due to the presence of chlorophyll a, though
other fluorescent components may also be present. The excitation
spectra differ from one color group to another, each having a unique
region for optimum excitation. Spectra within any given color group
are, however, remarkably similar as shown in Figure 3 for golden-brown
algae. Therefore, one can characterize the fluorescent excitation
properties of any algae by the color group to which it belongs. Note
that these cross sections were computed for single molecules of chloro-
phyll a_ and a spectral resolution of 5 om. This will be important in
the analysis which follows.
It should be noted that no single excitation wavelength can be chosen
to uniformly stimulate chlorophyll a_ fluorescence in all algae. Spec-
tral overlap also precludes selective excitation of any one color
group of algae. One method of "measuring the concentration of chloro-
phyll a can be shown in the LIDAR equation given below.
II - 4
-------
"rec
where
SA
For single fluorescent
scatterer
or
f.(Az)n.
a
For four different
fluorescent scatterers
I =
A =
^D
f
e =
r
Ql =
n =
°f
P =
o
V
a =
optical efficiency of receiver
effective area of telescope primary mirror (m )
= detector bandwidth (nm)
= fluorescence bandwidth (20 nm)
receiver field of view (rad)
laser beam divergence (rad)
density of chlorophyll a (molecules/m )
2
•• effective fluorescence cross section (m )
laser output power (W)
laser wavelength (nm)
attenuation coefficient of water (m~ )
= fluorescent power received (W)
and subscripts f and 1 refer to fluorescence and laser wavelengths,
respectively. If all algae were equally excited at a given wavelength,
then the upper form of the equation (for single Of) would be appro-
priate. As previously shown, algae of each color group possess differ-
ent cross sections and therefore the bottom form of the LIDAR equation
is required. Since the algae color groups have different fluorescence
excitation spectra, the use of four wavelengths yields four equations.
These equations can thus be solved simultaneously to yield the unknown
chlorophyll a concentration contained in each color algae. In matrix
form this can be expressed as
II - 5
-------
p oo
recx 1'
Prec aj for all laser
wavelengths under consideration (if-50 nm - 650 nm), and the optical
window of water decreases with increased wavelength, at least Of must
be known to yield quantitative measurements. In open waters, data are
available indicating that a does not vary rapidly in time or space.
In estuarine and coastal waters changes are more rapid and frequent
measurements of a must be obtained.
II - 6
-------
Assuming a is known or measurable, the above matrix technique can be
used to determine the concentration of chlorophyll a in the algae pres-
ent in a body of water and the distribution of chlorophyll a in the
different color groups.
MJLTIWAVEnENGTH LIDAR INSTRUMENT
Figure k shows a schematic of the airborne LIDAR system which has been
designed and fabricated at Langley Research Center to demonstrate the
multiwavelength concept of chlorophyll a detection. The laser used in
the system is a unique four-color dye laser pumped by a single linear
xenon flashlamp. Figure 5 shows a cross-sectional view of the laser
head. The head consists of four elliptical cylinders spaced 90° apart
with a common focal axis. The linear flashlamp is placed along this
axis and its radiant emission is equally divided and focused into the
dye cuvettes located on the surrounding focal axes. Fluorescent dyes
form the active medium for the four separate dye lasers. A rotating
intracavity shutter permits only one color at a time to "be transmitted
downward to the water.
The resulting fluorescence from the chlorophyll a is collected by a
25.1)--cm-diameter Dall-Kirkham type telescope. The signal is then passed
through a narrow bandpass filter centered at 685 nm and on to the
photomultiplier (PMT) tube. The PMT signal is digitized by a waveform
digitizer and stored on magnetic tape for later analysis. The dyes and
the water for the flashlamp are kept at a uniformly cool temperature by
the refrigerator. The high voltage supply, charging network, coaxial
capacitor, trigger generator, and a spark gap along with a central con-
trol system complete the package. Figure 6 shows the completed system
prior to installation in a helicopter for flight evaluation.
Field tests have been performed to demonstrate the capabilities of this
new technique. Fjcperiments have been conducted from a fixed height
platform (George P. Coleman Bridge, Yorktown, Virginia) 30 meters over
the York River. This site was selected because it was convenient to
both Langley Research Center and the Virginia Institute of Marine
Science (VIMS). Ground truth data (chlorophyll a_ concentration, salin-
ity, and algae species identification) were supplied by VIMS using
standard water sampling techniques. The attenuation coefficient (at
632.8 nm) and temperature of the water were measured on site by Langley
personnel.
Measurements were made every half-hour on the evening of July 9, 1973,
and data are shown in Figure 7 along with ground truth data supplied by
VTMS. Chlorophyll a concentrations shown represent the total chloro-
phyll contribution of all color groups. A bioassay performed by VIMS
indicated a dominance of golden-brown (dinoflagellates) and green algae.
The ratio of golden-brown to green algae varied over the course of these
measurements and was in general agreement with observations obtained
with the LIDAR system.
II - 7
-------
On July 25, 1973, the LIDAR system was successfully flight-tested over
the James River between Hampton Roads and the Chickahominy River.
Flight altitude was 100 meters and the speed was 120 km/hr. The flight
path is shown in Figure 8 along with the chlorophyll a concentration
measured over the 138-kilometer round-trip flight. During each flight
leg the laser was fired at a rate of 0.5 pps. The data plotted in
Figure 8 represent average chlorophyll a concentration over each leg.
For example, leg No. l6 data represent an average of 63 laser firings
of each color or 252 shots total. There is sufficient data, however,
from each four-color cycle to determine chlorophyll a content without
averaging. In fact, the data collected over the entire flight (approx-
imately 75 minutes long) represent nearly 500 separate chlorophyll a_
measurements.
SUMMARY AND CONCLUSIONS
A multiwavelength laser fluorosensor has been developed to remotely
measure chlorophyll a concentration in living algae in natural waters.
Preliminary field operation of the instrument and technique has been
demonstrated from both fixed height and airborne platforms. The maxi-
mum operational altitude of the present system is approximately
300 meters (estimate based on data acquired at 100 meters). Laser
energies varied with color from 0.6 mJ (k^k.k nm) to 7.15 mJ (598.7 nm).
These values are well within the eye safe limits at the operational
altitude of 100 meters. Greater energies could be employed to accom-
modate higher altitude operation. System stability appears to be
excellent as evidenced by the fact that laser alinement has remained
constant for over 6 months.
The major disadvantage of all optical remote sensors of water constitu-
ents is their dependence on foreknowledge of a (or its components "a"
and "s" due to absorption and scattering, respectively) to make quanti-
tative measurements. This is true for the multiwavelength LIDAR tech-
nique as well. Data can only be analyzed quantitatively when a is
known. Studies are now underway to determine the feasibility of remote
a measurements by LIDAR techniques. It may be possible, with slight
instrument modifications, to monitor a simultaneously with the
chlorophyll a concentration.
ACKNOWLEDGMENTS
The authors are grateful for the cooperation of Drs. Franklin Ott and
Robert Jordan of the Virginia Institute of Marine Science for obtaining
laboratory algae sample and ground truth data and to Messrs. L. G. Burney,
C. S. Gilliland, and B. T. McAlexander of NASA Langley Research Center
for their assistance in the design, fabrication, and testing of the
LIDAR flight instrument.
II - 8
-------
EMISSION MONOCHROMATOR EXCITATION MONOCHROMATOR
H
H
I
VD
FLUORESCENCE DATA
RECORDING
oo
FILTER
SAMPLE
CELL
LIGHT
SOURCE
Figure 1.- Laboratory apparatus used in fluorescence excitation and
emission studies.
-------
EXCITATION SPECTRA
EMISSION SPECTRA
1=
o
t:
o
g
h-
.6
4
.2
r-Rhodosorus (Red)
Aphanothece /
(Blue-green)/
—\
i I i t
400 480 560 640 540 580 620 660 700 740 780
EXCITATION mVELEN6Tfi,nm EMISSION WAVELENGTH.nm
o
K.
O
UJ
O
z
UJ
o
(f)
111
QL
O
ID
0L
Eutrepia marina
(Green)
Chaetoceros
(Golden Brown) /
r — i — -I- — r-
400 480 ,560 640 540 580 620 660 700 740 780
EXCITATION \AAVELENGfH, nm EMISSION WAVELENGTH, nm
Figure 2.- Typical fluorescence cross-sections and emission spectra of
red, blue-green, green, and golden brown algae samples.
II - 10
-------
EXCITATION SPECTRA
GOLDEN BROWN ALGAE
CO
o
x
z
o
a
CO
CO
o
tr
o
UJ
o
UJ
o
CO
UJ
cr
o
-
u_
Monochrysis
lutheri
Chaetoceros
Gymnodinium
simplex
Chrysosphaeropsls
planktonicus
Pseudoisochrysis
paradoxa
360 440 520 600 680
EXCITATION WAVELENGTH, nm
Figure 3.- Effective fluorescence cross-sections of chlorophyll a_ in
golden brown algae. Data shown are normalized to single
molecule values with a 5 nm resolution.
- 11
-------
H
H
CHARGING
NETWORK
HIGH VOLTAGE
SUPPLY
REFRIGERATOR
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
TRIGGER
GENERATOR
H20 AND DYE PUMPS
BUBBLE FILTERS
HEAT EXCHANGER
XXXXXXXXXXXXXXXXXXXXXXXX
xxxxxxxxxxxxxxxxxxx
Figure 4.- Schematic of the airborne multi-wavelength LIDAR system.
-------
POLISHED
ELLIPTICAL
CYLINDER
COMMON
LINEAR
FLASHLAMP
DYE CELLS
DYE CELLS
/ 75 ST ALUMINUM'
/ f S S S f /
Figure 5.- Cross-sectional view of the four-color dye laser used for
fluorescence excitation.
II - 13
-------
H
H
Figure 6.- Photograph of complete LIDAR package prior to helicopter
installation.
-------
H
H
1
ro
^v
O»
3 3
ol
o.
o
o
0
"GROUND TRUTH "
ALOPE DATA
1830 1900 1930 2000 2030 2100 2130 2200
9 JULY 1973
Figure 7.- Field data acquired from a fixed height platform over the
York River near Yorktown, Virginia.
-------
7
'14
N
161
1 g
FLIGHT DATE -25 JULY 1973
ALTITUDE- 100 m
AIRCRAFT - BELL 204B
LASER NO. WAVELENGTH(nm)
1 598.7
2 4544
3 5390
617 8
START 1045 EDT
FINISH 1200 EDT
^^V NEWPORT j
>K>^EWS ''/$
¥^5v //«
1 c^ /X3\ / „ ,/ fc
^fes/P^T 1
«A HAMPTON
\ ROADS
m r^
\r^~ *P,
NORFOLK^/
\c=^^
-\TV S ^
xT
^^JW
\ *v~
Nk
c
O
A
G
0
A
O NW FLIGHT
ASE FLIGHT
2 -
1 -
n l2 '"
U 13 '14
10 ,98, 6&7 5 4
15 16 17 18
3 2 1
9 20 21
FLIGHT LEG
Figure 8.- FlighL data acquired over the lower James River
-------
REMOTE DETECTION OF WATER POLLUTION
WITH MOCS: AN IMAGING MULTISPECTRAL SCANNER
Gary W. 'Grew
NASA Langley Research Center
Hampton, Virginia
Presented at the Second Conference on
Environmental Quality Sensors
Las Vegas, Nevada
October 10-11, 1973
II - 17
-------
REMOTE DETECTION OF WATER POLLUTION
WITH MOCS: AW IMAGING MULTISPECTRAL SCANNER
Gary W. Grew
NASA Langley Research Center
Hampton, Virginia 23665
BIOGRAPHICAL SKETCH
Mr. Grev received his B.S. degree in Physics from Northeastern
University in 196l, and has done graduate vork at Virginia Polytechnic
Institute. From 1961 to the present time he has worked at the
Langley Research Center in basic and applied research. This research
includes such topics as: design and development of flight instrumenta-
tion, solid state detection devices, radiation sensors, radioactive
sensing techniques, meteoroid research, and remote sensing. He worked
on a meteoroid experiment on Explorer XXIII and was project engineer
and coinvestigator for the meteoroid experiments on Lunar Orbiter
flights I through V. Mr. Grew is currently investigating the applica-
tion of remote sensing of ocean color to environmental pollution
problems.
ABSTRACT
Aircraft flights to collect spectral data using MOCS (Multichannel
Ocean Color Sensor) are being conducted in an effort to establish
algorithms which correlate with and distinguish between water pollu-
tants, including algae and sediment. Data collected over Clear Lake,
California, New York Bight, and off Cape Hatteras demonstrate the
value of MOCS as a remote sensing tool in studying the hydrosphere.
A spectral signature extracted from the Clear Lake data has identified
the type of algae in the lake. The New York Bight and Cape Hatteras
data reveal a peculiar signature associated with the transition from
blue water to a turbid water mass. This transitional signature may
be useful as a calibration point for remotely determining the concen-
trations of the suspended materials in the surrounding waters.
INTRODUCTION
Remote sensing of ocean color is currently under investigation at the
Langley Research Center (LaRC) with the prime objective of developing
the capability of producing, by means of spacecraft and aircraft
instrumentation, periodic maps which display the distributions of
pollutants, including algae and sediment, in our oceans and inland
waters. This investigation is being conducted with MOCS (Multichannel
Ocean Color Sensor), a unique multispectral scanner designed and
developed under NASA contract by TRW, Inc, as part of the AAFE
(Advanced Applications Flight Experiments) Program. Under this
L-921U
-------
contract MOCS was successfully flight tested on the NASA Convair 990
over various water bodies. A flight was also conducted in the New
York Bight area in a program sponsored by the National Oceanic and
Atmospheric Administration (NOAA). Examples of data from both missions
are presented in this paper. Flights supported by ground truth are
planned for EPA over the Potomac River. Several flights have been
scheduled, but weather conditions were unfavorable.
With MOCS it appears feasible that a user agency can periodically over-
fly a given water body and generate, from the spectral data collected,
maps which identify the suspended materials in the water and which con-
tour their distributions. The feasibility of remotely detecting water
pollutants with MOCS has been demonstrated (Eefs. 1,2). The degree
to which pollutants can be distinguished and quantified by this tech-
nique requires further investigation.
Additional flight data supported by ground truth data are needed not
so much to test instrument capability, but rather to establish optimum
data processing techniques. Before contour maps of water bodies can be
automatically outputted, spectral signatures must be established that
correlate with the various suspended materials. Computer techniques
exist which automatically correlate multispectral data, separate data
points into classes, and generate maps which distinguish land and
water features by color coding. These classification programs can tell
the user nothing about what each color coded feature might be, but only
that each color shades areas that emit similar spectral signatures.
For land maps, the user must identify known features (e.g., wheat
fields) that correlate with each color. His task is diminished by
several factors: (l) Land features are often separated by sharp
boundaries, (2) land features are often unchanging (i.e., deserts)
or change gradually (i.e., vegetation), and (3) classified features
can be verified by ground truth days after the overflight.
The situation is quite different for water masses. Water boundaries
generally are not sharp and changes can occur rapidly. The establish-
ment of an algorithm that correlates with a particular feature is
hampered by the variation in the spectral signature of the water mass.
While the spectral signature emitted from various parts of a given
corn field can be fairly consistent, the signature from a water mass
can vary continuously between its extremities. Consider, for example,
the spectral signature of sediment. Because of the absorptive and
reflective properties of water and sediment, the upwelling spectral
signature may vary with composition, size distribution, and vertical
distribution of the sediment. In some cases, however, algorithms can
be established readily when only one pollutant is present. For example,
an algae bloom can be easily identified because it is highly concen-
trated at the water surface. Incident light on the bloom cannot
II - 19
-------
penetrate very far below the surface. The upwelling light presents
to the remote spectroradiometer a unique spectral signature of the
algae. An example of a bloom signature will be given later in this
paper.
The situation becomes more complex when mixtures of pollutants occur.
Each pollutant may have different absorption and scattering properties
and with multiple, interactive scattering the resultant upwelling
signature is not necessarily the sum of the individual signatures.
Furthermore, the upwelling signature may vary continuously when a
gradual transitional region between two water masses is traversed.
Advanced computer programs will be needed to automatically handle data
of this type, particularly in coastal regions. (The processing of open
ocean data may not be as complex since mixtures are rarer and the sus-
pended matter is primarily phytoplankton.) These programs must be
capable of classifying features, analyzing signature variations between
the classified features, and displaying this information to the user
in a meaningful form.
Examples of MOCS data for two different types of water masses are
presented in this paper: (l) An algae bloom which has a well-defined
signature that can be readily processed on a computer and (2) a plume
of suspended matter which is complex in composition and in spectral
signature. The new information extracted from both cases demonstrates
the value of a multispectral scanner, such as MOCS, as a tool for
studying processes that occur in the hydrosphere.
MOCS INSTRUMENT DESCRIPTION
The MOCS is a visible imaging spectroradiometer which performs multi-
spectral scanning electronically. It has no moving parts. MOCS was
specifically designed for measurements of small differences in ocean
color from space. It measures the intensity in 20 spectral bands at
each of 150 spatial sites of the ocean across the field-of-view. It
is unique in that it uses only one detector and, as a result, it is
compact and very light, weighing only 23 pounds.
Figure 1 is a schematic of the optical arrangement of MOCS and a
listing of its specifications. In operation, light from the water is
focused by the objective lens on the entrance slit. The instrument
is designed to form a high-quality optical image of the ocean surface
on the slit, so that light from one edge of the field-of-view is imaged
at one end of the slit, light from the center of the field is imaged
at the center of the slit, etc. The light is then collimated,
dispersed by a blazed'transmission diffraction grating, and reimaged
on the face of the image dissector. The resulting image consists of
a large number of adjacent spectra, each one composed of radiation
coming from a different site across the instantaneous field-of-view.
The spectra are scanned in sequence in a raster pattern on the photo-
sensitive surface of the tube. The resulting video signal is a
II - 20
-------
measure of the spectral intensities of the light coming from each of
150 spatial sites. The scan rate is such that the whole raster (one
frame) is read out in the time (286 msec) it takes the spacecraft
or aircraft to move forward over the ocean one resolution element .
The scan is then repeated at a rate of about 3-5 frames/sec to give
contiguous coverage of the ocean. The f i eld-of -view of the MOCS for
two different altitudes is given in Table I.
The present MOCS system has an alternate mode of operation made
possible by changing the image dissector. In this mode the spectral
resolution (5 nm) is increased by a factor of 3 at the expense of a
factor of 3 reduction in spatial, resolution.
The output of MOCS is fed to an A/D converter and stored on magnetic
tape. The bit rate from the converter is 1^0 Kbits/sec. A detailed
description of MOCS and its associated electronics can be found in
Reference 1. Figure 2 is a photograph of MOCS.
SPECTRAL SIGNATURE OF ALGAE
An example of a well-defined spectral signature of a water mass con-
sisting essentially of one pollutant was obtained from MOCS data of
Clear Lake, California, on the NASA Convair 990 mission. The lake was
overflown on June 28, 1972, at an altitude of 37,^00 feet. The map in
Figure 3 shows the flight path across the lake and the total field-
of-view (2 by 18 miles) of the MOCS along the path.
False color maps of Clear Lake, shown in Figure k, were generated from
the MOCS data. The algorithms to use are:
IT -| 4 - IQ t
Green Band: Y, = -iiij - «J- (l)
J T
xio,,3
and
Red Band: Y. = (2)
<] T
10, J
where 1^ 4 is the magnitude of band i of the spectrum from spatial
element j of the MOCS data. The bands 9, 10, 11, 19, 20 correspond,
respectively, to center wavelengths of 528, 5^3, 558, 678, and 693 nm.
Selection of the two algorithms is based on data (discussed in Ref. 2)
which support the assumptions that the green band map shows the distri-
bution of all particulate matter near the lake's surface, whereas the
red band map shows only the distribution of algae. Therefore, the
II - 21
-------
strong feature in the lower part of the lake, which appears in both
maps, is assumed to "be an algae bloom.
Using a computer program developed at LaRC, regression analysis was
performed on MOCS data points within this bloom in an attemp-t to
extract a spectral signature of the algae. A linear regression
equation was assumed of the form
YJ = ^ij + Bi
where Yj is the value of a given algorithm for spectrum j 9 and A^
and BJ_ are constants for each band. Several algorithms were tested,
including those of Equations (l) and (2). The results were essentially
the same (Fig. 5). Figure 5 is a plot of A^, the slope of -the regres-
sion equation for each of the 20 spectral bands. In this example, the
algorithm
YJ = '20,0 - ^.j w
was used. In essence, this plot shows the spectral signature of the
algae bloom. The ordinate is the relative change of the signal in band
i with a relative increase in the algae concentration (assuming the
algorithm in Equation (k) correlates directly with concentration). With
an increase in concentration, the signal in the blue bands decrease due
to absorption by the plant pigments. The chlorophyll a, absorption
peaks in bands U and 19 are clearly evident. In the green and red bands
the signals increase as expected due to scattering. The known scatter-
ing peaks in the green (band 12) and the red (band 20) standout. The
crossover point, known as the hinge point, between absorption in the
blue and scattering in the green occurs at band 10. At the hinge point
no signal change occurs with a change in algae concentration. Duntley
(Ref. 3) reported the hinge point for marine algae to be at 523 run.
It appears that for this fresh water specie the hinge point occurs at
5^3 run. The algorithm in Equation (k) was presented in order to show
where this hinge point occurs. If Equation (2) or (3) were used, the
normalization factor I]_Q would normalize the data to band 10 and
thereby obscure the fact that the hinge point occurs at that band.
Figure (6) shows absorption spectra of several different phytoplankton
species that were measured in the laboratory by the author on a Gary lU
spectrophotometer. The spectral signature in Figure 5 looks very
similar to the inverted spectrum of Anacystis marinus, a bl/ue-green
algae. The predominant organism in the lake one day after "the flight
test was a blue-green algae (Aphanizomenon). These data demonstrate,
perhaps for the first time, that a remote sensor can be used to
identify the type of algae in the water in addition to mapping its
relative distribution.
II - 22
-------
NEW YORK BIGHT DATA
A water mass more complex than the algae bloom just presented is
exhibited in the New York Bight data. The assumed complexity is based
upon the observed variations in the spectral data along a path across
the water mass. The data reveals a peculiar transitional signature -
a sequential variation in the spectrum - that might be significant in
establishing remotely reference points of known concentrations within
water masses.
MOCS data collected in the New York Bight on April 7, 1973, was part
of the Marine Ecosystems Analysis (MESA) Program being conducted by
NOAA. The New York Bight mission was organized and directed by the
National Environmental Satellite Service (NESS), a branch of NOAA.
Data were collected with various instrumentation aboard both aircraft
and boats. The general objectives of this mission were to expand our
knowledge of the coastal marine processes and at the same time demon-
strate the value of remote sensing in achieving that objective.
NASA-LaRC participated in the mission by flying MOCS in an NASA
Wallops C-5^ aircraft. Ten passes, each approximately 30 n. mi. long,
were made over the Bight at an altitude of 17,500 feet. The ground
resolution (Table l) of MOCS was 35 feet by 70 feet with a total swath
width of 5,250 feet.
Figure 7 is an ERTS I image (600 - 700 nm band) of the New York Bight
obtained on the morning of the mission. The morning sky was cloudless
and clear; the sea was very smooth. The afternoon was marred by a
continuous buildup of haze. An unusually large plume of suspended
matter was clearly evident in the Bight fed by the Hudson River.
Fairly sharp rectangular boundaries of this plume can be seen in
Figure 7- Beyond the plume is an acid waste dump. MOCS data taken
over this dump has been used to construct a false color map (Fig. 8).
This map contours the region in which the upwelling signal is greater
than 50$ of the maximum upwelling signal in the 573-nm band. At the
outer boundary of this map the strength of the acid is 50$ of that at
the center of the dump. To demonstrate the sensitivity of the instru-
ment, the map is color coded in 5$ intervals down to the 60% level;
the last two intervals are uncolored. The sharpness of the contour
lines illustrates the excellent resolution of MOCS.
Many interesting signature variations are exhibited in the MOCS data
of the Bight. Correlation studies will be performed on these data when
the ground truth data become available. As an example of these varia-
tions, the data along one flight track (designated it5-1) will be
presented. This flight path is shown in Figure 9-
Plots of data along the center of the track for three spectral bands
are shown in Figure 10. The ordinate has been adjusted such that the
II - 23
-------
upwelling signal from blue water is equal to zero relative signal. Each
data point in Figure 10 is an average of four spatial elements which
reduce the resolution to ?0 feet by 1^0 feet. A change of one frame
count corresponds to an advancement of l^tO feet along the flight line.
For purposes of discussion, the ocean along the track has been divided
into regions, designated Bl, B2, etc., in Figure 10.
The plots demonstrate the complex variation in the upwelling signal as
the aircraft flew over blue water, crossed the plume "boundary, passed
over the plume, and approached land. Region 1 consisted of relatively
clear blue water, although coincident signal variations in the three
bands in the latter half of the region indicate small features. In
Region 2, the signals for the three bands increase as the plume boundary
is approached. Within the plume, Region 3, the relative magnitudes of
the three bands diverge. The blue band (U68 run) decreases, the orange
band (603 nm) increases, and the green band (5^3 nm) remains rela-
tively constant. In Region 1» the blue and green bands fall and rise
together while the orange band remains relatively constant. Notice that
the blue band has dropped to zero relative signal. In Region 5 all
bands rise. The divergence that occurred in Region 3 appears to be
reversing in Region 6. However, as land is approached, bottom reflec-
tion appears to influence the directional variance of the signals in
both Regions 6 and 7.
It is instructive to examine the relative signal variations between
bands by plotting the data as shown in Figures 11 and 12. Figure 11 is
a plot of every third point in Figure 10 for the blue and orange bands.
The cluster of points which fall within each region is indicated in the
figure. The sequential variation of the data is clearer in Figure 12,
where lines are drawn between the averages of sets of 20 data points
along the track.
At least two models are suggested by these signal variations. In the
first model the variations are due to the size and depth distributions
of sediment along the track. Backscattered light from clear water is the
result of selective scattering, known as Rayleigh scattering, by water
molecules and small particles (Ref. h). Rayleigh scattering predominates
when the size of the scatterers are much smaller than the wavelengths of
the scattered light. The amount of scattering is inversely proportional
to the fourth power of the wavelength. In addition, water molecules
selectively absorb light much more strongly in the red end of the spec-
trum than in the blue, greatly reducing the amount of red light avail-
able for backscattering. However, as the size of the suspended particles
increases to the order of one micrometer, Mie scattering (Ref. U) becomes
significant. Mie scattering is predominantly in the forward direction
and is nonselective with wavelength. As a. result, scattering decreases
in the blue region of the spectrum and increases in the red region.
However, since the particles in the water are not of one size, the
analytical analysis becomes more complex.
II -.24
-------
With this model the data in Figure 10 could be explained as follows.
In Region 2, the particles are small and distributed somewhat evenly
below the surface. As the boundary is approached, the number of
particles increase. Within the plume, the particles are larger and
more numerous resulting in a considerable reduction in the photic zone.
The transition from a low concentration of small particles to a high
concentration of larger particles results in a decline in the signal in
the blue bands and a rise in the red bands.
A second model is suggested by the fact that algae absorbs light in the
blue bands and scatters light in the red bands. Therefore, the varia-
tion in signals in Begions 2 and 3 could also be influenced by the
transition from a region of primarily sediment to a region of algae
mixed with sediment. The ground truth may resolve these possibilities.
A very similar occurrence of this signal variation is seen in MOCS data
collected along the 36° parallel off Cape Hatteras. This flight was
part of the NASA Convair 990 mission. Data were collected at 37,300 feet
starting about Uo n. mi. offshore and ending at the shoreline. The simi-
larity between Figures 13 and lU for the Cape Hatteras data and Fig-
ures 11 and 12 for the New York Bight data is striking. Perhaps this
phenomenon is a fundamental characteristic of certain types of plumes
and can be useful in establishing reference points of known concentra-
tions. In other words, the transitional signature, the triangle in
Figure lU, could be uniquely associated with certain distributions of
the suspended matter. If known concentrations can be associated with
this signature, then the distribution ef the matter in the surrounding
region can be determined. Reference points of this nature would be
invaluable in the processing of remote sensing data.
CONCLUDING REMARKS
The examples of data presented in this paper demonstrate the value
of multispectral scanners, such as MOCS, as a remote sensing tool for
analysis of processes that occur within the hydrosphere. From the
Clear Lake data, a spectral signature was extracted which identified
a feature, observed in a false color map of the lake, as & blue-green
algae bloom. The false color map shows the relative concentrations of
this bloom. Generation of maps of this type which display absolute
values of concentration using remote sensing data appears to be quite
feasible.
The sensitivity of MOCS is demonstrated by the well-defined contour
lines (color coded in 5$ intervals) on the false color map of the acid
waste dump in the New York Bight.
The New York Bight data also revealed a signature associated with the
transition between two water masses. Examination of the Cape Hatteras
II - 25
-------
data showed the same transitional signature. The value of this signa-
ture has not been fully determined, but it has the potential of estab-
lishing reference points of known concentrations and could be invaluable
in the processing of remote sensing data of the hydrosphere. These
examples demonstrate the complexity of the signatures associated with
suspended matter. Signatures may vary gradually and continuously, not
only between water masses, but within them. To facilitate user applica-
tion of data of this type, computer programs are needed that can properly
assess the signature variations and display on maps in a meaningful form
the variations in the water associated with these signatures,
REFERENCES
1. White, P. G. ; Jenkin, K. R.; Ramsey, R. C.; and Sorkin, M.:
"Development and Flight Test of the Multichannel Ocean Color
Sensor (MOCS)." NASA CR-2311, 1973-
2. Grew, Gary W..: "Signature Analysis of Reflectance Spectra of
Phytoplankton and Sediment in Inland Waters." Remote Sensing of
Earth Resources, Vol. II, Benson Printing Company, Nashville,
Tennessee, 1973.
3. Duntley, Seibert Q. : "Detection of Ocean Chlorophyll From Earth
Orbit." Fourth Annual Earth Resources Program Review, Vol. IV,
Presented at the Manned Spacecraft Center, Houston, Texas,
January 17-21, 1972.
k. Williams, Jerome: "Optical Properties of the Sea." United States
Naval Institute, Annapolis, Maryland, 1970.
II - 26
-------
TABLE I. FIELD-OF-VIEW OF MOCS
Altitude
Field-of-viev
Swath width
Ground resolution
Aircraft
IT ,500 feet
17.1°
5,250 feet
35 by 70 feet
Spacecraft
,500 n. mi.
17.1°
150 n. mi.
1 by 2 n. mi.
II 27
-------
OPTICAL SCHEMATIC
RIGHT RED
BLUE
LEFT
FACE OF IMAGE
DISSECTOR
REIMAGING
LENS
DIFFRACTION
GRATING
COLLIMATING
LENS
SLIT
OBJECTIVE LENS
SPECIFICATIONS
400-700 nm SPECTRAL RANGE
15 nm SPECTRAL RESOLUTION
20 SPECTRAL BANDS
150 SPECTRA/SWATH WIDTH
17.1° FIELD OF VIEW
22.5 pounds
7.5w3tts
.35 cubic feet
FLIGHT DIRECTION
Figure 1. Optical schematic and specifications of MOCS.
Figure 2. Photograph of MOCS,
II - 28
-------
Figure 3. Flight path of HASA Convair 990 and field-of-view of MOCS
(2 "by 18 miles) over Clear Lake at 37,^00 feet.
I
A. GREEN BAND
I
Figure k
B. RED BAND
False color maps (originals in color) of Clear Lake generated
from green band and red band algorithms.
II - 29
-------
RELATIVE
SIGNAL
CHANGE, A.
1.2
1.0
.8
.6
.4
.2
0
-.2
-.4
-.6
-.8
-1.0
-1.2L
L
I
I -I I
I I I I
I 1 I I
408 438 468
i L
498 528 558 588 618 648 678
WAVELENGTH, nanometers
J I _L 1 L_J I I I 111
1234567
8 9 10 11 12 13 14 15 16 17 18 19 20
SPECTRAL BAND, i
Figure 5- Spectral signature of an algae "bloom in Clear Lake.
II - 30
-------
ANACYSTIS MARINUS
GYMNODINIUM
SIMPLEX
NANNOCHLORIS
OCULATA
ISOCHRYSIS GALBANA
1 1 1 1 1 1 1 1 1 1 1 1 1
300 400
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
500 600 700 800
WAVELENGTH , nm
Figure 6. Absorption spectra of sample phytoplankton species.
II - 31
-------
' ;
trt
•
Figure 7. ERTS I imagery (BAND 5, 600-700 nanometers) of the New York
Bight area on April 7, 1973.
II - 32
-------
FALSE COLOR MAP FROM MOCS DATA
NASA C-54 AIRCRAFT - ALTITUDE: 17500ft
AREA: 0.75 x 7.5 n mi
RED: HIGHEST ACID CONCENTRATION
BLUE: 60-65% OF HIGHEST CONCENTRATION
Figure 8. False color map of acid waste dump in New York Bight.
The outer "boundary is 50% of the acid strength in the center of
the dump.
II - 3;
-------
HUDSON
RIVER
STATEN
ISLAND
EW JERSEY
ROCKAWAY BEACH
TRACK 45-1
ATLANTI C
OCEAN
Figure 9. Sketch of the EFTS I imagery (Fig. 7) shoving approximate
boundaries of the plume, acid waste, and one track of the NASA C-5^
aircraft.
II - 34
-------
200
180
160
140
120
100
RELATIVE on
SIGNAL
60
40
20
0
-20
-Aft
o 468 nm BAND nyE -v '
D 543 nm BAND MARKER \ ' rROCKAWAY
O 603 nm BAND ,' '. BEACH
PLUME i 1 1
^BOUNDARY t _• • ('
.,^-.^vv •' '*" /•
'••• -/. i RROCKAWAY
.'•' \<>'.y .' INLET
/' ^w*^'^!
~" ,' ,!(•'•' |f
-3' i
''f$*\ • >
~ ; t JM; 'j,R „ /
• . £*;/' ' r*\ /^
^i^^¥' Xy^V
L
' Rl i p? i R3 i R/l iR5iRfij R7 I
i i i i i i i I
0 100 200 300 400 500 600 700 800 900 1000 1100 12001300
FRAME COUNT
Figure 10. Relative signals along center of track U5-1 for three
spectral bands. The signal levels have been adjusted such that
upwelling light over blue water equals zero relative signal.
II - 35
-------
280 r-
240
200
EVERY 3nd POINT
RELATIVE
SIGNAL,
603 nm
BAND
160
80
40
0
R6
R7
R4 R3?,:
R2
I
-40 0 40 80 120 160 200
RELATIVE SIGNAL,468 nm BAND
240
Figure 11. Relative signals of 603-nm band versus ^68-nm band for
data along track ^-5-1 in New York Bight. The regions designated
in Figure 10 are approximated on this plot.
II - 36
-------
280
240
200
160
RELATIVE SIGNAL, 120
603 nm BAND
80
40
0
-40
20 POINT AVERAGES
-40 0 40 80 120 160 200 240
RELATIVE SIGNAL,468 nm BAND
Figure 12. Relative signals of 603-nm band versus U6Q-nm band for
data along track ^5-1 in Nev York Bight. Lines are drawn between
averages of consecutive sets of 20 data points.
II -37
-------
1.6
1.2
.8
RELATIVE SIGNAL,
603 nm BAND
.4
-.4
EVERY 3nd POINT
-.4'
.4 .8 1.2 1.6
RELATIVE SIGNAL,468 nm BAND
2.0
2.4
Figure 13. Relative, signals 'of 603-nm band versus ^68-nm band for
data along 36° parallel track off"Cape Hatteras. (Compare with
Fig. 11.)
II - 38
-------
1.6
20 POINT AVERAGES
1.2
.8
RELATIVE SIGNAL,
603 nm BAND
.4
0
I I
ill
I, | II
-A 0 .4 .8 1.2 1.6
RELATIVE SIGNAL,468 nm BAND
Figure Ik I Relative signals of 603-nm band versus ^68-nm band for
data along 36° parallel track off Cape Hatteras. Lines are drawn
between averages of consecutive sets of 20 data points. (Compare
with Fig; 12.)
II - 39
-------
SUMMAEY OF NASA LANGLEY RESEARCH CENTER REMOTE SENSING ACTIVITIES
UNDER THE ENVIRONMENTAL PROTECTION AGENCY INTERAGENCY AGREEMENTS:
POTENTIAL FOR REGIONAL APPLICATIONS
James L. Raper
NASA Langley Research Center
Hampton, Virginia
Presented at the Second Environmental Quality Sensor Conference
Las Vegas, Nevada
October 10-12, 1973
II - 40
-------
SUMMARY OF NASA LANGLEY RESEARCH CENTER REMOTE SENSING ACTIVITIES
UNDER THE ENVIRONMENTAL PROTECTION AGENCY INTERAGENCY AGREEMENTS:
POTENTIAL FOR REGIONAL APPLICATIONS
James L. Raper
NASA Langley Research Center
Hampton, Virginia
BIOGRAPHICAL SKETCH
Mr. Raper is a Senior Aerospace Engineer with 15 years experience in
aerospace research projects at the Langley Research Center. He has
managed or assisted in the management of projects involving reentry
heat protection and communication, hypersonic transition, parachute
qualification, and rocket motor qualification. In the last year he
has managed the NASA/EPA Interagency Agreement for Water Pollution
Sensor Evaluation. In addition to establishing a flight-test program
to meet the needs of nine investigators, it was necessary for him to
devise an interim capability for multispectral scanner data analysis
and establish requirements and plans for a more permanent Langley
capability. He has recently joined Langley1s Environmental Quality
Program Office with responsibility for managing studies to define
sensor modules for future pollution monitoring satellites.
ABSTRACT
The purpose of this paper is to report the current status of all
except three of the activities being conducted under the NASA/EPA
Interagency Agreement for Water Pollution Sensor Evaluation. The
three topics excluded, the multiwavelength laser, MOCS, and Pollutant
Response, are separately reported in this session. A description of
eight investigations is included. Expected applicability of these
investigations to regional situations is described.
INTRODUCTION
Dr. Fletcher, the NASA Administrator, in his talk to the 37th Annual
Meeting of the National Wildlife Federation, on March 17, 1973> said
the following:
"No part of the changing, moving face of the globe we inhabit is free
of human influence or removed from human interest. We therefore can
afford to leave no part unmonitored. We need to know the condition of
our environment and in time to take appropriate action." Many of the
capabilities he described were identified as already possible; the
technologies exist or are under development. He said, "We are learn-
ing to interpret and as we do so, we learn what other related parame-
ters and phenomena we need to observe to get a total picture." Early
in his talk, Dr. Fletcher pointed out to the audience that the systems
L-9213
II - 41
-------
NASA is developing are just tools and must not be thought of as tech-
nological solutions in themselves to the serious environmental chal-
lenges, problems, and issues with which the world must cope. "The
basic purpose of these tools," he explained, "is to collect and com-
municate data from which information can be extracted."
ERTS-1 and Skylab are two of our best known uses to date of these
tools. There is also a large effort within NASA's Office of Applica-
tions to define applications for sensors within our pollution and
earth resources programs. The AAFE (Advanced Applications Plight
Experiments) program, Figure 1, at Langley is one such effort and has
the objective of developing an effective inventory of space applica-
tion experiments from which future missions and flight experiments for
approved flights may be proposed. AAFE does in fact bring instrument
techniques to the point where they can be applied to problems such as
those faced by EPA. We are also interested in investigating with the
User agencies the design of an aircraft remote sensing system to
investigate the usefulness of synoptic measurements of air and water
quality on urban and regional scales. Underscoring these Langley
activities is the recent designation of Langley as NASA's Focal Center
for Space Application's Environmental Quality monitoring. In this
role Langley is responsible to coordinate the agency's pollution
monitoring activities in a way which will most directly benefit
agencies such as the EPA.
Defining remote sensor applicability for EPA mission requirements is
the subject of the interagency agreement between EPA and NASA — in
existence since June 1972 — and the status of work under the agreement
is the subject of this paper. The interagency agreement, under the
cognizance of Dr. Harvey Melfi of Las Vegas NERC and John Koutsandreas
of EPA Headquarters, is designed to evaluate the applicability of
various existing remote sensors for detecting and monitoring various
types of water pollution. The agreement specifically directed inves-
tigations of pulsed laser systems, multichannel scanning radiometers,
multiband cameras, infrared scanners, and second derivative infrared
spectrometers. Subsequently, investigations in 12 areas were under-
taken in support of these objectives. Reports of activity with the
pulsed laser (Multiwavelength Lidar), the multichannel scanning
radiometer (MOCS), and multiband cameras (Pollutant Response) are
presented in the previous three papers. This paper will deal with
summary status reports of activities to date in the other areas.
Before proceeding to a discussion of activities being conducted under
the interagency agreement, it is important that three observations of
an introductory nature be stressed. First, if such agreements are to
provide maximum benefits within EPA it is necessary that more dialog
exist between the agencies with regard to specific EPA mission require-
ments, the manner in which EPA field surveys are conducted, and EPA
capabilities for data analysis. Second, as noted in Dr. Fletcher's
talk, the basic purpose of the remote sensing tools is to collect and.
communicate data from which information can be extracted. Although
II - 42
-------
there is some polishing to do on the collecting end of the present
investigations, the most work remains in developing ways to extract
information from the collected data relative to defining the water
pollution situation quantitatively as well as qualitatively. Third,
the use of remote sensing will probably never replace the requirement
for selected in situ sampling. Remote sensing appears to best fit the
situation where large areas are to be frequently covered and where the
synoptic view is required to provide an evaluation of the inter-
relationships between factors being sensed. Remote sensing probably
can best indicate where to concentrate the in situ sampling. Also,
in situ sampling will continue to find application because the remote
sensor may not be sensitive to the desired parameter or because of
technical or economic constraints associated with providing the
required spectral or spatial resolution.
SALINITY MAPPING
A 2.65-GHz S-band microwave radiometer has been developed as a part of
Langley1 s AAFE program to remotely sense sea state. A byproduct of
the data has been information about surface salinity. As indicated in
Figure 2, the 2.65-GHz microwave radiometer is insensitive to salinity
variation below about 10 ppt. Also, the sensitivity to salinity above
10 ppt does not meet measurement requirements. The 2.65-GHz radiometer
was used, however, to demonstrate the salinity mapping concept and to
provide confidence that a radiometer dedicated to salinity mapping
could be built. As indicated in Figure 3, a l.i|~ GHz L-band radiometer
system would provide the required operational characteristics. If the
0.1 K accuracy in brightness temperature demonstrated in the 2.65-GHz
system is maintained in the L-band system, a variation of 0. 2 to
0.3 ppt of salinity change will be detected over the range of 3 to
35 ppt of salinity which is the operational range of interest to EPA.
The L-band system is now being assembled and is expected to be ready
for EPA demonstration flight tests in August
Figure 3 shows the type of salinity survey that EPA would be
interested in conducting at periodic intervals. The map shown was
produced by massive manual sampling using boats. A comparison of the
resources required to sample just the lower half of the bay is pre-
sented along with the expected expenditure of resources when using the
microwave radiometer remote sensing approach. The radiometer approach
is equally applicable for monitoring changing conditions around a
powerplant, around a desalinization operation, or due to a massive
storm. Again, one of the obvious benefits of the salinity remote
sensor is its ability to cover quickly and frequently a large area and
provide a synoptic view of conditions with a small manpower
expenditure.
Multispectral Scanner Detection of Algae
One of the principal EPA interests with regard to monitoring water
quality involves detecting and monitoring water algae content. The
II - 43
-------
amount of algae is indicative of the amount of waste nutrients being
dumped into the water and is an indicator of the suitability of the
water for human consumption, recreation, and wildlife. The purpose of
this investigation was to use the multispectral scanner over an area
of varying algae concentration to accomplish two objectives. The first
is to determine the optimum spectral ranges which identify algae in the
water and which permit computer separation of the algae from other
features. The second is to determine whether it is possible to produce
a quantitative map of algae content for a water body such as a tidal
river. The NASA-JSC Earth Resources Program C-130 aircraft and 2k-
channel multispectral scanner were used to gather the data in October
1972 over the Patuxent River in conjunction with an intensive field
survey of water parameters which was being conducted simultaneously.
Analysis of the data has proved to be a major task. Langley did not
possess the capability for analysis of such imagery data and much
effort has been devoted to developing that capability. The Purdue-
LARS remote terminal located at the GSFC and the Bendix-Ann Arbor
analysis systems have been used for the Patuxent River data analysis.
Figure k shows the analysis approaches which can be used and points
out the dependence of analyses on ground truth data. Figure 5 shows
a portion of the Patuxent River and is the result of the supervised
classification of the nonwater areas with the Bendix analysis system.
The two dots along the upper bank of the river at the right side of
the image are the boats from which the ground truth data were taken.
Figures 6 to 11 present the results of supervised (since the features
to match were specified even though not known) classifications of
spectrally similar' areas in the image. For each classification, the
computer was instructed to locate other areas spectrally similar to a
selected training site. A training site around the boat was selected
because of the indicated high chlorophyll quantity, IkO mg/m^. Indi-
vidual examination of some of the Ik channels of multispectral scanner
data indicated other spectrally different areas of water and thus a
small portion of each was selected as a training site for classifica-
tion of the image. An unsupervised classification of the total image
area to define all the statistically different spectral classes has
not yet been attempted.
Figure 6 is a water classification based on a training set taken at
the boat and thus should contain all the area that responds spectrally
as water containing ~LkO mg/m3 of chlorophyll. Figure "{ is a water
classification based on a training set taken at the outflow of the
upper left stream into the river. Figure 8 is a water classification
based on a training set taken in the inland body of water just behind
the boat. Figure 9 is a water classification based on a training set
taken in the river to the left of the boat. Figure 10 is a water
classification based on a training set taken in the water at the left
of the sand bar at the river bend. Figure 11 is a water classifica-
tion based on a training set taken in the water in the center of the
river at the left of the boat. The principal point which can be made
II - 44
-------
with Figures 5 to 11 is that it is possible to use computer analyses
to identify spectrally different classes of water. Unfortunately, not
enough ground truth data exist with this set of data to identify each
class of water. Future experiments will be planned with more ground
truth so that a "calibration curve" can be constructed. The above
comments infer that the different water classes are related to differ-
ent amounts of chlorophyll which would permit contouring as desired.
Much work remains to be completed to verify this relationship. Even
though the specific multispectral scanner used in this investigation
was never meant for operational use, the potential for this type of
remote sensing device is great because of its ability to accurately
define the spectral signature of a pollutant. An understanding of the
multispectral scanner information will permit the development of effi-
cient and economical sensors for many types of water pollutants and
the understanding of information from other sensors.
For the immediate future more analysis of the present data will be
performed, and other remotely sensed data taken simultaneously will be
cross- correlated in an effort to develop means for identifying the
water classes.
ERTS INVESTIGATIONS
Many of the investigations under the interagency agreement have quite
naturally raised the question, "How applicable is current ERTS imagery
to EPA water pollution monitoring requirements?" We know from present
ERTS investigators' findings that the ERTS data indicate various water
pollutants. This question is underscored by the importance of the
synoptic view. There is also interest in how conclusions derived from
ERTS data compare with those derived from ground truth and from air-
craft remote sensors. Five separate investigations involving ERTS are
in progress and each has the common objective of defining how to use
ERTS data in combination with other remotely sensed and ground data to
provide a water pollution detection and monitoring capability.
Lake Eutrophication — The National Eutrophication Survey is a major
effort within EPA to define the eutrophic state of specific inland
lakes throughout the United States. A significant output of that sur-
vey will be data to be used in constructing predictive computer models
of the eutrophication process. The purpose of this ERTS investigation
is to determine whether data can be provided by ERTS which will permit
continuing evaluation of the eutrophic state of lakes and which can be
used in the computer predictive models. Figure 12 presents the current
status of this investigation. An ERTS-B proposal was submitted in
January 1973 and will be revised to reflect a more complete problem
definition and work plan prior to the proposal evaluations in January
Upper Bay Pollution - The types of water pollution found in a large
bay surrounded by highly industrialized sites and cities is very
different from that found in an inland lake. This investigation is
II - 45
-------
directed at defining the applicability of ERTS to detecting and monitor-
ing various water pollutants such as algae, sediment, oil, and sewage
in a large bay area — the Chesapeake Bay. Figure 15 presents the cur-
rent status of this investigation. An ERTS-B proposal was submitted
in January 1973 and will be revised to reflect a more complete problem
definition and work plan prior to the proposal evaluations in January
Land Fill Classifications — A problem of particular concern to the
State of Pennsylvania and EPA, Region III is the proliferation of solid
waste dumps throughout the state. Current laws regulate location and
type of dump to a land fill operation. A means of detecting noncompli-
ant dumps and for continuous surveillance of the operation of all
others is required. This investigation is directed at evaluating to
what extent ERTS data can be used in locating and monitoring land
fills. Figure l^t- presents the current status of this investigation.
Early efforts directed at manual viewing of the magnified ERTS images
proved inadequate, especially when attempting to determine if land
fills or dumps existed in unrecorded locations. Subsequent analysis
has depended upon the Purdue-LARS analysis approach for supervised
classification. The computer training sets were defined by aircraft
photographic imagery taken over known land fill locations. Results,
to date, indicate that land fill sites as small as 10 acres can be
detected in ERTS imagery. When a supervised classification of a large
part of an image was performed to pick out land fills, only about 10$
of the locations thus identified were actually land fills. This result
probably occurred because the land fills are nonhomogenous spectrally
and have limited unique features which are spectrally identifiable.
Also too imprecise knowledge exists of what is and is not a land fill.
Future analysis activities will concentrate on improving the degree of
land fill recognition. The aircraft 2l|~ channel multispectral scanner
data will be analyzed for land fill identification to determine
whether greater spectral resolution improves identification and to
better define the spectral signature of a "typical" land fill.
Acid Mine Drainage - EPA (Region III) and the States of Pennsylvania
and West Virginia are particularly affected by seepage from shaft and
open coal mines into the surrounding streams. The resulting acid
stream causes destruction of adjacent vegetation. Knowledge of the
source of the seepage is required in order to enforce present laws or
to take measures to prevent seepage from abandoned mines.. The purpose
of this investigation is to determine the most cost effective system
for detecting and monitoring the effects and extent of acid mine
drainage using ERTS as appropriate. ERTS should be most effective in
providing a synoptic view and in identifying large areas of vegetation
destroyed by acid mine drainage. ERTS is not expected to be able to
monitor the water quality in small streams. This will perhaps only be
possible with manual in situ sampling. Figure 15 presents the current
status of this investigation. The primary effort is contractual and
the contractor will present a detailed paper elsewhere in these pro-
ceedings. Figure l6 outlines the contractual tasks. The emphasis on
II - 46
-------
land use arises from the fact that land use maps do not exist for the
selected test area and from the requirement to correlate land use with
acid mine drainage effects. Figure 17 indicates the test site loca-
tion chosen for this study, which is the head waters for the Potomac
River. The rectangular outline indicates coverage provided by a
recent ERTS underflight mission. The end effort of the contract is to
define the optimum mix of sensing techniques for performing the
required monitoring. It will remain for future contracts, NASA-LaRC,
or EPA programs to evaluate the proposed technique.
Data Automation and Transmission — In order to be in a position to
react quickly to a pollution problem, one must receive monitoring
information rapidly and continuously and in a form which is readily
assimilated. At the present time EPA (Research Triangle Park KERC) is
receiving raw, air quality monitoring data weekly in magnetic tape
form. Technology exists for automating the conversion of these data
into a readily understandable form and for transmitting the informa-
tion twice daily to a selected location. The ERTS Data Collection
Platform is in fact specifically designed for this type of data relay.
The purpose of this investigation, see Figure 18, is to demonstrate
this technology by installing an automation/transmission system in the
District of Columbia CAMP station. Even though air quality parameters
will be monitored, the system to be demonstrated will be equally appli-
cable to water quality parameters. Figure 19 summarizes the require-
ments that the system is being designed to accommodate. Figure 20
presents a schematic of the present CAMP station arrangement and the
proposed noninterference automation/transmission system. For the sys-
tem demonstration, the data handling and converting computer is being
placed in the CAMP station in order to centralize the total effort into
one location. In an operational arrangement, data from a number of
monitoring stations could be transmitted to a single site where the
computer is located.
SOUTH RIVER SEWAGE OUTFLOW DETECTION
Large portions of the South River in Maryland are presently closed to
fishing because of the presence of coliform bacteria from private
septic tank seepage into the river. Personnel in the Anne Arundel
County Health Department are responsible for eliminating the septic
tank violations and for patrolling, the approximately li-30-mile perimeter
of the South River at periodic intervals to verify no further viola-
tions. The purpose of this investigation is to determine what, if any,
remote sensing techniques exist which will enable the Health Department
to detect violations by checking at frequent intervals. The approach
to date has been to use various films and filters, a radiation thermom-
eter, and a thermal IR scanner over known outfall locations in an
attempt to identify an approach for further study. Manual viewing of
the films has been inconclusive. Density slicing of the same film will
be performed in the near future. The thermal IR scanner used proved
not to have the required resolution for detecting small outflows, even
at the low flight altitudes employed. Future missions with a different
II - 47
-------
IB scanner are planned. The potential for application of remote sensing
to this problem is fairly obviousj however, the sensors on hand which
have been tried may not possess the required spatial resolution for such
a small seepage as that which one might expect from a private septic
tank.
CONCLUDING REMARKS
Our experience to date with investigations being performed as a part
of the NASA/EPA interagency agreement strongly underscore Dr. Fletcher's
comments about having the tools but not yet fully understanding how to
get information from the data. Developing methods for getting this
information will be the most important benefit for EPA. Also empha-
sized by our experience is the critical importance of ground truth
data for developing an understanding of the sensors capabilities.
Finally, this paper has tried to show in various ways that remote
sensing will not replace present methods of manual sampling. It is
more likely to raise enough questions which will require increased
manual sampling. Most importantly, such sampling will not be random
but will be at specific locations where problems are indicated.
ACKNOWLEDGMENTS
Acknowledgment is hereby made to the following Langley employess whose
work was summarized herein: Bruce M. Kendall, Ruth I. Whitman,
Robert W. Johnson, John B. Hall, Joseph W. Drewry, and James N.
Chacamaty.
II - 48
-------
OBJECTIVES
• TO DEVELOP AN EFFECTIVE INVENTORY OF SPACE APPLICATION EXPERIMENTS
FROAA WHICH FUTURE MISSIONS AND FLIGHT EXPERIMENTS FOR APPROVED
FLIGHTS MAY BE PROPOSED
APPROACH
• FUND THE DEVELOPMENT OF SELECTED EXPERIMENTS, BEGINNING AT A STAGE
WHEREIN THE APPLIED RESEARCH AND ENGINEERING FEASIBILITY HAS BEEN
ESTABLISHED AND LEADING UP TO, BUT NOT INCLUDING THE PRODUCTION OF
H PROTOTYPE HARDWARE
* STATUS
INITIATED INFY70
54 INSTRUMENTS FUNDED TO DATE
11 INSTRUMENTS TO BE FLOWN IN AIRCRAFT MISSIONS
6 INSTRUMENTS DEMONSTRATED IN AIRCRAFT MISSIONS
23 INSTRUMENTS SELECTED OR BEING CONSIDERED FOR SATELLITE MISSIONS
14 INSTRUMENTS CURRENTLY UNDER DEVELOPMENT
Figure 1. Advanced applications flight experiments.
-------
0
SALINITY, PARTS PER THOUSAND
10 15 20 25 30
35
H
H
Ul
O
CHANGE IN
BRIGHTNESS
TEMPERATURE.
deg.-K
10
15
L4GHz
t-BAJMD
Figure 2. Microwave radiometer, change in brightness temperature with change in salinity..
-------
CHESAPEAKE BAY
SURFACE SALINITY (0/00
AUTUMN AVERAGE
SALINITY SURVEY OF
CHESAPEAKE BAY
AND TIDAL RIVERS
LOWER BAY REQUIRES
o - 80 BOATS
-160 MEN
-129 STATIONS
- 2 DAYS
OR
- '2 MEN
- 2 DAYS
-129 STATIONS
- HELICOPTER
WHOLE BAY
WILL REQUIRE:
- 1 DAY
- 2 MEN
- AIRCRAFT
W/MICROWAVE
RADIOMETERS
Figure 3. Map of Chesapeake Bay showing average salinity variations of autumn.
II - 51
-------
SUPERVISED CLASSIFICATION
• IDENTIFY ALL PARTS OF THE IMAGE WHICH ARE
SPECTRALLY SIMILAR TO THIS AREA WHICH
GROUND TRUTH HAS IDENTIFIED
^ - UNSUPERVISED CLASSIFICATION
S • SEPARATE ALL PARTS OF THE IMAGE INTO SPECTRALLY
SIMILAR AREAS AND WHICH WILL BE IDENTIFIED AT A
LATER DATE
Figure k. State of the art in MSS data analysis demands complete dependence on ground truth.
-------
•H
H
Ul
CO
Figure 5- Patuxent River multispectral scanner data.
-------
«•*•/*-./••'
K
H
J
Lfi
Ilil&M&Jap;
Figure 6. Patuxent River — training set taken at boat.
-------
H
H
Ul
\ ..v vV
V*...*
Figure 7. Patuxent River — training set taken at entry of upper left stream into river.
-------
W
, s S
-•"V
H
H
CTi
/
.. '"^mm
Figure 8. Patuxent River — training set taken in inland body of water behind boat.
-------
H
H
I
Ul
^
,:">.,.,. ^
Figure 9- Patuxent River — training set taken in the river to the left of the boat.
-------
• v *
HI
H
I
CO
Figure 10. Patuxent River — training set taken in the water at the left of the river bend
sand bar.
-------
H
H
I
Ul
•: .v;-4>*..::--T:-v.^.v- -
t* . r V* i v •* ' . i» ' '^*
s ..;•" .*••••&?. '. .•:-.. -\
.. •* *•••.;.' •
:•'•'" i
Figure 11. Patuxent River — training set taken in the water in the center of the river at
left of boat.
-------
OBJECTIVE -
STATUS -
H
H
O
OUTLOOK -
TO DEFINE :
• CURRENT STATE OF EUTROPHICATION - UTILIZING ERTS
• WATER SHED USE, WATER USE
• NUTRIENT CYCLE
ONE AIRCRAFT PHOTO MISSION SIMILTANEOUS WITH NES AND ERTS OVER
• KERR LAKE
• LAKE CHESDIN
• CHICKAHOMTNY LAKE
ERTS 1 IMAGERY BEING PROVIDED AUTOMATICALLY
DATA ANALYSIS UNDERWAY
ERTS B PROPOSAL PENDING
EVALUATION OF SEASONALITY EFFECT
PROGRESS REPORT - JANUARY ' 74
COORDINATION - EPA, LAS VEGAS NERC
Figure 12. EFTS — lake eutrophication.
-------
OBJECTIVE - •
STATE OF EUTROPHICATION
INPUTS TO EXISTING POTOMAC AND PATUXENT MODELS
STATISTICAL CORRELATION - ERTS AND EPA SAMPLING
LOCATE POLLUTION SOURCES AND OUTFALLS
STATUS -
H
H
DEVELOPING FIELD TEST PROGRAM
EVALUATING PRESENT SENSORS FOR QUANTITATIVE RESULTS
ERTS 1 IMAGERY BEING PROVIDED AUTOMATICALLY
DATA ANALYSIS UNDERWAY
ERTS B PROPOSAL PENDING
OUTLOOK -
PROGRESS REPORT - APRIL 1974
COORDINATION - EPA, REGION III ANNAPOLIS FIELD OFFICE
Figure 13- ERTS — upper Chesapeake Bay pollution.
-------
OBJECTIVE - DEFINE UTILITY OF ERTS FOR LOCATING AND CLASSIFYING LANDFILLS AND
OTHER UNIQUE LAND SURFACE USES
STATUS - • THREE AIRCRAFT PHOTO MISSIONS FLOWN
• CORRESPONDING ERTS IMAGERY OBTAINED
• DATA ANALYSIS UNDERWAY
ANALYSIS - • SITE LOCATION - PHOTO AND COUNTY MAP
SEQUENCE * GRID MAP
• ERTS MSS DATA ANALYSIS VIA LARS
RESULTS - • SELECTION RATIO
r""1 ~- ' - MI—«™-«"-
H A
PROBLEMS - • EDGE EFFECTS
• NON-HOMOGENOUS
• UNIQUENESS
OUTLOOK - • FURTHER LARS TERMINAL ANALYSIS
• PROGRESS REPORT - APRIL ' 74
COORDINATION-EPA, REGION III HEADQUARTERS
Figure ~Lk. ERTS - landfill classification.
-------
H
H
cr>
to
OBJECTIVE -
STATUS -
OUTLOOK -
TO DEFINE MOST COST EFFECTIVE SYSTEM FOR DETECTING AND
MONITORING EFFECT AND EXTENT OF ACID MINE DRAINAGE UTILIZING
ERTS AS APPROPRIATE
• CONTRACT EFFORT UNDERWAY AND CONTINUING
• PRIOR DATA BEING ASSEMBLED
• ONE NEW AIRCRAFT MISSION FLOWN
• ERTS 1 IMAGERY BEING PROVIDED AUTOMATICALLY
• RESULTS OF DATA REVIEW PHASE DUE NOVEMBER ' 73
• RESULTS OF DATA ANALYSIS PHASE DUE FEBRUARY ' 74
• FINAL CONTRACTOR ORAL PRESENTATION APRIL ' 74
COORDINATION - EPA, REGION III HEADQUARTERS
Figure 15- ERTS — acid mine drainage.
-------
SURVEY AREA: WEST VIRGINIA - POTOMAC RIVER BASIN
TASK 1. DATA SURVEY (2 MONTHS)
a) REMOTE SENS ING-AIRCRAFT AND SPACECRAFT
b) IN SITU MEASUREMENTS
c) LAND RECORDS
TASK 2. DATA ACQUISITION AND ANALYSIS (4 MONTHS)
a) IDENTIFY LAND USE ACTIVITIES
b) LOCATE AREAS AND POINT SOURCES OF POLLUTION
I c) TOPOGRAPHY DISTURBANCES
* d) DRAINAGE PATTERNS AND POLLUTANTS
TASK 3. DEMONSTRATION PROGRAM PLAN (2 MONTHS)
a) OPTIMUM SURVEILLANCE PROGRAM
b) CONTRACTOR'S REPORT
Figure 16. Contract effort.
-------
H
H
CTi
U1
W. VA.
SCALE IN MILES
5 0 5 10 15 20
Figure 17. Mine drainage pollution study - north branch Potomac River.
-------
OBJECTIVE -
TO DEFINE AN APPROACH FOR AUTOMATING THE CONVERSION OF
IN-SmJ POLLUTION MEASUREMENTS INTO ENGINEERING UNITS AND
TRANSMITTING THESE DATA TO A CENTRAL DATA COLLECTION POINT
VIA THE ERTS DCP
STATUS -
CONTRACT STATEMENT OF WORK COMPLETE
H
H
OUTLOOK - • CONTRACT AWARD BY JANUARY ' 74
• COMPLETION OF SYSTEM DEFINITION TRADE STUDIES MARCH ' 74
• START SYSTEM CHECKOUT NOVEMBER ' 74
COORDINATION - EPA, RESEARCH TRIANGLE NERC
Figure 18. ERTS — data automation and transmission.
-------
SENSORS
SEVERAL SENSORS ARE REQUIRED TO DETECT POLLUTANTS AND MONITOR
METEOROLOGICAL CONDITIONS
DATA VALIDATION
ROUTINE CALIBRATION
DETECTION AND CORRECTION OF SENSOR FAILURE
CALCULATIONS
• CONVERSION AND AVERAGING FOR EPA STANDARD
I—I
H
^ DATA AVAILABILITY
• ON SITE
• AEROMETRIC DATA BANK AT EPA
• RESPONSE TO LOCAL AND STATE REQUESTS
DATA TRANSMISSIONS
• DISTRIBUTION TO LOCAL USERS
• TO CENTRAL FACILITY OF MULTISTATION NETWORK
• FROM REMOTE SITE
• TO EPA DATA BANK
Figure 19- Operational requirements for in situ air quality monitoring.
-------
H
H
CD
/ —
/
/
f
SENSOR
— — — — —
SIGNAL
TIONING
, •.
1
MANUAL RECORD.
CONVERT, REPORT.
VALIDATION,
ZERO/SPAN CAL
1 CHART
1 RECORDER
i
J
| PRESENT SYSTEM
INPUT
A/D CONVERTER,
SCANNER
CLOCK
TPROPOSED SYSTEM ADDITION '
LIFO
DATA STACK
I
TRANSMISSION
MODEM
DATA
HANDLING
SYSTEM
r
24v
ELECTRONIC
UNIT
POWER OFF
CLOCK
Figure 20. Data flow at camp and proposed system addition.
-------
THE USE OF NEAR-INFRARED REFLECTED SUNLIGHT
FOR BIODEGRADABLE POLLUTION MONITORING
Walter E. Bressette
NASA Langley Research Center
Hampton, Virginia
and
Dr. Donald E. Lear, Jr.
EPA Annapolis Field Office
Annapolis, Maryland
Presented at the EPA Second Environmental Quality Sensors Conference
Las Vegas, Nevada
October 10-12, 1973
II - 69
-------
THE USE OF NEAR-INFRARED REFLECTED SUNLIGHT
FOR BIODEGRADABLE POLLUTION MONITORING
Walter E. Bressette
NASA Langley Research Center
Hampton, Virginia 23665
and
Dr. Donald E. Lear, Jr.
EPA Annapolis Field Office
Annapolis, Maryland 21U01
BIOGRAPHICAL SKETCH
Mr. Bressette is a Senior Research Scientist with 25 years experience
in aerospace research, all at the Langley Research Center, including
research in the areas of visibility and thermal control of aerospace
vehicles, passive communications, geometric geodesy, solar energy col-
lection, and broadband photometric observation of satellite surfaces.
In recent years, he has applied his research skills and background
knowledge in scattering, absorption, and transmittance of electromag-
netic radiation to applications of remote sensing of water pollution
problems .
Dr. Lear, a graduate of the University of Rhode Island's Advance School
of Oceanography, is presently Chief of the Biology Section at the EPA
Annapolis Field Office. His responsibilities in EPA include surveil-
lance of the waters of the Chesapeake Bay and its tributaries. He has
had many years experience in sampling, analyzing, and reporting of
water quality conditions in the Chesapeake Bay area.
ABSTRACT
On October 2, 1972, a pattern of chlorophyll a_-containing phytoplankton
(algae) was detected from 3 kilometers altitude in a series of near-
infrared photographs of the Potomac River "Salt Wedge Area." Densitom-
eter traces over the film images, related to in situ measurements of
chlorophyll EL concentrations that varied from U to >3000 ug/i, revealed
a phytoplankton "bloom" threshold in the near infrared between the
concentration of 3k and 51
The photography also revealed bottom features through 2 meters of water
and made it possible to integrate chlorophyll a_ concentrations over a
l6-square-kilometer area to demonstrate this remote sensing technique
for biodegradable pollution monitoring.
L~9091 II - 70
-------
INTRODUCTION
The NASA Langley Research Center is working with the Environmental
Protection Agency, Office of Monitoring, in a joint program with the
purpose of developing the capability to determine synoptic pollution
levels and distributions of biodegradable pollutants through the use of
a. chlorophyll detection system. The objectives of the program are:
first, to determine if the spatial and temporal changes of the chloro-
phyll produced by the growth of phytoplankton (algae) can be measured
remotely, and second, to determine if these measurements can be
employed in an index of the stress on the ecosystem caused by bio-
degradable pollutants. The program will then develop a large area
chlorophyll detection system (aircraft and satellites are to be con-
sidered) to be coupled with in situ measurements to provide the
necessary synoptic observations required for biodegradable pollution
monitoring and control.
The remote sensing approach that we have chosen is broad-band optical
filtering of reflected sunlight from an altitude sufficiently high
so that a synoptic view can be obtained. It is envisioned that a
broad-band optical filtering system consisting of more than one
broad-band filter will be required to separate phytoplankton reflec-
tance from suspended sediment reflectance. At the present time
photographic film is being used as the sensor in order to develop
the optical filter system. However, in a biodegradable pollution
monitoring mode the developed broad-band optical filtering system
will be adapted to an electronic readout sensor; perhaps, solid state
readout by silicon diodes, thus providing a rapid readout system
most responsive to EPA's monitoring requirements.
THEORY
Sunlight reflected from plants that contain chlorophyll varies as a
function of wavelength as shown by the solid curve in figure 1.^ The
generally accepted color classification for the various wavelengths is
also shown along the abscissa for" comparison. The reflectance of
chlorophyll-containing plants is approximately 0.05 in the violet and
blue region, increases to 0.15 in the green region, decreases to 0.05
in the orange and red regions, and then increases abruptly to 0.5 to
0.6 in the near-infrared (NIR) region. Since the human eye is unable
to detect NIR, chlorophyll-containing plants are predominantly visible
in the green region of the spectrum. However, NIR sensors can detect
the spectral region where the high reflectance of chlorophyll-containing
plants results in the greatest reflected solar energy.
II - 71
-------
To detect phytoplankton floating below the surface of water, the absorp-
tion of light by the water must be considered. The dashed curve in
figure 1 is the exponential (1/2.7) absorption attenuation length of
sunlight in distilled water as a function of wavelength, as obtained
from reference 3. Thus, the penetration of sunlight in distilled water
is approximately 27 meters in the blue region of the solar spectrum,
decreases to 16 meters through the green region, and is a meter or less
in the NIR region.
Comparing the reflectivity of chlorophyll-containing phytoplankton with
the absorption of sunlight by water in figure 1 shows that when phyto-
plankton is within the upper meter or on the surface of the water, it
can be detected in the NIR region. It is also probable that the magni-
tude of the reflected solar energy in the NIR will also depend upon the
concentration of phytoplankton in the water as has been shown by Duntley
for the green and yellow regions.
EXPERIMENTAL AREA
On October 2, 1972, between the hours of 11:58 a.m. and 12:17 p.m. e.d.t.,
a photographic mission was flown at 3 kilometers altitude over the
Potomac River "salt wedge area," a distance along the river of approxi-
mately 30 kilometers by an NASA, Wallops Station, C-5^ aircraft. The
flight lines are shown in figure 2. The tidal action in this area, where
the salt water normally interfaces with the fresh water inflow, retards
the fresh water flow, deposits some of the suspended sediment, and
concentrates nutrient wastes from the sanitation plants located up river,
Since the late 1930's waste water discharges in the Washington metro-
politan area have increased the nutrients in the Potomac River - phosphorus
has increased tenfold, and nitrogen fivefold.5 These nutrients under
spring and summer conditions result in massive "blooms" of phytoplankton
(blue-green algae, primarily Anacystis cyanea), from Gunston Cove to
Maryland Point. Below Maryland Point and above the Route 301 Potomac
River Bridge the amount of phytoplankton decreases abruptly, primarily
because of increased salinity and a decrease in nutrients.5 On October 2,
1972, personnel from EPA, Annapolis Science Center, who obtained the in
situ water measurements of chlorophyll a_, light penetration, salinity,
etc. (tabulated in Table l), reported that in some places heavy concen-
trations of phytoplankton were visible on the water.
EXPERIMENTAL METHOD
The Wallops aircraft contained a bank of four Hasselblad cameras.
Pertinent information concerning cameras, film, filters, and exposure
are listed in Table 2. Three of the Hasselblad cameras were equipped
with the Wratten filter selections shown in figure 3 where transmit-
tance as a function of wavelength is plotted for the three filters.
The number 58 filter transmits reflected sunlight only under the curve
in the blue-green-yellow region. The number 12 filter transmits
reflected sunlight under the curve in the green-yellow-orange-red
region with the cutoff in the red determined by the film used (Table 2).
The number 89B filter transmits reflected sunlight under.the curve in
the NIR region with the cutoff again determined by the type of film
used (Table 2).
II - 72
-------
Black and white film, which was used in this experiment, was flown with
all three filters. In addition, a HasseTblad camera was used with con-
ventional color film and a haze filter to provide color reference. The
film-filter systems just described were selected to isolate the green
and NIR reflectance of phytoplankton as identified in the theory section
of this paper as well as to identify other color features such as the
red Maryland clay on the river bottom. The four cameras were synchro-
nized; this resulted in four sets of photographs (28 photographs per set).
RESULTS AHD DISCUSSION
As predicted from our preliminary study, the photographs taken through
the NIR filter on black and white film revealed many features in or on
the water that contrast strongly with the background of the water as
shown in figure U. Pictures 1 through 7 which show many features in the
water are from the flight line of the upper reach of the river where the
in situ data indicated very high chlorophyll a_ concentrations and very
low salinity. Pictures 9 through 15 are from the flight line over the
"salt wedge area," and as can be seen in pictures 10, 11, and 12, the
features are concentrated on the side of the river opposite the salt
water intrusion for the flood tide conditions existing at the time of
these pictures. Along the flight line over the lower reach, pictures IT
through 22, where the salinity is above six parts per thousand, the
features are absent and the water is uniformly black, with some increased
radiance in places from sun-glint, shallow bottom areas, and powerplant
smoke. The consistency of the quantity of features within the fresh
water area and the" very high in situ measurements of chlorophyll a_ in
the same area identify the features in the water in pictures 1 through
Ik as chlorophyll a_-containing phytoplankton.
The positive prints that are seen in figure 4 were obtained by over-
development from a normally developed negative that, in turn, was obtained
with a camera setting that was one full stop open from the recommended
Kodak opening for the camera shutter speed, sun angle, and altitude of
the flight mission. As you can see in all the prints, the solid white
areas, which are land, are considerably overdeveloped, washing out the
land features normally emphasized in standard processing procedures. The
best exposure (which controls the ratio of density or contrast)? for detec-
tion of the chlorophyll a_-containing phytoplankton in the water cannot be
determined from just one set of photographs. However, overdevelopment
of positive transparency and prints is essential, as can be seen by com-
paring figure k with figure 5- In figure 5 are presented essentially
the same pictures as in figure k, only the development time of the posi-
tive prints is normal, resulting in the definition of many land features.
However, in these prints the water is essentially black; only the area
where surface phytoplankton was reported and very high chlorophyll a_
concentrations measured, reflected sunlight.
Figure 6 includes a densitometer trace across print number 10 of
figure k, and its location is shown by the arrows on an identical print
included on the right of this figure. The densitometer trace, which is
the solid line, is presented relative to the density trace over the
-------
unexposed portion of the film (R) on the vertical scale and on the hori-
zontal as distance from one edge of the river (X), in kilometers. In
the figure the unexposed film is identified as A, the land area as B,
and the water areas as C. As can "be seen, the difference in R between
B and A is large compared to the difference in R "between C and A, and
between C and the radiance from the chlorophyll a-containing phytoplankton
in the majority of places. Therefore, since development time increases
the difference in film density or contrast,' to enhance the phytoplankton
radiance relative to the water, it is necessary to overdevelop the posi-
tive transparencies and prints - thus overexposing the highly radiant
land area, but not the highly absorbing water area at the NIR wavelengths.
Also presented in figure 6, as the dashed curve, is the depth of the river
bottom in meters, obtained from a marine map, plotted against distance
across the river in kilometers. 'The purpose of this curve is to establish
that the river between 0 and 1 kilometer, where there is a broad increase
in R on the densitometer trace, labeled bottom radiance, is approximately
2 meters deep. The justification for concluding that the increased
radiance in this area is due to sunlight penetration to, and reflected
from, the bottom can be established by comparing the two simultaneous
pictures on the right of figure 6. In the upper portion of the top
photograph, just below the land, taken through the NIR filter that passes
only red and infrared radiation, there is an increase in radiance as
evident by the light area, while the same area of the bottom photograph,
taken through the number 58 green filter that passes only green radiation,
there is a decrease in radiance as evident by the area being dark in
relation to its surrounding area. Therefore, the increase in radiance
in the NIR photograph in this area cannot be due to chlorophyll a-
containing phytoplankton, nor is it due to white light (foam), because
the radiance is not increased in the same area of both the green and NIR
photographs, and must be due to either red or infrared radiation.
Dr. Lear has observed the bottom as being hard, red Maryland clay.
R obtained from densitometer traces of NIR photographs (Table l) like
the one seen in figure 6, over each area where the in situ chlorophyll a_
measurements were made, is plotted in figure 7, represented by the
circles, against the measured chlorophyll a_ concentrations. The solid
curve is faired through these data points. Also shown in figure 7 is
the visibility depth (S), in meters, as seen by the eye, of a 30-cm
white disk (Secchi disk) that was lowered into the water at each of the
data stations. The dashed curve is a fairing through S, represented by
the squares, that were also plotted against measured chlorophyll a_ con-
centrations. The dashed curve consists of two straight lines intersect-
ing at the chlorophyll a_ concentration of ^0 ug/£ which you can see fits
the data reasonably well.
Observation of the faired values of R between the values of chlorophyll a_
concentrations of k and 3^ yg/& indicates that in the NIR R is not sensi-
tive to chlorophyll a_ concentration over this range of concentrations.
However, the faired values of S over this same range of chlorophyll a_
concentrations indicate that the light penetration into the water is
decreasing with increasing chlorophyll ^concentrations. Thus, the
II - 74
-------
absorption of light with the increase in chlorophyll *i concentrations
over the range of *i to 3^ yg/£, reduces the photic zone in the water,
but does not affect the upwelling NIR radiation back through the sur-
face of the water. Since the upwelling light back through the surface
of the water is not influenced in this range of chlorophyll a_ concen-
trations, observations of the back-scattered light from scattering of
the atmosphere and reflection off the surface of the water as a function
Pt ' O
of sun altitude angle0'" can be made, providing other contaminants in
the water (such as suspended sediment) do not cause upwelling radiation
through the surface of the water in the NIR wavelength.
The very low value of R in figure 7 over the range of k to 3*+ ug/Jt of
chlorophyll a_ concentrations suggests that suspended sediments are not
significantly contributing to the upwelling light even though it is
common knowledge that the Potomac River carries a heavy sediment load.
This sediment load is confirmed by S of this figure which has a maximum
penetration into the water of only l.U meters. However, it must be
remembered that S was obtained with the human eye which responds best
at the green wavelength of light (the attenution length of green light
in distilled water approaches 20 metersS), and since the reflection
from suspended sediment is a function of both quantity and particle
size-1-^ it could be possible that the size of the sediment particles in
the upper 2 meters of the river, for the river flow conditions on the
day of this mission, are below the threshold value for upwelling of
sunlight at the NIR wavelength. Therefore, the absence or the detection
in the NIR of suspended sediment when chlorophyll a_ concentrations are
below *JO ug/£ could provide information at one of the two wavelengths
required for the measurement of suspended sediment.^ The other wave-
length would be around 523 nanometers where laboratory measurements
show the reflectance of chlorophyll a_ to be independent of concentra-
tions between 1 and 30 yg/£ as it is in the NIR in this experiment.
Above chlorophyll a_ concentration of 3^ pg/S- in figure 7, R increase
and the S decrease, if any, is slight. Thus with chlorophyll a_ concen-
trations above 3^ Mg/£, the situation is reversed, with the photic
zone essentially constant and the upwelling light increasing. The
transition point cannot be precisely determined from this experiment
because of lack of data in the critical range of chlorophyll a_ concen-
trations , but it is defined as being between 3k and 51 Vg/& and is
labeled "threshold for phytoplankton 'blooms'" on the figure.
The mechanism for the reversal of R and S cannot be determined from the
data of this photographic mission, nor can the percentage of upwelled
light indicated by the increase in R that would have been used for
photosynthesis by the phytoplankton.
The values of R from a, series of densitometer traces can be employed
to map the chlorophyll a_ concentrations in a "bloom" area as seen in
figure 8 for a ^-kilometer square of the river. As you can see in
figure 8, chlorophyll a_ concentrations are very erratic, and would be
impossible to describe or even average correctly with the limited ground
II - 75
-------
truth data obtained during this photographic mission. However, it is
possible to integrate the densitometer data that were used to construct
figure 8 to determine the percent area coverage of various incremental
ranges of concentration of chlorophyll a_ as shown in Table 3. Table 3
shows that only 23 percent of the water area in figure 8 contains chloro-
phyll ji concentrations that are below the threshold concentration for
producing phytoplankton "blooms." The largest percent of "bloom" area
is between the chlorophyll a_ increment of kQ to 60 ug/fc with the "bloom"
area of each chlorophyll a_ concentration increment progressively reduced
with increasing chlorophyll a_ concentrations. Information of the type
shown in Table 3 over a period of time would be strong input for monitor-
ing the stress of a body of water from biodegradable pollution.-'-
CONCLUDING REMARKS
On October 2, 1972, a photographic flight, at 3 kilometers altitude,
using Hasselblad cameras, 'was flown over the Potomac River "salt wedge
area." During the same time in situ water measurements detected chloro-
phyll £i concentrations that varied from k yg/£ to >3000 ug/£.
From an analysis of photographs taken through the 89B near-infrared (NIR)
Wratten optical filter and a selected picture taken through the number 58
Wratten green filter, along with corresponding ground truth data, the
following conclusions can be made:
1. ' Overdevelopment of positive transparencies and prints from a negative.
that was overexposed in relation to the Kodak recommended normal
exposure setting through an NIR filter greatly enhances the contrast
between phytoplankton "blooms" and the water background.
2. The penetration of sunlight into water can be photographed through
an 89B Wratten NIR filter to reveal bottom features as deep as
2 meters .
3. Using an NIR camera system, the threshold for phytoplankton "blooms"
is between the chlorophyll a_ concentrations of 3^ and 51 ug/&.
k. The upwelling of NIR light through the surface of the water is inde-
pendent of chlorophyll a_ concentrations below a value of 3^
5- The penetration of sunlight into the water as determined by the eye
from lowering of a 30-cm white disk (Secchi disk) into the water
decreased with increasing concentrations of chlorophyll a_ below
the concentrations of 3^ to 51 yg/£5 and remained nearly constant
with further increase in chlorophyll ^concentrations.
6. Detection of upwelling light in the NIR coupled with detection near
523 nanometers has a potential for measurement of suspended sediment
when chlorophyll a_ concentrations are below 3^
II - 76
-------
7. Synoptic pictures in conjunction with limited ground truth data
makes detection and mapping of chlorophyll a_-containing phyto-
plankton "blooms" possible over large areas of water - thus pro-
viding a strong input for monitoring the stress in a body of
water from biodegradable pollutants .
REFERENCES
•k)dum, E. P., "Fundamentals of Ecology," Third Edition, W. B. Saunders
Company .
Katzoff, S. , "The Electromagnetic-Radiation Environment of a Satellite,
Part 1, Range of Thermal to X-Radiation," NASA TN D-1360, 1962.
, F. N. , "Oceanic Environment," Chapter II, Hydronautics, Edited
by Sheets, H. E., and Boatwright, V. T. , Jr., Academic Press, 1970.
^Duntley, S. Q. , "Detection of Ocean Chlorophyll From Earth Orbit,"
Section 102, Hth Annual Earth Resources Program Review, Volume IV,
Presented at the Manned Spacecraft Center, Houston, Texas, January 17-21,
1972.
o, J. A., Jaworski , N. A., and Lear, D. W. , "Current Water Quality
Conditions and Investigation in the Upper Potomac River Tidal System,"
Technical Report No. Ul, Chesapeake Technical Support Laboratory, Middle
Atlantic Region, Federal Water Quality Administration, U.S. Department
of Interior, May 1970.
Eastman Kodak Company, "Kodak Wratten Filters, for Scientific and
Technical Use," Twenty-second Edition, First 1966 Printing.
?Mack, J. E., and Martin, M. J. , "The Photographic Process," First
Edition, McGraw-Hill Book Company, Inc., page 211.
8Piech, K. R., Walker, J. E. , "Aerial Color Analyses of Water Quality,"
Proceedings of the American Society of Civil Engineers, Volume 97,
No. SU2, November 1971.
9payrie,TR. E. , "Albedo of the Sea Surface," Journal of the Atmospheric
Sciences, Volume 29, July 1972.
•^Williams, Jerome, "Optical Properties of the Sea," United States Naval
Institute, Annapolis, Maryland, 1970.
II - 77
-------
Table 1. EPA, Mid-Potomac Estuary Phytoplankton Surveillance
Data, Surface Samples, October 2, 1972
H
I
-J
03
Sta. No.
1
2
3
1*
5
6
7
8
9
10
Location^
N 32
Below Br.
West Side
Buoy CN
Nun 25P
Popes Cr.
C 3
C 7
C 9
N10
Nl6
C19
Time 3
ll*15
ll*08
11*03
1355
13^6
1337
1328
1320
1306
1253
Salinity
Cond.
Tide ymhos °/oo
Slack 12.1
Slack 10.5
Slack 10.5
P 10.3
F 9.8
F 8.6
F 8.0
F 7-1
F 5-7
F 1*.2
7-5
6.6
6.1*
6.7
6.0
5.U
1*.8
1*.U
3.1*
2.7
Temp.
°C
21.0
20.5
21.U
21.1*
21.3
21.1*
21.2
21.2
21.7
20.9
DO
mg/£
5-8
6.0
6.1*
6.0
—
6.5
6.9
7.0
7.0
Secchi1*
Disc-m.
1.1*1*
1.52
1.80
1.18
.67
1.23
.93
• 95
-77
.82
Chloro-
phyll a
ygm/£
007.5
930.0
31*. 5
16.5
10.5
l+.O
10.0
9.0
332.0
6.0
Wind
dir.
N
N
N
N
N
N
N
N
N
S
Wind
vel. , See
mph state
0-2
0-2
3-6
3-6
3-6
3-6
3-6
3-6
3-6
2-5
Calm
Calm
Ripples
Ripples
Ripples
Ripples
Ripples
Ripples
Ripples
Ripples
R5
0.9!*
-7U
.87
.78
.90
-77
.78
.80
1*.58
.91*
EPA, Annapolis Science Center, Annapolis, Maryland 211*01.
^Coast and Geodetics Survey Navigation Map #559-
^Eastern daylight time.
The visibility depth of a 30-cm-diameter white disk.
^Relative reflectance, as determined by difference between light transmittance of densitometer traces
over the in situ data point location on the film relative to traces over the unexposed portion of the
film.
-------
Table 1. EPA, Mid-Potomac Estuary Phytoplankton Surveillance
Data, Surface Samples, October 2, 1972 - Concluded
H
H
Sta. No.
11
12
13
lH
15
16
17
18
19
20
Location2 Time^
Buoy 22
Off
Potomac
N26
Off
Acquia
N30
N3H
Off Va.
Shore
NHO
Mallow1
CHl
Buoy HH
12H2
1230
Cr.
1222
1210
Cr.
1201
115H
11H6
1137
s Bay
1130
1120
Tide
F
F
F
F
F
F
F
F
F
F
Cond.
umhos
3.20
2.75
2.H5
1.85
1.80
1.25
1.00
.75
.Ho
• .25
Salinity
°/oo
2.1
1.7
1.5
1.2
1.0
.85
.6 •
-5
.2
.1
Temp.
°C
20.7
21.3
20.5
21.3
20.2
21.0
21.1
20.8
20.7
20.65
DO
7-5
7.6
7-5
7-5
7.5
7-H
7. It
7.1
6.7
7.3
Secchi1*
Dis-m.
0.82
.H6
.59
.72
.62
.67'
.6H
• 72
.72
.59
Chloro-
phyll a_
57
>3128
783
1192
79
328
51
559
76
27 H
Wind
dir.
S
S
S
S
S
S
S
S
S
S
Wind
vel. ,
mph
2-5
2-5
2-5
2-5
2-5
2-5
2-5
2-5
2-5
2-5
Sea
state
Ripples
Ripples
Ripples
Ripples
Ripples
Ripples
Ripples
Ripples
Ripples
Ripples
R5
_____
18. HO
6.2H
5.80
2.38
H.30
1.56
6.19
2.80
H.52
-'-EPA, Annapolis Science Center, Annapolis, Maryland 2lUoi.
^Coast and Geodetics Survey Navigation Map #559-
^Eastern daylight time.
The visibility depth of a 30-cm-diameter white disk.
^Relative film transmittance, as determined by difference between light transmittance of densitometer
traces over the insitu data point location on the film relative to traces over the unexposed portion
of the film.
-------
Table 2. Sensor Complement and Camera Settings
Camera
1. Hasselblad
2. Hasselblad
H 3. Hasselblad
I
g 1*. Hasselblad
Focal
length
(mm)
1*0
1*0
1*0
1*0
Filter1
58 (green)
12 (yellow)
89B (NIR)
HF-5 (haze)
Film
format
(mm)
TO
TO
TO
TO
Film type
21+02 Black &
White
2i*02 Black &
White
2l+2l* Black &
White NIR
SO-39T Color
AEI3
6
1*0
28
12
Speed14
(sec)
1/250
1/250
1/250
1/250
f ^
number
5.6
11
8
5.6
Kodak Wratten filter number.
2
Kodak film number.
-'Kodak recommended aerial exposure index.
4
Actual exposure.
-------
Table 3. Area Distribution of Chlorophyll a_ Concentrations
From Integration of Data for Figure 8
Chlorophyll a_ concentrations, Area covered,
percent
23
UO to 60 36
60 to 100 18
100 to 175 6
175 to 300 5
300 to U60 3
>1000 6
II - 81
-------
H
H
CD
K>
1/e ATTENUATION IENGTH OF DISTILLED WATER
(REF. 3)
REFLECTANCE
REFLECTIVITY OF
CHLOROPHYLL
PLANTS (REF. 2)
30
20
ATTENUATION
LENGTH , m
10
.4 .5 .6 .7 .8 .9 1.0 1.1 1.2
COLOR-| V B G Y 0 R- -|R
WAVELENGTH , M
0
Figure 1. Comparison of the reflectance of chlorophyll-containing plants with the attenuation
length of sunlight in distilled water.
-------
N
oo
u>
GUNSTON COVE
MARYLAND POINT
tr SALT WEDGE
AREA"
ROUTE 301
BRIDGE
SEWAGE TREATMENT PLANTS
BLUE PLAINS
ARLINGTON
ALEXANDRIA
FAIRFAX-WESTGATE
FAIRFAX-LITTLE HUNTING CREEK
FAIRFAX-DOGUE CREEK
PISCATAWAY
ANDREWS AIR FORCE BASE -NO. 1&4
FORT BELVOIR - NO. 1 & 2
PENTAGON
INS ITU DATA POINTS
FLIGHT LINES
Figure 2. Normal location of Potomac River "Salt Wedge Area" relative to location of in situ
data points, flight lines, and sanitation plants.
-------
H
H
1
00
TRANSMKTANCE
100
80
_ FILM
CUT OFF~*j
,<-— "1
1 <"> I~ IITFTI /
60
40
20
0
COLOR
58 FILTER
89B FILTER
1R FILM CUT OFF
4
v 1
B
.5
1 G
.6
i Y
0
.7
R
.8
-IR
.9
1.0
1
.1
1.
WAVELENGTH,
Figure 3. Filter selection and film cutoff for detection of phytoplankton.
-------
H
I
00
Ul
17
21
22
Figure k. A series of photographic prints along the Potomac River from Possum Point to Bluff
Point, taken through the number 89B near-infrared Wratten optical filter, from 3 kilometers
altitude.
-------
H
H
I
CO
26
Figure 5. Essentially the same series of pictures as in figure k only the development time
for the positive prints is reduced in order to define the land features.
-------
cc
-0
14
12
10
8
R
4
2
0
\ __
i
i
i
i
B
PLANKTON -
RADIANCE
• C -
BOTTOM
RADIANCE
B
0
2 "A"
4 B
6 C
DEPTH,
8 meters
10
12
14
X, km
DENSITOMETER TRACE
RIVER BOTTOM DEPTH
UNEXPOSED FILM
LAND AREA
WATER AREA
89B NIR
FILTER
DENSITOMETER TRACE
58 GREEN
FILTER
Figure 6. Densitometer trace across the number 10 near-infrared (NIR) photograph of figure
showing radiance of land, water, phytoplankton, and the river bottom, relative to the
unexposed film, in the form of relative film transmittance, R, and river bottom depth
versus distance across the river.
-------
H
H
C»
00
10 r-
R 1
D
D
D
ooo
-.-B—a :
H k-
THRESHOLD FOR
PHYTOPLANKTON "BLOOMS"
10
METERS
-J .1
10 100
CHLOROPHYLL a CONCENTRATIONS, ptg/*
1000
Figure 7- Relative film transmittance, R, from densitometer traces and 30-cm Secchi disk
depth, S, versus measured chlorophyll a concentrations.
-------
H
H
I
03
DISTANCE Y, km
CHLOROPHYLL a. ,
1 2 3
DISTANCE X, km
c-1000
500
r- 100
50
10
Figure 8. Distribution of chlorophyll a_concentrations over 16 square kilometers of the Potomac
River obtained from densitometer traces of picture 6 in figure 4.
-------
SESSION III
AIR QUALITY. SENSOR DEVELOPMENTS
CHAIRMAN
MR. CHARLES E. BRUNOT
OFFICE OF MONITORING SYSTEMS -, OR&D
-------
ST. LOUIS
REGIONAL AIR MONITORING SYSTEM
JAMES A. REAGAN
Presented at the
Second Environmental Quality Sensor Conference
October 10-11, 1973
NATIONAL ENVIRONMENTAL RESEARCH CENTER
LAS VEGAS, NEVADA
III - 1
-------
The St. Louis Regional Air Monitoring System
James A. Reagan
ABSTRACT
The Regional Air Monitoring System (RAMS) consists of 25 sta-
tions observing gaseous, particulate, meteorological, and radia-
tion parameters. Distribution of the stations is on four concen-
tric circles, centered on the arch, and at distances of U, 9, 20,
and 1*0 kilometers. The system is designed for a 90% data capture,
excluding scheduled calibration. Scheduled maintenance was designed
in conjunction with remote calibration capabilities to reduce op-
erational manpower requirements.
Parameters monitored at one or more stations include: sulfur
dioxide, hydrogen sulfide, total sulfur, carbon monoxide, total
hydrocarbons, methane, nitrogen dioxide, nitric oxide, ozone, visi-
bility, wind speed, wind directions, temperature, relative humi-
dity, pressure, temperature differential, solar radiation, three
component wind speed, and particulate loading. Each station has
independent operational capability, including in situ recorders.
Stations are linked via a telephonic communication network to a
data center. Routine monitoring of the data stream is made to
flag observational or system monitoring anomalies.
The system is designed to be a research tool, implying a
high quality of data, yet fabricated and operated at minimal ex-
pense. Technology spinoff.will include system design specifica-
tions, performance specifications, operation, and troubleshooting
III - 2
-------
manuals for system components. Data from the network will "be
available to requestors through a convenient distribution chan-
nel.
Ill
-------
INTRODUCTION
The Regional Air Pollution Study or RAPS will develop mathe-
matical simulation models of atmospheric processes affecting the
transport and concentration of air pollutants. Extensive source
and receptor data will be collected. Additional coordinated ex-
periments to elucidate interim species, reaction rates, and energy
"balances will "be performed.
Any of these models may "be represented by the relation of the
change in time with the change in space equals the losses and addi-
tions from reactions and sources and due to diffusion.
This vector differential has many submodels which require develop-
ment .
.Figure 1, displaying the schematic of the model, is a diagram
of the overall system. Sulfur and particulates primarily from sta-
tionary sources, along with nitric oxide and hydrocarbons from mo-
bile sources, are transported through the atmosphere. In route they
undergo a series of complex reactions with one another resulting in
new species formation and alteration of existing ones. Ambient air
quality is translated into effects on the people and the environment.
Ill - 4
-------
H
H
H
I
(Jl
SOURCES
Fuel Combustion
Industrial
Steam Electric
Residential
Transportation
Road- Vehicles
Aircraft
Railroad
Vessels
Solid-Waste Disposal
Process Losses
TRANSPORT
Diffusion
Portage
Turbulence
TRANSFORMATIONS
Solar Radiation
Chemical Reactions
Decay
AiR CLUALiTY
Ambient Levels
Sulfur Dioxide
Aerosols
Carbon Monoxide
Nitric Oxide
Hydrocarbons
Ozone
Nitrogen Dioxide
EFFECTS
Health
Visibility
Flora
Fauna
Materials
Welfare
Figure 1
-------
The RAMS measures portions of each box except Sources. Mete-
orological measurements give data for the characteristic transport.
Incident solar energy of the right wavelength drives some of the
critical chemical transformations. All of the pollutants are moni-
tored by appropriate air quality sensors. Resultant effects on
visibility are also measured. This system then is designed to pro-
vide many direct measurements on the model parameters.
GENERAL SYSTEM
RAMS consists of 25 stations situated throughout the St.
Louis area. Each station has a full compliment of basic air qual-
ity and meteorological monitoring equipment. They are linked to-
gether via a telecommunications network, and all data are sent to
the central data acquisition system on a one minute polling frequency.
Several stations have special sampling equipment which is placed
because of site or area considerations.
The system is designed to achieve a 90 percent data capture
rate. Scheduled maintenance is excluded from this. We felt that
the equipment is not state-of-the-art, albeit close to that. There-
fore, the system was awarded on a performance contract basis as long
as certain standards of technical excellence were met or exceeded.
Currently, the schedule calls for the first station to be delivered
in late January 197^, with the remainder following at a rate of one
per week.
SITING CRITERIA
Dr. Francis Pooler, in his paper on the "Network Requirements
for the St. Louis Regional Air Pollution Study,"2 discusses the
III - 6
-------
number and position of stations with the guidelines of the study
objectives, as follows:
1. Extensive spatial coverage is necessary for including
the entire metropolitan area.
2. For large networks the marginal increase in know-
ledge with the addition of each station declines after
the nth station.
3. Simultaneous measurements should be made at all sites
for a long, enough interval of time to include all the
varying weather regimes.
U. Station costs become fixed after about ten stations.
Most manufacturer discounts do not decrease after the
10th item.
5. The same measurements should be made at all sites.
6. Minimize labor costs, even if reasonable increased
capital costs are initially incurred.
T. Colocate meteorological measurements with air
quality measurements for convenience.
We settled on a subjective number of 25 stations as being ade-
quate to supply our needs. This was reached by considering four con-
centric circles of six stations each around a central station. This
has been modified as shown on the map.
To arrive at the final network layout, several factors were
considered:
1. At least one station should be in the upwind sector,
regardless of the wind direction.
2. Comparison of upwind-downwind measurements is de-
sireable; therefore, each upwind site should have a
downwind counterpart.
3. Place sites so that the station density is roughly
proportional to pollutant gradient.
k. Incorporate existing monitoring efforts as much as
is feasible.
Ill - 7
-------
The first criteria can be met by placing four rural sites at
approximately 90 degrees azimuth spacing. Existing data show that
the gradient slackens off out to 10 kilometers from the Arch, with
most concentrations occuring in that area. The St. Louis City/
County network and the Illinois EPA network constitute most of the
existing surveillance system. Therefore, the rings were set at
distances of U, 9, 20, and ^0 kilometers from the central station.
Actual station location is within less than a kilometer for the
center city stations, increasing to a 5-kilometer area on the outer
ring. Density was also increased in the third ring to eight stations
since most of this ring is in the heavy surburban area. The four
outer stations are designed to monitor background levels.
Several specific criteria must be followed as much as possible.
1. The site should be representative of a reason-
ably large area.
2. The site should be 1 kilometer or more from a major
traffic artery.
3. Site should not be in the lee of a building.
k. There should be no significant obstruction to the
air flow higher than one-tenth the distance to the ob-
struction from the point of measurement.
5- Terrain should be representative of the surrounding
area that is not in a depression.
6. Power and communications should be available, for
they can be very expensive to bring any great distance.
Locations for sites were defined to allow the chance of finding in-
expensive, acceptable sites. An absolute coordinate location is
impossible to achieve in an urban area.
Ill - 8
-------
STATIONS
Each station is a 10-by-l6-foot shelter mounted on concrete
piers. A free standing 10 or 30 meter tover is adjacent to the
station on the northerly side. The shelter is of prefab construc-
tion with white walls and a flat roof with steel mesh covering to
allow roof access and utilization. A four-ton cooling system with
good internal air circulation is provided to maintain a constant
temperature of 72° ± 3°. Instrument heat is sufficient to require
cooling to temperatures near the freezing point of water. Dual
exits are provided for quick exit as well as an automatic heat sens-
ing and Freon fire extinguishing system. Air pumps are mounted in
a sound-proofed box which doubles on top as bench space. Gas
cylinders are placed in a fire-proofed compartment with forced air
draft to prevent the build-up of explosive gas concentrations. Hy-
drogen is provided to the instruments by means of a catalytic
separator in lieu of pressure bottles.
The air quality sensors 'have been provided with zero and span
gas systems capable of remote actuation. This is an example of a
capital equipment expenditure which should pay for itself within one
year by reduced personnel costs. Critical air flows, power levels,
range settings, and calibration status are monitored to assure
proper component performance.
Air quality instruments were selected to:
1. Detect levels below normal ambient level minima.
2. Have noise generally less than one percent of full
scale values.
3. Have zero and span drift no worse than one percent
a day or 3 percent in three days.
Ill 9
-------
k. Operate without failure for more than a week.
These specifications are general, and each instrument has a speci-
fic set applicable to it.
Instruments selected are not on "board, so they must still be
considered tentative. They include good analyzers, performance and
price considered, but are not necessarily the only instruments hav-
ing good performance, low price, or both. The choices certainly
cannot constitute an Agency endorsement, since we have not had an
evaluation of the application of Finagle's laws to this system.
Each station will have an ozone monitor based on the chemi-
luminescent method. The probable instrument is the Monitor Labs
8U10.
Each station will have a NO box capable of direct WO and NO
-*C X
determination and KOp inference. The probable instrument is the Moni-
tor Labs 8S*UO.
Each station will have a total hydrocarbon, methane, carbon mon-
oxide monitor based on chromatographic separation. The probable in-
strument is the Beckman 6800.
Twelve stations will have a total sulfur monitor based upon
flame photometric detection. The other 13 stations will have a sul-
fur gas chromatograph capable of total sulfur, hydrogen sulfide, and
sulfur dioxide. Most station locations should have only sulfur di-
oxide impinging upon them. This is due to the primarily local source
nature of any other sulfur gas. Since the total sulfur analyzer is
cheaper by half than the chromatograph, as well as more dependable,
we split the stations in half for these measurements. Chromatographic
III - 10
-------
analyzers will be placed where other sulfur gases maybe expected or
are known to be present. The probable total sulfur analyzer will be
the Meloy SA 185, and the Chromatograph, the Tracer 270 HA.
Each station will have a nephelometer for visibility and fine
particulate determination. This will be the MRI 1550, the only one
on the market.
Each station will have two gas bag samplers for the deter-
mination of hydrocarbon concentrations by laboratory chromato-
graphs. If the bag material is suitable, they may be used for the
collection and analysis of tracer compounds such as sulfur hexa-
flouride and freon 11.
All meteorological instruments will likely be provided by
MRI with one exception.
Each station will have wind speed, direction, and ambient tem-
perature.
Twelve stations will have a temperature differential between
ground level and 30 meters.
Barometric pressure will be measured at seven stations, es-
sentially the outer perimeter and -one diameter through the system.
Dew point will be measured at each location, probably with the
Cambridge 880.
All meteorological instruments were ordered with the pro-
vision that they meet standard catalog specifications. The deter-
mination of relative humidity or dew point is considered to be es-
pecially important since this heavily influences aerosol formation.
Solar radiation measurements will be made at up to six stations
to determine the incident solar output. All equipment will probably
III - 11
-------
III - 12
-------
be provided by Eppley and have common batch optical glass filters.
Measurements made include total incident radiation, total 300-
395 nanometer radiation, and total 300-695 nanometer radiation us-
ing pyronometers. Pyrheliometers with filter wheels consisting of
filters having no cutoff, 395 nm cutoff, It75, 530, 570, 630, 695,
and 780 nm cutoffs. Pyrogeometers measuring 3-50 urn will also be
used.
Spare parts for the instruments were established as spare in-
struments in the amount of 10 percent for all instruments except the
chromatographs, for which 20 percent will be ordered.
The outputs from each sensor are converted by an Xincom A/D
converter to digital input to a PDP-8/M computer with l6K core, Per-
tec magnetic tape drive 9 channel, 800 bpi density and ASH-33 tele-
type. Addition of new equipment is easy because f the extra control
and data sampling capabilities of the data acquisition system. Data
is continuously recorded in situ, allowing the telecommunications link
to be lost with no loss of data except in real time.
CENTRAL FACILITY
A PDP-11 AO computer with. 32KL core, three tape drives, three
1.2 megaword dishes, console, line printer, card reader, graphics
printer/plotter and CRT is the base of the data acquisition system. A
PDP-11/05 front end processor handles station telecommunications
while the foreground of the 11/40 handles peripheral devices and th.e
Disk Operating System as kept in background.
Ill - 13
-------
Each station/instrument is polled each minute, and a one-
minute average is transmitted to the central computer. Each station
samples once each half second and stores the sum until converted
and transmitted. Raw voltages are written to magnetic tape, which
is subsequently processed to produce a tape with validated one-
minute values. Hourly averages are formed in the computer and
kept for up to a month on the random access storage.
The primary purpose of the central computer is to handle
large volumes of data from the network, as well as other sources, such
as special mobile samplers or aircraft. It can be used to validate
and get a quick look at data, but it is not for the testing of models
or any large scale correlation work. It is intended that the total
data base will be accessible by the Univac 1110 system which will be
at the Eesearch Triangle Park.
SUMMARY
RAMS is a portion of the overall Regional Air Pollution Study.
It is designed to be a research tool and not an enforcement moni-
toring network. The contractor selected for fabrication and opera-
tion is the Rockwell International Science Center of Thousand Oaks,
California. Anyone desiring detailed descriptions and specifica-
tions should contact me at the EPA office in St. Louis.
Ill - 14
-------
BIBLIOGRAPHY
Allen, Philip W. Regional Mr Pollution Study - An Overview. APCA
Annual Meeting, Paper No. 73-21.
Pooler, Dr. Francis. Network Requirementsfor the St. Louis Re-
gional Air Pollution Study.
Ill - 15
-------
\ J\
L
RAPS/RAMS PROPOSED STATION LOCATIONS
H
M
M
SIAIIWS Mtt AMAKWD tH COKEXTMt CtKUS
UOWD STATIC* 18) AT OlStMCTS OF *. ».
K). MO 40 KlLOlCTERS. THE SIMS Ml AT
n*t WPIOHWU CCKTCR or w *H* WOE*
CWSIKWT1W MWIKC fBOH LESS IHM A
K1LOKTER-& HAOlUi AT 1« ONTIW. S1ATIOH
TO AKVI A S KliaCUII UOIUS AMXMD IW
OUII* ClJttU OF « SIATIWS,
-------
LONG PATH OPTICAL MEASUREMENT OF
ATMOSPHERIC POLLUTANTS
Material prepared for discussion with EPA Regional
S & A Representatives, NERC-LV, October 10-11, 1973
By
Andrew E. O'Keeffe, Chief
Air Quality Measurement Methods Branch
Chemistry & Physics Laboratory
NERC-RTP
At the time when this symposium was first being organized
early this summer, John Koutsandreas asked me to participate by
describing that work of my branch related to remote sensing of
pollutants. I promptly breathed a sigh of relief and explained to
John that my branch, being totally concerned with the measurement
of pollutants in ambient air, has no work in progress aimed at
remote sensors. I went on to explain that we have given considerable
thought to the concept and have concluded that, since there exist no
physical or legal restraints upon our access to the air around us,
there is no demonstrable need for sensing its pollutant content
remotely.
It turned out, however, that my sigh of relief was premature.
It seems that the symposium organizers were aware of certain of
our ongoing work and felt that it was within their sphere of
interest. Naturally, I felt complimented by their interest, and
willingly joined in a semantic discussion that soon led to a
redefinition of remote or non-contact sensors. They are, for the
purposes of this paper, those devices that accomplish the measurement
of an atmospheric pollutant without physically transferring a sample
of the air to an observation chamber within the instrument. In
order to clarify this definition and thus avoid later confusion, let
us look at two illustrative examples. An instrument that measures
a pollutant by observing its spectral properties within a closed
cell into which an air sample is pumped is obviously in contact
with its sample. A second instrument that performs the identical
function by observing the spectral properties of the pollutant along
the path of a light beam projected from the instrument to an external
reflector and thence back to the instrument is - by this definition -
non-contact, although a purist might argue that the reflected external
III - 17
-------
beam effectively stretches the instrument out and wraps it around
the observed sample. I repeat that, for the purposes of this
paper, the latter is defined as a non-contact operation.
The program of the Air Quality Measurement Methods Branch
includes a single segment comprising non-contact measurements as
I have just defined that term. The remainder of my presentation
will consist of descriptions of the several tasks presently operating
within that segment.
Presently, there are three active contracts within the area of
long path optical measurement of ambient air pollutants. As all
three are based on the same basic principle a single description
of that principle will apply to all three and thereby save both your
time and mine.
Monochromatic light having a wavelength corresponding to a
principal absorbance maximum of the pollutant to be measured is
projected outwardly from a source to a retroreflector, which returns
the beam to its origin. There sensitive detectors read the
intensities of the transmitted and received beams. Electronic
logic devices compare the two light intensity values and display
a signal representing the attenuation of light of the selected
wavelength as the beam traversed the path source-reflector-detector.
In the simplest case the light attenuation whose measurement
we have just described serves as a direct measure of the concentration
of the pollutant of interest. But Nature is seldom simple, so in any
real-life situation we will observe an attenuation that is the sum
of that caused by the pollutant, plus that caused by other gases
present - e.g., CO-, water -, plus a further loss caused .by tur-
bulence of the air in the optical path. If we are to obtain an
accurate measure of a pollutant we must find ways of eliminating
the extraneous effects.
Potentially interfering contributions to light attenuation, if
they are wavelength-independent, can be effectively eliminated by
reading the difference in attenuation between two wavelengths, one
being at a peak, the other at a valley, in the absorbance spectrum
of the pollutant of interest. This strategy is capable of almost
completely eliminating the turbulence effect, provided the two wave-
lengths are observed over a period (milliseconds) less than the time
required for a significant change to be induced by turbulence. Its
capability for eliminating interferences due to other absorbing gases
III - 18
-------
is less, to the degree that the spectra of such gases possess
structure at or near the wavelength of observation. This capability
can usually be extended, however, by observing at three, four or
more wavelengths in a manner such that some algebraic combination -
say A-B+C-D, for a simple example - is dependent on pollutant
concentration but independent (or relatively so) of the interferent.
The light sources used in the several projects now active are
characterized by being highly monochromatic. That is, their entire
output lies within a very narrow wavelength band. This, together
with the added quality of tunability, gives them great flexibility,
in applying the strategy which I have just described in order to
attain great specificity for any given pollutant.
Before discussing the individual contracts, I would like to
spend a moment on the rather high level of apparent redundancy among
them. In most of our program we reserve the sponsoring of parallel
efforts aimed at a single objective for those needs for which we
cannot accept the failure of a single selected approach to reach the
objective within a fixed time frame. This is nearly equivalent to
saying that redundant strategies are applied only in situations
approaching the crisis category. In the present case, that of long
path optical measurement, our reason for funding parallel - possibly
redundant - efforts is based, not on any overwhelmingly important.
need, but on the simultaneous emergence of three separate proposals,
all already partially developed through other funding, each having
about the same probability of success, and each involving a highly
specialized technical team that will be disbanded if funding support
cannot be provided. Under these circumstances we feel that we are
fully justified in supporting all three contractors.
In a contract with the General Electric Company we are
investigating the application of gas lasers to long path instru-
mentation. This effort centers on, but is not restricted to, the
carbon dioxide laser. Carbon dioxide when excited within the cavity
of a laser initially emits a series of some 70-odd spectral lines
in the 9-11 micron region. An external grating and chopper operate
to sequence four lines, preselected by computer comparison of all
possible lines with the absorbance spectrum of the pollutant of
interest (in this case ozone). The four light pulses, each carrying
light of its own unique wavelength, exit from the instrument, traverse
the atmosphere that is being observed, and are returned (again traversing
the same path) by a retroreflector situated some 1 or 2 kilometers
away.
Upon returning to the instrument, the intensity of each pulse
is measured and compared with that measured for the same pulse a
III - 19
-------
few microseconds earlier on its outward journey. The differences,
or attenuations, of all four wavelengths become the coefficients
of a set of equations whose solution equals the average ozone
concentration along the optical path just described. An on-line
computer derives the necessary equations through an empirical
calibration procedure and thereafter serves to,accomplish their
solution.
The instrument that I have described is presently being
compared with point-sampling instruments that physically traverse
the same path. Assuming reasonable agreement in this test, it is
planned to assemble a prototype instrument on a trailer mount and
to carry out extensive field evaluation.
A second contract, with the Lincoln Laboratories of Massachusetts
Institute of Technology is, in general concept, sufficiently similar
to that just described to eliminate the need for a separate overall
description. Its most salient feature is the use of a tiny crystal
diode laser as its light source, analogous to the gas laser previously
described in connection with the General Electric program. Lincoln
Laboratories have developed the art of synthesizing such crystals in
a form that exhibits spectral resolution several orders of magnitude
finer than is possible with gas or other lasers. By appropriate
control of composition they can achieve rough tuning close to a
desired wavelength, and can perform fine tuning to a precisely
selected valufe by controlling the operating current. Described in
simple terms, the resolution of such a laser is to that of a tunable
gas laser as the gas laser is to a grating monochromator. Resolution
of this order is an extremely powerful tool in attaining the needed
single-substance specificity in long-path pollutant measurement,
since it multiplies the ability to discriminate against potential
interferents. To counterbalance the obvious advantages of Lincoln
Labs' diode lasers, we are faced with two principal disadvantages
inherent in them. One is that their power output is extremely small,
thus requiring the most sensitive light detectors for their utilization-
and these must be operated at or near the temperature of liquid helium.
The other is that the diodes themselves must be operated at cryogenic
temperatures in order to maximize their resolving capability. For-
tunately it is possible to mount laser and detector in fairly close
physical proximity to one another, so that both demands for cooling
can probably be satisfied with a single device. Lincoln is now
carrying out studies to determine the optimum tradeoff between cooling
requirement and quality of performance of the system. It appears
quite likely that they can assemble a system capable of operating on
III - 20
-------
commercially available mechanical refrigeration, thus avoiding
the need for supplying liquid helium to a field installation.
A breadboard assembly of Lincoln's best effort to date will
be compared with moving point-sampling instruments in a preliminary
evaluation scheduled for later this month.
Our third extramural effort in long-path optics consists of a
grant to Tulane University. A group under Dr. Hidalgo there
is studying gas lasers similar to those being used in the General
Electric program. Their effort involves theoretical and experi-
mental application of more sophisticated tuning techniques to the
carbon dioxide laser, with their results being checked out through
actual measurement of atmospheric ozone and comparison of results
with moving point samplers.
Ill - 21
-------
A REVIEW OF AVAILABLE TECHNIQUES FOR COUPLING CONTINUOUS GASEOUS
POLLUTANT MONITORS TO EMISSION SOURCES
James B. Homolya, Research Chemist
National Environmental Research Center, Research Triangle Park, N.C,
Chemistry and Physics Laboratory
INTRODUCTION
The state-of-the-art in the development of source level
pollution monitors has reached the stage'at which several viable
detection methods are available. Instrument systems have been
designed which are based on either process analyzer concepts or
modification of ambient air sensors. Measurement techniques
such as nondispersive infrared or ultraviolet absorption have
long been incorporated in the analysis of process streams whereas
flame photometry, chemiluminescence, and electrochemical trans-
ducers have found initial applications in ambient air monitoring
instrumentation where high sensitivity is required.
Nearly all of these sensors have the required sensitivity,
freedom from interferences, and adequate response time where
application for source level pollution monitoring presents no
problem on the detectors themselves. But we find that for moni-
toring stationary sources of gaseous emissions, consideration
must be given to the complete monitoring system. Stack emissions
usually contain corrosive gases at elevated temperatures. Such
streams may have a high dew point temperature or include parti-
culate matter of varying composition and size. A monitoring
III - 22
-------
system must be capable of continuously extracting a sample from
these types of sources» transporting it to the detector, and
conditioning it, if necessary, for an accurate analysis.
It is important that the extraction, transport, and condi-
tioning of the sample be consistent with the analytical method
involved. At present, there are three sampling conditioning
techniques available. These consist of: a) a "brute-force"
approach; b) dilution techniques; or c) in-situ measurement.
SOURCE LEVEL SAMPLE CONDITIONING
Figure 1 illustrates a typical measurement system for
extractive monitors having application for the analysis of S0~
or oxides of nitorgen from combustion sources. The gas sample is
withdrawn from the stack via a filtered-probe and passed through
a water removal system (usually a refrigerated dryer) before
entering the analyzer itself. For lone term operation, the water
condensate is continually removed and the probe filter is
periodically back-flushed with compressed air to remove entrained
particulate matter. The svstem also contains some provision
for the introduction of zero and calibration gases. Several
potential sources of error can exist in such a sampling system
prior to the instrument detector. Sample integrity can be
destroyed by: (1) chemical reaction with surface materials,
(2) chemisorption on particulate matter, (3) solution in the
water condensate, and (A) leakage of sample lines. The
Environmental Protection Agency has contracted an investigation
III -23
-------
into these problem areas. Hopefully, design criteria can be
established for analyzer-interface combinations applied to
general stationary source categories..
An extractive monitoring system for combustion sources
requiring minimal sample conditioning is illustrated in Figure 2.
An air aspirator is utilized to extract a sample from a source
on a continuous basis. As our experience has demonstrated that
sample pump failure has been a major problem area in continuous
source monitoring, the use of aspirators could be advantageous
if a source of plant or instrument air is available at the
system installation site. This particular system, incorporated
as part of a research study discussed later in this paper, is the
DuPont 460/1* S09 and NO analyzer. Stack gas is passed
£m X
through a heated sample cell positioned between an ultraviolet
energy source and a phototube detector. Sulfur dioxide and the
nitrogen oxides in the sample gas are analyzed in sequence. A
split-beam photometer is utilized by measuring the difference
in energy absorption at 280 nm and 436 nm measuring wavelengths
and at a 570 nm reference wavelength. The 280 nm wavelength is
chosen for SCL measurement with the 436 nm wavelength selected for
N02 measurement. Since nitric oxide has little absorbance in
the visible and ultraviolet, conversion to NCL is required for its
* Mention of Company name or product is not intented to
constitute endorsement by EPA.
Ill - 24
-------
measurement. The system achieves this conversion by reacting
NO in the sample with oxygen at high pressure. By measuring
the reaction product at roughly 90% completion, a sequential
analysis of S0«, N0«, and NO can be accomplished every 15
minutes. Between each analysis sequence the gas cell, heated
sample line, and probe filter are backflushed automatically with
air to remove particulate matter from the system and obtain a
"zero" for the photometer.
The particular design of an extractive monitoring system
depends largely on the characteristics of the emission source.
For example, Figure 3 illustrates a typical monitoring system
configuration for analyzing the atmospheric emissions from
sulfur acid plants. In this arrangement, the probe filter has
been eliminated because of the absence of particulate matter in
the source emission stream. The filter has been substituted with
a coalescing device to collect sulfuric acid mist before con-
taminating the analyzer. This type of system has been widely
used in conjunction with many of the nondispersive infrared
analyzers.
SAMPLE DILUTION TECHNIQUES
Dilution techniques can offer an advantage in sample con-
ditioning by eliminating heated sample lines and water vapor
removal systems, if the stack gas sample can be quantitatively
diluted as close to the source as possible. Such devices are
based on controlled flow, permeation sampling, or mechanical means
III - 25
-------
Figure 4 illustrates a controlled-flow dilution system
for monitoring stationary sources. The stack gas sample is ex-
tracted via a filtered probe and transported to the dilution
network and analyzer by heat-traced sample line. At this point,
the source sample is quantitatively diluted with air by a con-
trolled- flow/orifice combination. In this particular system,
sulfur dioxide is measured by a flame photometric detector.
Principally, the dilution concept is incorporated in this device
to dilute the SOn level into a range of linear detector response.
The dilution ratio is in the range of 1000 to 1. The sample
stream must be filtered from particulates to avoid altering the
orifice dimensions which would change the dilution ratio. In
addition, a constant temperature must be maintained at the
dilution network.
More recently, the controlled-flow approach has been applied
2
to an in-situ dilution system. In this application, a
specially-constructed sampling probe acts as the orifice but
with attainable dilution ratios ranging from 2:1 to 20:1.
Diffusion or permeation sampling devices have been widelv
Q /
reported in the literature. ' A general configuration for such
systems is illustrated in Figure 5. A source sample stream is
introduced into a chamber divided by a membrane permeable to the
gaseous component of interest. Gases permeating the membrane are
swept from the chamber by a carrier stream and delivered to the
analyzer. Both FEP Teflon and silicone polymers have been used
III - 26
-------
as rcerr^rane tnateri si for SCL and NO dilution. In addition,
polymer tubes have been substituted for the membrane. In this
manner, the tube is enclosed in a temperature controlled chamber
and the sample stream passes over the outer surfaces of the
tube with carrier gas flowing through the tube. The desired
dilution ratio is dependent upon: (1) the permeability of the
membrane or tube to the component of interest; (2) the surface
area of the membrane or tube and its temperature; and (3) the
volume flow of the carrier gas stream. Systems utilizing this
technique are commercially available from several manufacturers.
However, the permeation samplers still require extraction of a
source sample which must be filtered and held at ari elevated
temperature prior to entering the diffusion chamber.
Recently, a mechanical device has been developed in our
laboratory to quantitatively dilute a source sample in-situ,
eliminating the need for heated sample lines and probe filters.
The system, illustrated in Figure 6, utilizes a rotating disc
containing sample chambers of known volume. The disc is sand-
wiched between two stationary discs having sample inlet ports to
allow gas exchange between the sample chambers and the stack gas
environment. In practice, the dilution head is inserted into the
emission stream. Rotation of the sample disc effects gas ex-
change at the sample inlet ports and at a mixing chamber into
which a diluent gas is introduced. The diluted sample stream
is then analyzed by an ambient air analyzer. The dilution
III - 27
-------
ratio depends upon: (1) the number and volume of sample chambers;
(2) the rotational speed of the sample disc; and (3) the volume-
tric flow of the diluent gas. Operation of the sampler has been
demonstrated in the field by the continuous analysis of the S02
emissions from a 190 megawatt pulverized coal boiler.
Figure 7 represents a typical 24-hour segment from a week's
continuous operation during which the "disc diluter" was coupled
to a conductometric ambient SO^ monitor. The resultant SO^
emissions, based on a dilution ratio of 1600:1 are plotted
against the net load, in megawatts, from the boiler turbine
generator. The diluter/analyzer combination appears to follow
the trends in power output quite consistently. Further develop-
ment of the dilution system is being carried out under contract
to determine its range of applicability. Also, design modi-
fications have been incorporated to allow in-situ calibration of
the dilution head.
IN-SITU MONITORING
In-stack measurement avoids any extraction of sample by
utilizing the sample stream itself as an analysis chamber. These
instrument systems employ electro-optical detection which can be
arranged in three differing configurations.
A folded-path design places the energy source and receiver
at the same location. In this manner the energy beam enters the
emission stream through a. slotted probe and is reflected back
into the instrument. For large stack or duct diameters, the path-
Ill - 28
-------
length of measurements mi.ght be representative of a relatively
small portion of the stack diameter.
A double-ended system is one in which the source and
receiver are located at opposite ends of the stack diameter.
However, some instruments still might require the use of a slotted
pipe extending across the stack to either prevent misalignment
of the optical beam or restrict the absorption pathl^ngth to
maintain a linear detector response.
Recently, an investigation has been completed which com-
pared both extractive and in-situ electro-optical instrumentation
for the measurement of SO^ emissions from a pulverized-coal
power generating; boiler. An assessment was made of individual
system performance under field conditions. To accomplish this,
particular areas of interest in this study included: (1) an
investigation of the effects of variations in fuel composition,
boiler operating; conditions, and particulate matter on the
various measurement systems; (2) a correlation of instrument
response with standard EPA compliance test methods; and (3)
a determination of several instrument operating criteria such as
zero drift, span drift, and maintenance.
This study was carried out at the Duke Pow^r Company's
River Bend Steam Station in Charlotte, North Carolina, from
Januarv through March, 1973. Table 1 outlines the instrumentation
utilized during the study. Sulfur dioxide levels were simul-
taneously monitored by three discrete systems. They consisted
III - 29
-------
of the DuPont 460/1 source monitoring system for S0?, NO, and
N0x; the CEA Mark IV in-situ S02 system; and the Bailey Meter
Company SC^ source analyzer. A in-stack transmissometer measuring
opacity and a beta-gauge mass particulate monitor which were
installed in the source stream as part of parallel research
studies in progress at the time provided supporting measurements.
An instrument was also used which was able to provide a continu-
ous record of stack gas velocity and temperature.
The instruments and sampling probes were installed in the
stack of a 150 megawatt wall-fired boiler. The power generating
unit was equipped with both hot and cold electrostatic precipi-
tators containing a total of 14 stages. Therefore, the particu-
late loading in the stack could be varied in finite increments
over a wide dynamic range. A small building was erected at the
base of the stack to house the instrument control units and a
digital data acquisition system. The output signlas from all of
the monitoring devices were coupled to are Esterline-Angus 2020
digitizer with a teletype print out. In addition, each measure-
ment was recorded on stripcharts.
Figure 8 is an illustration of the optics utilized in the
CEA in-stack SCU correlation spectrometer. In this system,
light from a tungsten halogen lamp is collimated and then reflected
off the flat zero/read mirror. When this mirror is in the "rea'd"
position, light is directed into the probe and the probe mirror
directs the light beam back into the spectrometer. If SCL is
present in the probe slot, it will absorb energy in regularly
III - 30
-------
spaced bands at 3025^.. Light passing through the entrance
slit is reflected off the modulator mirror to the diffraction
grating. The grating disperses the light, spatially displaying
a focussed absorption spectra of SO^ at the exit mask. The
optical center of the modulator mirror is tilted at a slight
angle with respect to its axis of rotation, causing the angle at
which light strikes the grating to vary as the motor rotates.
This in turn causes the absorption spectra to scan the exit
mask in a circular fashion, creating a series of harmonics whose
intensity represents the SCU concentration.
When the zero/read mirror is in the "zero" position, light
is redirected into the spectrometer, by-passing the sample probe,
to provide a zero check. If a temperature-stabilized gas cell
containing a known SCL concentration is introduced into the path
while the mirror is in the "zero" position, an instrument span
check can be made. The output signal is affected by temperature
variations of the absorbing gas in the sample slot. The effect
is governed by the Charles' Law Relationship and spectral band-
broadening at high temperatures. A 6°C- change in stack gas
temperature alters the output signal by 3 percent. This could be
significant in applications in which there are wide fluctuations
in the emission temperature.
Figure 9 illustrates the operation of the Bailey SCL source
monitor. The system consists of a source housing and a receiver.
The source contains two hollow cathode lamps, a reference sensor,
III - 31
-------
and optics. The receiver contains another identical sensor.
A slotted pioe is provided with a inetered flow of purge air to
each housing to maintain a definite optical pathleneth through
the gas to be analyzed. The electronics are contained in a
cabinet. The source lamps were chosen to emit ultraviolet energv
at two closelv spaced wavelengths. In operation, the lamps
are pulsed alternately on and off and out of phase with each
other. Under these conditions, the sensors are detecting the
absorption of UV enerey at two discrete wavelengths closply
spaced such that the extinction coefficient of particulate matter
remains the same. Therefore, the output signals represent the
average SCL concentration across the pathlength determined by
the slotted pipe.
After installation, the instruments were operated continuously
for three months. For nearly two-thirds of this period, they
were left unattended as an attempt to assess their true relia-
bility in an actual installation. An example of the information
being obtained in the program is outlined in Table 2 which
summarizes the data from an experiment to determine possible
particulate interferences on the measurement systems. In this
experiment, the particulate concentration in the stack gas stream
was systematically increased to yield a range from 2 1/2 percent
through 48 percent stack opacity. During this time, the instru-
ments were operated continuously and the data logger was cycled at
10-second intervals to closely approximate integration of each
III - 32
-------
instrument signal. Concurrently, a series of compliance test
(Method 6) samples were taken as a reference. During the course
of the experiment, net load of the boiler output increased
which resulted in a proportional increase in the S0? concentration
by some 50 percent as seen from the Method 6 analyses.. Through-
out the period, data obtained from the DuPont and CEA systems
did not show an appreciable effect from the increasing parti-
culate level. However, the Bailey monitor did appear to respond
to the changes in opacity since the relative error in SO^ con-
centration as measured by the instrument was 28 percent higher
than that of the Method 6 sample obtained at a stack opacity of
2 1/2 percent. The relative error shows a consistent decrease
with an increasing opacity level to an error of approximately
10 percent at a stack opacity of 47 percent.
At this time, it is felt that the cause of the interference
is related to the pulse frequency of the hollow cathode lamps.
Data from the in-stack transmissometer indicates that the parti-
culate concentration varies dynamically over very short time
intervals. These fluctuations can occur either in or out of phase
with the pulsing rate of the Bailey sources. If they occur out
of phase, the energy absorbed during the pulse of one source
would occur at a higher background of particulate matter relative
to the second source, appearing as an erroneous measurement.
Figure 10 serves to illustrate this characteristic behavior. In
this figure, the transmissometer recording is shown as well as
III - 33
-------
reproductions from the Bailey and CEA stripcharts. The DuPont
system did not contain a continuous recorder because of its
sequential operation. The analyzer's S0~ analysis sequence was
held for the experiment and the data logger output for the
DuPont instrument was used to represent its analysis in Figure
10. At the present time, reduction of all of the monitoring
data is underway and will be presented in a future publication.
In addition, a similar study has been planned for the fall of
1973 to investigate the performance of certain in-situ and
extractive NO and NOj monitoring systems.
SUMMARY
In summary, we have seen that there are several approaches
which one can take to continuously monitor a gaseous source
emission stream. The extractive approaches involve certain
degrees of sample conditioning dependent upon not only the nature
of the emission source, but also the detection technique being
employed. The conditioning techniques include filtered probes,
heated sample lines, water vapor removal systems, or a variety of
dilution devices coupled to ambient air monitors. Dilution
systems might offer the advantage that both ambient air and source
monitoring could be accomplished by the same instrument. A
recent investigation has demonstrated the viability of in-situ
monitoring for SO- which, in effect, eliminates all sample con-
ditioning.
Ill - 34
-------
REFERENCES
1. McCov, J. —"Investigation of Extractive Sampling Interface
Parameters," Walden Research Corporation, EPA Contract
68-02-0742, February 1S73.
2. Rodes, C. E. "Variable Dilution Interface System for
Source Pollutant Gases," Proceedings, Analysis Instrumenta-
tion, II, 125-128 (April 1973).
3, McKinley, J. J. "Permeation Sampling—A Technique for
Difficult On-Stream Analyzers," Proceedings, Analysis
Instrumentation, 10, 214 (May 1972).
4. Rodes, C. E.f R; M. Felder, and J. K. Ferrell. "Permeation
of Sulfur Dioxide Through Polymeric Stack Sampling Interfaces,"
Envir. Sci. Tech., 7, 545 (1973).
5. Homolya, J. B. and R. J. Griffin. "Abstracts," 164th
National Meeting of the American Chemical Society, New
York, New York, No. WATR-84, August 1972.
6. Hedley, W. H. "Construction and Field Testing of a Com-
mercial Prototype Disc Diluter," Monsanto R-esearch Corpora-
tion, EPA Contract 68-02-0716, January 1973.
7. Federal Register, "Method 6--Determination of Sulfur Dioxide
From Stationary Sources, December 23, 1971, po. 24890-24891.
Ill - 35
-------
FIGURES
Fipure 1. Typical Measurement System for Combustion Sources
Figure 2. "Pumpless" Monitoring System.
Figure 3. Sulfuric Acid Plant Monitoring System.
Figure 4. Controlled-Flow Dilution System.
Figure 5. Diffusion Diluter/Analyzer Network
Figure 6. Disc Diluter
Figure 7. SCL Concentration, ppm vs Net Load, Mw, March 17,
197.2.
Figure 8. CEA-MK IV In-Stack S02 Monitor.
Figure 9. Bailey SCL Source Analyzer.
Figure 10. SC>2 Concentration, ppm and Stack Opacity, 7o,
0900-1400, March 1, 1973.
Ill - 36
-------
VENT
MONITOR
' 000
FLOWMETER
' r oi
I
MX!
FLOW
REGULATING
VALVE
FUNCTION
SELECTOR
VALVE
H
H
H
I
U>
AUTOMATIC
CONDENSATE DRAIN
AND SAMPLE BY-PASS
MOUNTING
HARDWARE
PURGE GAS INLET
ZERO GAS INLET
SPAN GAS INLET
SELECTOR VALVE
FILTER LOADING
INDICATOR
Q
SAMPLE PUMP
COOLANT PUMP
HEATED SAMPLE
LINES
THERMOSTAT
COOLANT REFRIGERATION
RESERVOIR UNIT
-------
RECORDER
IT1
PHOTOMETER
SAMPLE CELL
1
S BALL —}
a
j-U
PROGRAM^ TvTcUUM BREAKER l™^ ^ ^
J * 9 \ *•.*** Ixl*
i
LAMP
CONTROL
STATION
I
i
LJ
v~,
\» •
\*k •
1
f
HXJ-kx^ *
j
I IX^>I ^*^ ^S*-i ^ • i .. ^
A!R
02 CYL.
ASPIRATOR
3
I
AC POWER -- ~~ --- f-
L
_,„/•
_ ,
^
, —
1
t
POWER
SUPPLY
1
rsU ^
-t^si
T •
k pvl^f ! ,
U"si 5
PRESSURE REGULATOR
© OXYGEN GUAGE
TRAP
— a
4
^FLOW !
INDICATOR
i
HEATED I
COMPARTMENT
J
STACK
I i
FILTER PROBE
ELEC SOLENOIDS
HEATED TUBING
SAMPLE IN
J
00
H
H
EXHAUST
-------
ACID MIST FILTER
H
H
H
VO
FLOWMETER
ANALYZER r-
I
LJ
SECONDARY FILTER
VENT TO
ATMOSPHERE
u*
t
CALIBRATING GAS
-------
PRESSURE
REGULATOR
AND GUAGE
HYDROGEN
INLET
V
AIR INLET
SAMPLE INLET f
N0.1
NO
NO
CALIBRATION :
INLET
A
EXHAUST TO
EXTERNAL AMB
COLD PIPE AIR
SOLENOID
SAMPLE SOLENOID
VENT TO
CABINET
4 I/min
ORIFICE
NEEDLE
NEEDLE VALVE
VENTURI
ASPIRATOR
TRANSDUCER ^ PRESSURE
REGULATOR
AIR TRANSDUCER
CALIBRATION
SOLENOID
PRESS TO
TEST SOLENOID
AUTOMATIC
HYDROGEN
SHUTOFF
SOLENOID
HYDROGEN
SHUT OFF
VALVE
•'1*01
i r^
NO. 3 /*p~f
v-n VENT TO
-4— EXTERIOR OF
NEEDLE VALVE ^ COLD PIPE ^ . I CABINET
HYDROGEN "™
°-
-------
SOURCE SAMPLE
STREAM
~\
H
H
H
\
MEMBRANE
CARRIER STREAM
FILTER
VACUUM PUMP
DILUTER CARTRIDGE
VACUUM PUMP
CARRIER BLEED LIME
TO ANALYZER
ANALYZER
-------
H
H
H
to
STACK GAS FLOW
DILUTED SAMPLE
RETURN TUBE
SAMPLE INLET PORT
SAMPLE INLET PORT
TO STACK WALL
AIR INLET TUBE
DISC DRIVE
/^SAMPLEDISC
SAMPLE CHAMBER
-------
200
190
180
I H-H-H-H I I i-U-U-l
H
H
I
w
o
CVI
2
160
150
140
130
120
110
900
80°
700
600
500
400
300
200
100
— HOURLY AVERAGE LOAD
O INSTANTANEOUS LOAD
DILUTION RATIO -1600:1
12 1
8 9 10 11 NOON 12
TIME OF DAY, hours
7 8 9 10 11 12
-------
SAMPLE SLOT
H
H
H
STACK GASES
1.
2.
3.
4.
5.
6.
1.
8.
9.
10.
11.
12.
13.
14.
TUNGSTEN HALOGEN LAMP
COLLIMATING LENS
ZERO/READ MIRROR
LAMP FOCUSING LENS-PRIMARY
PROBE MIRROR
LAMP FOCUSING LENS-SECONDARY
FILTER
ENTRANCE SLIT
MOTORIZED CALIBRATION CELL
MODULATOR MIRROR
DIFRACTION GRATING
EXIT CORRELATION MASK
EXIT MASK LENS
PHOTO.MULTIPLIER
-------
SOURCE
J_
FLUE GASES
\\\\\
REFERENCE
SENSOR
\\\\\\
FIRST HALF CYCLE
S02 MEASURING
SENSOR
H
H
Ui
SOURCE
FLUE GASES
\\N\\S
REFERENCE
SENSOR
WSSN
SECOND HALF CYCLE
-------
m
,»
- r- -«*1
! i
m
im
m
m
IK
soo
«0
me
TWEOFMT.Imn
UN
DO POUT
m
im
ISO
IWt OF DAY.
IM UW
TtK OF 0»y. IHWI
a
III - 46
-------
Table 1. Instrumentation Utilized During Study, January 1-
April 1, 1973.
Table 2. Effect of Increasing Stack Opacity on Instrument
Response, March 1, 1973.
Ill - 47
-------
INSTRUMENTATION UTILIZED DURING STUDY
A, SQ2 MEASUREMENT
1, DuPoNT W/l SOURCE MONITORING SYSTEM
2, CEA/BARRHIGER f-K-IV S02 STACK MONITOR
3, BAILEY METER Co, S02 STACK MONITOR
B, SUPPORTING MEASUREJCNTS
1, EPA METHOD 5-S02
2, EPA METHOD 54kss PARTICULATES
3, TRA.NSMISSOMETER-OPACITY
4, BETA GAUGE-MASS PARTICULATES
5, PMC AUTOPITOMETER-COfNlTINUOUS STACK GAS VELOCITY, TEMPERATURE
C, DATA ACQUISITION
1, EA 2020 + TELETYPE
2, STRIPCHART RECORDERS
III - 48
-------
S/WPLE
A
B
C
D
E
F
OPACITY
2,5%
3,0
15,0
17,5
45,0
47,5
rcnioD 6
460PPM
475
675
619
661
682
DuPONT
451, - 1,9%
438, - 7,8
577, -14,5
632, +2,1
664, + 0,5
678, - 0,6
CEA MKIV
424, - 7,8
416, -12,4
550, -18,5
599, - 3,2
624, - 5,5
638, - 6.4
BAILEY %
590, +28,2
570, +20,0
680, + 0,7
720, +16,3
730, +10,4
750, +9,9
H
H
H
-------
SESSION IV
WATER QUALITY SENSOR DEVELOPMENT
CHAIRMAN
MR. A. F. MENTINK
NERC-CINCINNATI
-------
ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
OFFICE OF MONITORING SYSTEMS
INSITU SENSOR SYSTEMS FOR
WATER QUALITY MEASUREMENT
By
K. H. Mancy
Professor of Environmental Chemistry
School of Public Health
The University of Michigan
Ann Arbor, Michigan 48104
Presented at the "Second Conference On
Environmental Quality Sensors", October
10 - 12, 1973. National Environmental
Research Center, Las Vegas, Nevada.
IV _ 1
-------
OUTLINE
I. INTRODUCTION IV-3
II. SENSOR PERFORMANCE CHARACTERISTICS IV-5
1 - Primary Performance Characteristics IV-5
2 - Secondary Performance Characteristics IV-7
III. IN-SITU SENSOR SYSTEMS IV-8
1 - Oxygen Membrane Electrodes IV-8
2 - Specific Ion Electrodes IV-13
a - Solid State Specific Ion Electrodes IV-14
Silver Halide Specific Ion Electrodes IV-14
Silver Sulfide Specific Ion Electrodes IV-15
Lanthanum Trifluoride Electrodes IV-17
b - Liquid Ion Exchange Specific Ion Electrodes IV-18
The Calcium Electrode IV-18
The Nitrate Electrode IV-18
c - Gas Sensitive Potentiometric Membrane IV-19
Electrodes
IV. ON LINE MONITORING WITH SAMPLE CONDITIONING IV-20
V. CONCLUSIONS IV-20
VI. BIBLIOGRAPHY IV-22
APPENDIX - A CONTINUOUS MONITORING BY AUTOMATED CHEMICAL IV-23
SYSTEMS
APPENDIX - B COMMERCIALLY AVAILABLE SPECIFIC ION ELECTRODES IV-24
IV - 2
-------
I - INTRODUCTION
Monitoring is defined by the United Nations as the "process of repetitive
observing for defined purposes of one or more elements or indicators of the environ-
ment according to pre-arranged schedules in space and time, and using comparable
methodologies for environmental sensing and data collection." Present trends in
pollution control activities indicate an increasing reliance on automated quality
monitoring systems. This stems from the fact that manual measurements are, for the
most part, inefficient and limited by the frequency with which samples can be
collected for analysis. The time delay between sampling and analysis can result in
certain changes in the characteristics of the sample, causing it to be less repre-
sentative of the water body from which it was collected. The time involved in
making and reporting the analysis is also time lost before any water pollution
corrective action can take place. In addition, when viewed in relation to the
availability and cost of labor, automatic monitoring systems are considered to be
the least costly system. Nevertheless, our justification of automatic monitoring
systems should be accompanied by a realization of the limits of dependency on such
systems.
Generally speaking, automatic monitoring techniques can be categorized into
non-contact remote sensing procedures and contact in-situ or on-line continuous or
discrete sample analyses.
Remote sensing techniques are based on the measurement of electromagnetic
radiation reflected from the surface of water bodies over large geographical areas.
The concept of remote sensing using aerial surveillance is most appealing from cost
viewpoint. Unfortunately, the technique suffers from certain limitations, i.e.
IV - 3
-------
(a) detection is limited to a small number of quality characteristics, (b) results
usually exhibit poor accuracy and precision and (c) measurements are essentially
confined to the survace properties of water. Nevertheless, remote sensing has
been used with various degrees of success to measure temperature, oil, chlorophyll,
turbidity, and color. Typical applications include (a) location and typing of
waste outfalls, (b) assessment of oil spills, (c) algal biomass estimation, (d)
selection of sampling sites, and (e) identification of mixing zones. Detailed
discussion on this subject can be found elsewhere (1, 2, 3).
Insitu monitoring techniques rely on the direct placement of the sensor in
the environment to be measured. Consequently the sample step is eliminated and
the sensor response will be proportional to certain physical or chemical
characteristics of the acqueous phase according to established relationships.
It should be realized, however, that in-situ sensor responses may be significantly
influenced by water flow, temperature and a number of chemical and biological
interferences which may prevail at the monitoring site. These factors should
be taken into consideration whenever in-situ measurements are concerned.
On-line automated analyses are based on withdrawing water from a given
site, either as a continuous stream or discrete samples, and allowing it to flow
through the sensor system. This may be done with or without sample pretreatment.
Sample pretreatment may include temperature control, filtration, extraction,
dissolution, dilution, digestion and reagent a-ditions. In fact most of the
procedures which are usually carried out by an analyst, are automated and
performed on a stream of samples moved by a fixed-speed peristaltic pump, e.g.
the Technicon Auto~Analyzer (4). A schematic diagram of basic Auto-Analyzer
operations, and a listing of parameters for which automated procedures are
available are shown in Appendix A. Auto-Analyzer techniques find widest
application in laboratory operations where large samples of water are handled daily.
IV - 4
-------
There have been certain attempts, however, to use this technique for water
quality monitoring where the autoanalyzer is kept in a trailer on a river bank
or on board ship. The reader is advised to consult the following references for
typical applications of Auto-Analyzer techniques 4, 5.
Discussions in this paper are primarily concerned with a number of electro-
chemical sensors applicable to in-situ as well as on-line monitoring procedures.
This includes descriptions of currently available and newly developed sensors and
typical monitoring applications.
II - SENSOR PERFORMANCE CHARACTERISTICS
In monitoring systems the most critical part is the sensor system and the
reliability of measurement is mostly dependent on the reliability of the sensor
system. This is true whether the sensor is an electrode, a thermistor or a
photoelectric cell. A clear understanding of the operation characteristics of
the sensor and its dynamic response is essential. This is based on proper
calibration, servicing, maintenance and alertness for small clues that may
indicate malfunction.
Primary sensor characteristics are defined in terms of (a) sensitivity,
(b) response time, (c) selectivity, (d) long term stability, (e) accuracy, and
(f) precision. Secondary sensor characteristics are those which define the
environmental effects, e.g. (a) temperature, (b) hydrostatic and hydrodynamic
forces, (c) ionic strength, (d) pH, (e) sunlight, etc.
1 - Primary Sensor Characteristics;
Sensitivity is usually defined in terms of the smallest change in the
measured variable that causes a detectable change in the indication of the
instrument. It specifies the lower limit of detection of the sensor. Sensitivity
IV - 5
-------
is directly proportional to the slope of the curve relating the signal magnitude
to the amount of detectable material present. This will reflect directly on the
ability to ascertain a difference between the signal and backgroun noise at the
detection limit, i.e. given adequate precision, the greater the sensitivity, the
better the tdetectability.
The limit of detection of analytical method is the lowest concentration
whose signal can be distinguished from the blank signal. This value depends on
the sensitivity of the method, as well as the signal-to-noise ratio required to
discern the response due to a sample. Advances in electronics have brought about
the design of instruments with greater inherent stability and, therefore, lower
limits of detection. Use of an on-line digital computer in fast-sweep derivative
polarography has permitted the resolution of closely spaced peaks and extended
the analytical sensitivity of the technique by more than an order of magnitude.
The speed of the sensor response to changes in the test solution is
referred to as the "response time." It is an indication of the time needed for
the sensor signal to follow 90, 95 or 99 percent of instantaneous full scale
change in the measured variable. The response time should be specified for each
sensor indicating whether it is dynamic or static sensor response.
Selectivity of the sensor refers to the effect of interferences resulting
from detectable ions or molecules other than the species of interest. Since all
sensor systems cannot achieve absolute or 100 percent selectivity, then it is
important to specify the selectivity limitations in a given test solution. If the
type and amount of the interfering species are known, then it is possible to
incorporate the term "selectivity coefficient" in the sensors sensitivity expressionn.
Also, in certain cases it is possible to incorporate interferences effects in the
sensor calibration curve. This can be done by means of the standard addition
technique where known amounts of the raeasured ions are added to the test solution
and the proportional signal values are recorded.
IV - 6
-------
Long term stability usually refers to the change in the sensor's
performance characteristics with time. This is used to decide on the frequency
of checking the calibration or servicing the sensor. Long term stability is a pro-
perty of the particular system and is dependent on the presence of interferences
and the physiochemical characteristics of the test solution.
Drviations of results by a given sensor from the "true" value define the
accuracy of the system. If the source of error is found, and it is possible to
correct for it, this is called "determinate error." If the deviation from the true
value is compounded indiscriminately by many small errors, it is simply a "random
error." Random errors are subject to statistical treatment of the data.
Precision is defined in terms of the reproducibility of the sensor
measurement. The more scatter in successive readings, the less precise are the
measurements. Usually, precision is closely identified with random errors and
statistical theories.
2 - Secondary Sensor Characteristics;
Secondary sensor characteristics refer to the effect of environmental
variables. This can be a result of changes in the sensors primary characteristics
or changes in the physiochemical characteristics of the test solution. For
example, temperature -ffects on conductance measurement are quite complex since
the temperature coefficient is dependent on both ionic strength and temperature.
The conductivity of sea water was found to increase by 3 percent per degree increase
in temperature at 0°C, 2 percent increase at 25°C and about 5 percent increase at
30°C. It is therefore advisable to measure relative conductance than absolute
values. This is done by measuring the ratio of the conductance of the test solution
to that of a reference solution at the same temperature. Thermistors or resistances
can be used instead of the reference solution.
It is always advisable to establish the primary and secondary sensor
IV - 7
-------
characteristics for each sensor independently before using it for field applications.
Not only these characteristics will vary from one sensor type to another, but also
differences between two sensors of the same type and from the same manufacturer
may occur.
Ill - IN-SITU SENSOR SYSTEMS
Currently available in-situ electrochemical sensor systems applicable to
water quality monitoring are given in Table I. This includes conductometric,
potentiometric, and voltammetric electrode systems. Potentioraetric systems
include the glass electrode for pH measurement, inert metal electrode (platinum
or gold) for the measurement of oxidation-reduction potentials and potentio-
metric membrane electrodes which will be discussed later.
Workers in the field have expressed doubts regarding the utility of
oxidation-reduction potential measurements in surface waters. In fact, only
under anoerobic conditions (anoxic waters), and in certain industrial waste
effluents, can such measurements give meaningful results.
1 - Oxygen Membrane Electrodes:
At the present tim6 it appears that voltammetric membrane electrodes
are the only available sensors capable of in situ analysis of dissolved. The
unique features of such electrode systems is that the membrane separates the
electrode from the test solution. Two main types are presently available, the
voltammetric type and the galvanic cell type (6). The two types are similar in
operating characteristics; in the voltammetric type an appropriate potential
source is needed, however, while the galvanic type is basically an oxygen
energized cell.
Oxygen membrane electrodes have three main components: the membrane,
the oxygen-sensing element:, and the electrolyte solution. The membrane, the
unique feature in such c.Vctrode systems, serves .in three different capacities.
IV - 8
-------
Table 1. Electrochemical Sensors
(a) Conductometric
(b) Potentiometric
1. Glass Electrode
2. Inert Metal Electrode
(Redox Potential)
3. Potentiometric
Membrane Electrodes
n
RT I
constant + -— In ja.+K a,
zF ! i J J
']
pH » -log aK+
pE » -log a «
Cationic » pM
Anionic = pA
Eu/(2.3 RTF"1)
a.
-log aA-
(c) Voltammetric Membrane Electrodes- . . _
(Dissolved Oxygen) i, = jzFAP ~r-|an
u « TQ O I \Jj<
L J ^
(1)
(2)
(3)
(4)
(5)
(6)
(7)
L =• specific conductance
K « cell constant
c
C, ° ionic concentration
\, •= ionic equivalent conductance
z. *= ion valency
E ** measured electrode potential
m
P a the Faraday constant
K. B selectivity coefficient
m
» diffusion current
electrode surface area
« membrane permeability
coefficient
membrane thickness
IV - 9
-------
First, the membrane acts as a protective diffusion barrier separating the sensing
element from the test solution. Since plastic membranes are permeable to gases
only, oxygen molecules pass through, but electroactive and surface-active
contaminants present do not. The possibility of poisoning the sensing element
Is thus minimized.
Second, the membrane serves to hold the supporting electrolyte in
contact with the electrode system and thus makes it possible to determine oxygen
in gaseous samples as well as in nonaqueous solutions such as industrial wastes.
The third advantage is that the membrane constitutes a finite diffusion layer,
the thickness of which is independent of the hydrodynatnic properties of the
test solution.
A detailed discussion of the principle, operating characteristics,
applicability, and limitations in the use of oxygen membrane electrodes may be
found elsewhere (6).
Polymeric membranes used with oxygen membrane electrodes shew selective
permeability to various gases and vapors. Gases reduced at the potential of the
sensing electrode (e.g., SO- and halogens) cause erroneous readings, but these
gases rarely exist in .a free state in aqueous systems. Other gases capable
of permeating plastic membranes may contaminate the sensing electrode or react
with the supporting electrolyte, e.g., C0_ and H_S.
Oxygen membrane electrode systems have been used extensively in
laboratory and field analysis as well as for continuous monitoring purposes.
Some of the main operational problems associated with the use of these electrode
systems have been the effect of mixing in the test solution on the electrode
sensitivity and short terra stability which necessitates frequent calibration.
In monitoring operations, the accumulation of inert or biological material on the
membrane surface has caused a let of nuisance. In addition, it has been practically
IV - 10
-------
impossible to reproducibly change and mount the membranes without a change in
the membrane thickness and permeability characteristics.
As indicated in equation 7 the steady state diffusion current .i, is
a
directly proportional to the activity of molecular oxygen, a, and not necessarily
to the concentration, C. The activity and concentration terms are interrelated
by the following equation
a = YC (8)
where Y is the activity coefficient of molecular oxygen. It is only under
ideal conditions where Y will be equal to unity and the activity can be replaced
by concentration in euqation 7. For example in estuarine and sea waters or
brine wastewaters Y will be greater than unity - a phenomenon described as the
"salting-out"effeet. Also, the presence of soluble organic matter in the test
solution, e.g. alcohol, will result in Y values less than unity - a phenomenon
described as the "salting-in" effect. Accordingly, the application of the
oxygen membrane electrode in natural and wastewater should take account of these
factors.
The salting-out effect can be expressed as follows:
In Y = KS I (9)
where K is the salting out coefficient and I is the ionic strength. By
s
substitution of 8 and 9 in 7, it follows that
i. = fn FAP J] eKsI C (10)
d u m D a
Equation 10 can be reduced by introducing the proportionality constant in
a
replacement of term in brackets in equation 10,
i =
-------
Ionic strength values can be approximated by measurement of electrical conductance,
using Kg as a proportionality coefficient,
1= <}> eK'sI C (12)
u 3
where L is the specific conductance of the test solution. Equation 12 offers a
simple relationship to compensate for the salting-out effect using specific
conductance measurement. This compensation can be done automatically by a simple
analog circuit (6).
The main operational problems with available dissolved oxygen electrode
systems are (a) lack of stability, (b) easy fouling of the membrane and loss of
sensitivity, (c) dependency of sensitivity on flow, and (d) the poisoning effect of
H S. Regardless of these limitations, dissolved oxygen probes are widely used, and
with some extra effort in maintenance and checking, they yield reliable data.
Recent studies in this laboratory resulted in the development of an
advanced dissolved oxygen electrode system which does not suffer from the above
mentioned limitations (7,8). The newly developed dissolved oxygen probe utilizes a
three electrode voltammetric membrance cell of a special design. Measurement is
conducted in the form of pulse voltammetry at a predetermined frequency according
to the following relationship
„,..,
where D_ is oxygen diffusivity coefficient in the inner electrolyte solution layer
and .t is the pulse time. A comparison between equations 7 and 13 indicates that
the pulse current, ip, is neither dependent, on the membrane thickness, b, nor on
the membrance permeability coefficient, Pm. At appropriate pulse frequency, the
electrode sensitivity will not be influenced by depositions on the membrance surface
(fouling) and will be independent of mixing of the test solution. In addition, for
the same dissolved oxygen content i is approximately a hundred times greater than i
P d'
IV - 12
-------
To this end, the pulse oxygen membrane electrode is capable of exhibiting much
higher sensitivity and more stable performance characteristics than the presently
available steady state systems.
2 - Specific Ion Electrodes;
Perhaps the most significant advances in electrochemical sensor.
development in the last decade or two, is the emergence of potentiometric membrane
electrode systems, commonly referred to as specific ion electrodes or selective
ion electrodes. These electrode systems offer the possibility of monitoring a
wider range of anions and cations in natural and waste waters. The oldest of these
electrodes is the pH glass electrode which is available from a variety of sources
in the United States. pH electrodes with documented characteristics are available
from Beckman Instruments, the Coleman Division of Perkin Elmer, Corning Glass
Works, Leeds and Northup, Sargent Company, and H. A. Thomas, to mention the main
suppliers. European products include Radiometer in Denmark.
Glass electrodes responsive to monovalent ions, e.g..sodium, potassium,
silver, ammonium, are made by Corning Glass Works, Beckman Instrument and Orion
Research. Recently developed solid state specific ion electrodes including
single crystal, cast disk, or pressed pellet types are available from Orion
Research, Beckman Instrument, and Coleman Division of Perkin Elmer. Liquid ion
exchange electrodes are also available from Corning Glass Works, and Orion
Research. Heterogeneous membranes of the Pungor type are available from Radelkis
Electrochemical Instruments in Budapest, through National Instrument Laboratories,
Rockville, Maryland. Specific Ion Electrodes are also produced by Radiometer in
Denmark, Crytur in CSSR, and Phillips in Holland.
Specific ion electrodes of interest to water quality monitoring can be
conveniently classified into (a) solid state, (b) liquid ion exchange and (c) gas
responsive electrode systems.
IV - 13
-------
a. Solid State Specific Ion Electrodes
In these electrode systems the membrane contains a fixed sum of certain
kind of ions, the structure of which is constant in time. The solid ion exchange
membranes can be either homogeneous (single crystal, a crystalline subjstance in
a disk or pellet form, or glass which is considered to be solid with regard to the
immobility of the ionic groups), or heterogeneous, where the crystalline material
is included into a matrix made from suitable polymer. Heterogeneous solid state
electrodes are frequently known after their inventor E. Pungor (9-11).
A listing of commercially available solid state specific ion electrodes
is given in Appendix B.
Silver Halide Specific Ion Electrodes; In solid state electrodes of
the silver halide type, the silver ion is the charge carrier and does not involve
the diffusion of halide ions. There are three main types of silver halide solid
state electrodes. The aides type, developed by' Pungor at. al. (9, 10, 11) is
heterogeneous membranes composed of silver halide precipitated in silicone rubber
matrix. The latest version of these electrode systems is composed partly of
silver halide precipitate, incorporated into a silicone rubber matrix, and partly
of compact silver halide.
Homogeneous solid state membrane electrodes offer better performance
characteristics than the heterogeneous types. The membrane is made from a mixture
of silver halide and silver sulfide. The silver sulfide, which is much less
soluble than any of the silver halides, serves to increase the membrane conductivity
without affecting the membrane potential. These membranes have been found to be
more applicable for measurement of Cl~, Br~, I~ and CN~ ions. Main interfering
ions are sulfides and thiocyanates.
IV - 14
-------
The potential determining ion in these electrode systems ±s the Ag and
the Nernst relationship can be expressed as follows
E •= constant 4- — In [Ag ]
and since
[Ag+] = -4^T (15)
Where K. . is the silver chloride solubility product, then
Effi = [constant + |T- In K^Cl* ~ f^ ln tcl~l (16)
Of particular interest to water quality monitoring are the Cl and CN electrodes
which are being used with considerable success.
Silver Sulfide Specific Ion Electrodes: These electrode systems rely on
the use of highly conductive, sparingly soluble Ag?S membranes. The most applicable
types are those made of homogenous membranes of compressed pellets of Ag_S supplied
by Beckman, Orion, Corning, Philips and Coleman. These electrodes respond only to
S2 , Ag and Hg2 ions, and to a certain degree to Cn . Theoretically speaking, the
sensitivity to sulfides should follow a Nernstian relationship, to very low detection
limits of 10 20M. In order to achieve low detection limits it is recommended to use
plastic containers instead of glass vessels, since silver ions adsorb strongly on
glass.
Similar to halide electrodes, the potential determining ion is the Ag and the
Nernst relationship can be expressed as follows
1RT 4
E = constant 4- — In [Ag ]
m r
and since
K.
0.5-
[Ag+]
^T1 (18)
IV - 15
-------
where K._ „ is silver sulfide solubility product, then
U2F
- -z
This electrode is also responsive to other forms of free sulfides, HS
E - [constant 4- f~- In K, c]-~ In [S2~] (18)
"
and H_S, through the following equilibrium relationships
where KI and K2-are the first and second acidity constants for H S. Consequently,
if the pH is known, the following equilibrium relationships can be used to calculate
[HS~], [ELS] and total sulfides [S ] ,
log ' [ST] - [S2-] + log &£- + I|-l + 1 (2Q)
rw+i?
log [H2S] • log [S2 ] + log £y- (21)
2"
log [HS~] = log [S2"] 4- log -- (22)
Another type of the silver sulfide electrodes include those electrodes responsive to
4- + 4-
Cu , Pb or Cd . The membrane is made of a pressed mixture of silver sulfide and
divalent metal sulfides, e.g. PbS, CdS or CuS. The electrode is primarily responsive
to Ag activity which is dependent on the S2 activity, which in turn depends on the
activity of the divalent metal in solution according to the following equilibrium
relationships
[A/] [S2~] - K (23)
[M2+] [S2-] = KMS (24)
from which
—lQ.5
IV - 16
-------
Since
TDfp 11
E = const. + — In [Ag ]
m r
it follows that
E = const. + — In g? -f — In [M2 ]
m 2F ^ 2F
This electrode system is selective for M2 provided that
(KMS> ~ (KAg2S>
and MS solubility is sufficiently smaller than the concentration of M2 in the test
solution. In acidic media H may interfere. Nevertheless, the operating range is
from 10 ''M to 10 5M. The main interfering ions are Hg2 and Ag for the cupric ion
electrode and Hg , Ag and Cu2 for the cadmium ion electrode. The cupric ion
electrode may lose its selectivity in presence of Cl according to the following
reaction
Ag-S + Cu2+ + 2C1~ = 2AgCl + CuS (27)
2+
In this case the electrode may lose its selectivity to the Cu and gain sensitivity
to Cl~.
Lanthanum Trifluoride Electrodes; This electrode system is selective to
F ions and utilizes a membrane made of single crystal of LaF_ which has been doped
with europium (12). The Nernstian response is as follows
E = constant - ^ In [F~] (28)
m r
The operating range is between 1M to 10 6M. At low pH values the electrode potential
will decrease due to the formation of HF or HF» ions. Aluminum ions form stable
complexes with fluorides (AlFg3") over a wide pH range. At high pH the OH inter-
feres with the electrode's selectivity (13). The lanthanum trifluoride electrode
finds increasing applications in monitoring fluoride ions in natural waters, and
fluoridated water supplies (13, 14, 15).
IV - 17
-------
b. Liquid Ion Exchange Specific Ion Electrodes:
The calcium electrode.- Specific ion electrodes of this kind are based
on liquid membranes, in which are dissolved electroneutral salts of the ion of
interest and ion-exchanger. The membrane is formed in an insoluble solvent
layer (15). A typical example of these electrodes is the calcium or the
divalent cation (Ca2+ and Mg2+ or hardness) electrodes. The membrane in this
electrode system is a solution of calcium dialkyl phosphate in dioctylphenyl
phosphonate. The electrode exhibits a Nernestian response,
R T I 2+1
E - constant + -^ In jCa (29)
m Zr I J
The operative concentration range is 10 tt to 10 M. The electrode shows
a significant hydrogen ion sensitivity and is usableouly within the pH range
6-11. The calcium and hardness electrodes have been used by numerous investi-
gators for water quality analysis (14, 15, 19). The application of the
calcium electrode for water quality monitoring purposes is primarily limited by
the lack of stability and potential drift of these electrodes.
The Nitrate Electrode; In this electrode system, the ion exchanger is
methyltricapryl ammonium ion and the solvent is 1-deconol (16). The operative
—. — 9—
pH range is 4 to 7. Major interferences are HCO- , Cl , SO, and N0~ .
Similar to above mentioned calcium electrode, the nitrate electrode suffer from
lack of stability and potential drift which limit its application for monitoring purposes.
In addition, the electrode sensitivity to commonly present interfering ions, such as
_ - 2-
HCO- , Cl and SO, cannot be neglected.
A listing of commercially available liquid-ion exchange electrodes is
given in Appendix B.
IV - 18
-------
c. Gas Sensitive Potentiometric Membrane Electrodes;
Electrode systems of this type are available for the measurement of
CO-, NH, and SO . The CO electrode has been in use for over two decades (17, 18
19), while the NH_ and SO. are recently developed.
These electrode systems utilize a gas permeable plastic membrane to
separate a thin layer of an electrolyte solution from the test solution. The
electrolyte solution layer is in direct contact with the sensor system, e.g.
glass electrode. Measurement is based on allowing a soluble gas e.g. CO- in
the test solution to permeate through the plastic membrane and partition itself
between the inner electrolyte layer and the test solution. Subsequent changes
in the pH of the inner electrolyte layer, as determined by a glass electrode, are
directly proportional to C0? content in the test solution. The plastic membrane
may be polyethylene, teflon or rubber silicone.
In the case of the CO- electrode, the pH of the electrolyte layer will
be directly dependent on partial pressure of carbon dioxide across the plastic
membrane. The inner electrolyte layer is made of a mixture of KC1 and NaHCO- (17),
The electrode response may be improved by adding carbonic anhydrase (1 mg per ml)
to the inner electrolyte layer in order to accelerate the CO- solubility reaction.
Potentiometric measurement in the electrolyte layer requires the use of a
reference electrode, which is in electrolytic contact with the glass electrode.
Different electrode designs are reported in the literature (17, 18).
IV - 19
-------
IV ON LINE MONITORING WITH SAMPLE CONDITIONING
In view of the above discussion, it is apparent that certain electrode systems
can only function appropriately within restricted pH range or in the absence of
certain interferences. Consequently it seems imperative to modify the sample
physicochemical characteristics prior to measurement. Under these conditions in-
situ analysis is not possible. Sample conditioning includes (a) filtration,
dialysis or maceration, (b) dissolution, digestion or extraction, or (c) reagent
addition for pH control or masking of interferences.
This approach has been applied for on-line monitoring of drinking water
quality (13, 14, 20). On-line monitoring of continuous sample stream or
discrete samples were made for the measurement of (a) total and free fluorides
using the lanthanum trifluoride electrode, (b) copper, lead and cadmium using
differential anodic stripping voltammetric techniques, (c) alkalinity using
pH glass electrodes and (c) nitrates and hardness using liquid-ion exchange
specific ion electrodes.
V. CONCLUSIONS
In-situ analysis in which the sensor is placed directly in the environment
to be measured offers an ideal approach for water quality monitoring. Nevertheless
the application of in-situ monitoring techniques is limited to a few number of
parameter for which there are available sensor systems. In addition the response
of available sensors may be significantly influenced by water flow, temperature and
a number of chemical and biological interferences which may prevail at the
monitoring site. Recent developments in using pulse voltammetric techniques
IV - 20
-------
minimize the influence of these environmental factors on the application of oxygen
membrane electrodes.
Recently developed specific ion electrodes offer the opportunity for
monitoring a variety of anions and cations. Solid state specific ion electrodes
have been found to be more applicable for water quality monitoring than liquid
ion exchange electrodes. In most applications sample conditioning may be
required to maintain a given pH range or mask certain interferences.
IV - 21
-------
VI BIBLIOGRAPHY
1. Wezernak, C. T. and F. C. Polcyn, in "Instrumental Analysis for Pollution
Control" editor K. H. Mancy, VIII, 165, Ann Arbor Science Publishers, Inc.
(1971).
2. Proceedings of International Symposium on Remote Sensing of Environment,
The University of Michigan, Ann Arbor, Michigan (1969, 1970, 1971 and 1972).
3. Wezernak, C. T. and F. C. Polcyn, Proceedings of the 25th Purdue Industrial
Waste Conference, Purdue University, Indiana (May 1970).
4. Allen, H. E., in "Instrumental Analysis for Pollution Control" editor K. H.
Mancy, VI11> 135» A*10 Arbor Science Publishers, Inc. (1971).
5. Proceedings of Technicon Symposia on Automation in Analytical Chemistry,
published by Mediad Inc. 202 Mamaroneck Ave. White Plains, N. Y. 10601 (1966-
1972).
6. Mancy, K. H. and T. Jaffe "Analysis of Dissolved Oxygen in Natural and
Waste Waters", Public Health Service Publication No. 999-WP-37 (1966).
7. Schmid, M. and K. H. Mancy Chimia, 23:398 (1969). (Switzerland)
8. Schmid, M. and K. H. Mancy Schweiz L. Dydrol;, 32, 328 (1970) (Switzerland)
9. Pungor, E., K. Toth and J. Havas, Acta Chim. Acad. Sci. Huug. 48, 17 (1966)
10. ibid, Mikrochim. Acta, ^8, 690 (1966)
11. Pungor, E. Anal. Chem. _39_, 28A (1967)
12. Lingone, J. J., Anal. Chem. 40, 935 (1968)
13. Kelada, N. P. Ph. D. thesis, The University of Michigan, (1973).
14. McClelland, Nina I., and K. H. Mancy. J. Amer. Water Works Assoc. 64, 795
(1972).
15. Durst, R. A. (Editor) Ion Selective Electrodes, National Bureau of Standards
Special Publ. No. 314, U. S. Govt. Printing Office, Washington, D. C. (1965).
16. Koryta, J. Anal. Chim. Acta, 61, 329 (1972).
17. Severinghaus, J. W. In Handbook of Physiology, Respiration II, Amer. Physiol.
Soc., Washington, D. C. 61, 1475 (1965).
18. Kempen, L. H. J., H. Deurenberg and F. Krenzer, Respiration Physiol. North
Holland Pub. Co., Amsterdam, 14, 366 (1972).
19. Orion Research Inc. "Analytical Methods Guide" Angus (1973).
20. Schimpf, W. K, Ph. D. Thesis, The University of Michigan (1971).
IV - 22
-------
APPENDIX-fl
CONTINUOUS MONITORING BY
AUTOMATED CHEMICAL SYSTEMS
(Technicon Instruments Corporation
Tarrytown, New York 10591)
I - BASIC OPERATIONS
Sample
preparation
Digestion
Dilution
Dissolution
Dialysis
Extraction
Filtration
Maceration
Instrumental
transducers
Colorimeter
Fluori meter
Flame spectrophotometcr
Atomic absorption spectrophotometer
Electroannlytical systems
II - PARAMETERS FOR WHICH AUTOMATED
PROCEDURES ARE AVAILABLE
Acidity (total)
Alkalinity
Aluminum
Ammonia
Arsenic
Boron
Bromide
Cadmium
Calcium
Carbon dioxide
Chloride
Chlorine
Chlorine demand
Chlorine dioxide
Chromium
Cobalt
COD (see oxygen, chemical)
Copper
Cyanide
Fluoride'
Hardness
Iodide
Iron
Lead
Lithium
Magnesium
Manganese
Mercury
Nickel
Nitrogen (Kjeldahl)
Nitrates
Nitrite
Nitrogen (organic)
Oxygen (dissolved!
Oxygen demand (chemical)
PH
Phenols
Phosphate
Potassium
Selenium
Silica
Silver
Sodium
Specific conductance
Strontium
Sulfate
Sulfide
Sulfite
Surfactants
Synthetic detergents
Temperature
Turbidity
Zinc
IV - 23
-------
APPENDIX-fl
COMMERCIALLY AVAILABLE SPECIFIC ION ELECTRODES
J - SOLID STATE ELECTRODES
Beclcman Solid Stale Membrane Electrodes
Addresses: 2500. Harbor Boulevard, Fullerton, Calif., 92634, U.S.A. (P.O. Box 1, Glenrolhcs, Fife, Scotland).
Ion
F-
C!-
Br-
|-
S3"
Cua*
Model
39600
39604
39602
39606
39610
396121
Molar Range
10"
10°-
10°
10°
10°
jo-
-*jo-e
•5 x 10-B
— JO"
-io-B
— io-°
• (lower
limit)
pH Range Resistance (Megohms) Response Time Principal Interferants
(Seconds)*
0-*
0-*-
0--
0-'
0 —
0 —
33f
14
14
14
14
<5
< i
< 1
<0-25
<0-25
<
<
<
<
<
3
2
2
3
3
OH-
Br;l-;S2- and CM '
I-;S2- andCN"
S2'
Ag* and Kg**
Operative temperature range (°C) : —5 -» 100.
Overall size, length x diameter, (cm): 12-8 x 1-25.
* Defined as the time required to obtain a 90 per ccni response to a stcpchangc from 10"1 ~» !0~3 M concentration in a
stirred solution. Will also depend on concentration and viscosity of samples. For the copper electrode up to several minutes
may be required.
t At 10"' M fluoride.
} Like the Coleman solid state electrodes contains no internal filling solution.
Orion 94 and 96 Scries Solid State Specific ton Electrodes
Addresses: 11 Blackstone St., Cambridge, Mass., 02139, U.S.A. (E.I.L., Richmond. Surrey)^A".
Ion
CNS-
CN-
F'/La3*
F-/La"+
Na»
ci-
cr
Br-
I-
S- J
•<
Ag* 1
Cua+
Cda>
Pba*
Model Total Molar
Concentration
Range*
94-58
94-06
94-09
96-09
94-11
94-17
96-17
94-35
_ 94-53
94-16
* 94-29
94-48
94-S2
10" -* 10-"
10"
10"
10°
10'-
10"-
30°-
10°
30°
10°
10°
10'
— >
-*
— .
-5
•5
•5
— »
~*"
—
—
to-"
10-"
]0-o
X lO'6
x 10-6
x 10-6
x 10"e
10-'
io-7
lO'1
10-
;o-'
pH Range]
3 —
0 —
0 —
3 -»
0 —
0 —
0 —
0--*
0 —
0 —
0-*
1 -<•
2-»
14
14
11
11
12
13
13
14
14
14
14
14
14
14
Resistance (Megohms) Mentbrarc Material Principal Interferant ions
<30
< 1
<30
<200
<30
<30
< 10
1 --5
< 1
< 1
< J
< 1
AgSCN 4- ASaS S-- , Hga " and Cu3+ must
be abseht
Agl 4- A«5S S7" must be absent
Lap3 4.
LaF3 +
AgCI +
AgCI +
Agflr 4-
Agl 4-
Agj
Ag2S 4-
AgaS ,-
Ag,S-f
Fu*
F.u{
Ag!S
Ag,S
AfiaS
S
S
Cus
CdS
I'bs
•\
>OH- only interference
Ae*
T £
>S2~ itiust be absent
S2 " must be absent
S3" musrbc ab.»cnt
None as lai as examined
ilg3* mist be absent
S--, Ag* ami Hg2* must
be
abivnt
Af.\ !!g3' and Cu" must
Ay4, l!i^'1* and Cua* mu*
be ubsc.:t
.,
IV - 24
-------
S- LIQUID ION EXCHANGE ELECTRODES
Some Coming Liquid Ion Selective Electrodes
Addresses: Corninc Glass International, Mcdfold, Mass., 02052, U.S.A. (3 Cork St., London W.I)
(EEL, Hallstcad, Essex)
Ion Made!
Number
Cl- 476131
NOj- 476134
Caa* 476041
Caa+ -Mga* 476235
Working
Concentration
Range
JO-MO'6 M Nad
10°-10-8MKNO3
100-10-5MCaCla
100-10-BMasCa3*
Operating Temperature
pit Range Range (°C)
1-12 in
ID-'MNaCl
2-5-10 in
10-3MKNO3
5-10 in
10"3'M CaClj
5-10 in
)0-3MCaClj
10-50
0-50
10-60
10-60
Principal Inierferants
i- > cior >
NO3- «= Br-
cior > r
None given
(BaZ4=Sr:l* >
Overall size (length X diameter) cm : 12-3/12-7 x 1-5S.
Lifetime : 10-15 days (where quoted) before recharging with specified liquid ion exchanger.
Time response (s) : Generally < 60 for solutions differing by not more than ten-fold concentration
changes.
Resistance (Megohms) : 500 quoted nominally for Cast and Ca3* - Nfgs+ electrodes only.
Minimum sample size (ml) : 10 quoted for Ca2+ and Ca3* - Mg7+ electrodes only.
. |uPr'Snt m air for ^ort P*<"iods.
storage : |jon cxcnangCr jjquid removed for prolonged periods.
• Merely listed as such without any selectivity values.
Orion 92-Series Liquid Ion Exchange Membrane Electrodes
Addresses: 11 Blackstone Street, Cambridge, Mass., 02139, U.S.A. (E.I.L., Richmond, Surrey)
Ion
ci-
NO3-
BF«-
ClOr
Pb3+*
Caa+
Divalent
Ca3* -Mga+
Model
92-17
92-07
92-05
92-81
92-29
92-82
92-20
92-32
Molar
Activity Range
10- MO-
10-MO-5
10- MO-5
io°-io-s
10- MO'8
10-MO-8
lOMO'8
.CMC-
pH Resistance Principal Inierferants Mobile Exchanger
Range (Megohms) (where KMte > 1) SfVef
2-11
2-12
2-12
4-10/11
4-7
3-5-7-5
5-5-11
5-5-11
<30
<30
<30
<30
<30
<25
<»
Br-,1-, NO3-,CIO«- NR4*
and OH -
r,ClOs- andCIO4-
1-
OH-
Fea*
2naMButseeTabIe2) "
2n3+, Feat,Ni2t
and Cua+
•[Ni(phen)3]3*
kR-S-CHa-COa-
-(AlkylO)aPOa-
Operative temperature range (°C): 0-50.
Overall size, length x diameter (cm): 14-9 x 1-7 (slightly larger dimensions were once quoted).
Minimum sample size (ml): 3-5 in 50 ml beaker: 0-3 in disposable Orion microsample dish.
Reproducibility: Drift, repeatability and response time characteristics arc generally comparable with those of good quality
pH electrodes.
Storage: Can be air-stored or immersed in appropriate standardized ion solution.
Electrode life: About 1-3 months without renewal of ion exchange liquid.
Dollar cost: 195
* Now withdrawn from Orion 1969 Research Guide. lCat/lJ61).
t Iron replaces nickel in the ion exchanger material for
-------
Iff- HETEROGENEOU MEMBRANE ELECTRODES
I'ungor-Radclkis Heterogeneous* Ion Selective Kleclrmfes
f Protcch, 40 High St., Riekniansworih, Hcris (Advisory Services).
Addresses < Simac Instruments Ltd., Bridgeway Mouse, Bridge Way, NS'hitton, Mkldlcsox (Technical Services).
(_Radelkis Olcclrochcmical Instruments, Budapest 62, Hungary.
Ion
ci-
Br'
I-
sa-
so.1-
po,3-
F-
Loboratory
Model]
OP-C1-7111!
OP-Br-7Hl|
OP-I-7il^j
OP-S-7]If
**
**
**
Molar
Range
10-' -10-'
10"1 -* 10"'
10"1 — »• 10"'
io-1-*jo-1T
io-»-»io-5
10-' -* 1Q-5
Maximum
Resistance
10
10
10
10
Membrane
Material^
AgCI
AgBr
Agl
BaSO4
BiPO4
LaFs-CaFj
Principal
Jnlerferanls
Sa- ; I" ; Br
Sa-; I"
sa~
Operative temperature range (°C) : 1/5 -» 90
Overall dimension (length x diameter) cm : 11-5x1 for 711 Models.
Storage : Can be dry stored after a distilled water rinse.
Response time (s) : 15 -» 60 (Up to 3 min quoted by Rechnilz and co-workers for some electrodes.)
References 18 and 20.
Cost : $60
* The terms "homogeneous" and "heterogeneous" do not relate to the function, but to the composition of the electrodes.
(Pungor, Reference 102).
t Industrial number OP-Ion-722. The OP-lon-700 series carry side pin electrode connector plugs, while the OP-Ion-711
iypcs arc filled with shielded cabling.
} Pure siliconc rubber > II x id9 ohms.
II Dispersed in siliconc rubber matrix (50 \vt per cent of silver salt).
U These four electrodes may be used to assay Ag" ions after appropriate calibration but do not necessarily cover the same
range.
** Not commercially available.
IV
26
-------
SECOND CONFERENCE ON ENVIRONMENTAL QUALITY SENSORS
NATIONAL ENVIRONMENTAL RESEARCH CENTER
LAS VEGAS
OCTOBER 10 and 11, 1973
CONTINUOUS MONITORING BY ION SELECTIVE ELECTRODES
Julian B. Andelman, Ph.D
Professor of Water Chemistry
Graduate School of Public Health
University of Pittsburgh
Pittsburgh, PA 15261
ABSTRACT
The use of ion selective electrodes for analyzing natural and treated
water, as well as sewage and industrial wastes, has been increasing.
Among the species that have been so determined are fluoride, chloride,
sodium, ammonia, nitrate, calcium, total hardness, sulfide and cyanide.
Although ion selective electrodes are potentiometric sensors and are
thereby inherently adaptable for continuous monitoring, they must be
used for this purpose with care because of such concerns as interferences,
the need for pH adjustment and reference electrode instability.
The addition of a variety of pH buffers, salt solutions to control
ionic strength, chelators to negate the effect of interfering metals,
and ion indicators has been utilized to improve the capabilities of these
measurements. However, in order to adapt such procedures for continuous
analysis requires a more sophisticated monitoring system than simply the
insertion of a sensing and reference electrode into a waste stream or
natural water.
Such sophisticated monitors have been developed and are commercially
available using ion selective electrodes as sensing devices, along with
the addition of various chemicals to optimize the procedures. This paper
discuss techniques that are utilized in such systems, giving as an
example the cyanide monitor for analyzing waste waters.
IV - 27
-------
Ion selective electrodes have been available commercially for several
years (1). They have been used increasingly in analyzing water and waste
water samples in both field and laboratory applications. Most of these
uses, however, involve the analysis of single samples, rather than contin-
uous monitoring. This paper will focus on their capabilities for the
latter, discussing the problems and advantages in such use. Commercially
available ion selective monitors will be described.
One of the most important characteristics of ion selective electrode
systems relating to their use in continuous monitoring systems is their
dc voltage output. This is shown in equation 1, relating the electrical
potential read-out to the concentration, C , of ion "x", which is being
,n.
sensed by the detecting electrode in conjunction with a reference
electrode.
E(mv) = E + b log (C Y + k C Y + ) (1)
<. J a 5 >. x x y y y. J ^ J
The electrode inherently senses activity, rather than concentration, so
that CY must be multiplied by Y the activity coefficient, in this
.A. A
equation. Also, the electrode, designed to sense ion "x", may sense other
ions, "y" etc., so that their activities must also appear in equation 1
as additive terms, each being multiplied by its selectively constant ky, etc,
The coefficient b is primarily a function of temperature and the charge
on the principal ion being sensed, while Ea is determined principally by
the nature and construction of the reference and ion selective electrodes.
These variables in equation 1 must be carefully controlled or considered
in designing an ion selective electrode monitoring system. In this regard,
IV - 28
-------
Table 1 lists the major factors affecting each of these variables. Since
many ions being sensed are involved in equilibrium reactions, their
concentrations CY and C.. may be influenced by temperature, T, ionic
A y
strength, I, pH, and the presence of complexing agents. The sulfide and
fluoride electrodes are good examples of such an effect of pH} since the
weak acid forms HF5 H2S, and HS~ will be formed as the pH decreases.
Similarly, fluoride may be readily complexed by aluminum(III) and thus
its free ion, which is that sensed by the electrode, reduced in concentration,
This was shown, for example, by Harwood(2), who compared the results of
using EDTA and CDTA to chelate aluminum so as to minimize its complexation
of fluoride. Table 2 presents Harwood's data comparing the relative
efficiencies of these two chelating agents to improve the ability of the
electrode to sense the total fluoride in the presence of varying amounts
of added aluminum. It is apparent that CDTA is more effective, and that
certainly, without the addition of such chelating agents, it is apparent
that less fluoride would be detected than that present in the solution.
The preceding discussion indicates that T, I, pH and the presence of
natural complexing agents can all effect the free concentration of the
ion being sensed. These must be controlled, therefore, in a continuous
monitoring system in order to accurately monitor concentration. This may
be done by the addition of pH buffers, inert salt solution to control
ionic strength, and chelating agents if the interfering ions- are metals.
Similarly, the temperature in the sensing system can be controlled. Also,
the extent to which other ions may be sensed, as reflected by ky, etc.
must be considered. For the fluoride electrode, the only such ion-in
fresh waters that may interfere is hydroxide, but this is also readily
IV - 29
-------
controlled by pH (1). However, for other electrodes, such as calcium, it
is necessary to consider the value of ky for magnesium. Again, in natural,
fresh waters the additive term for magnesium in equation 1 is unlikely to
be significant (1).
Ea, as indicated in Table 1. can be affected by T, I. and the nature
3.
and construction of the ion selective and reference electrodes. Although
temperature and ionic strength can be controlled in a monitoring system,
there exists a serious likelihood of potential drift due to the inherent
characteristics of the two electrodes in the system. It is, thus, desirable
to choose electrodes with minimum drift characteristics, but also to
standardize the system periodically.
Even when there is no complexation or equilibrium mechanism to affect
the concentration of a free ion being sensed, variations in I, the ionic
strength, can significantly affect the activity coefficient, y > as
indicated in Table 1. This effect on Ymay be even more important than on
Ea or Cx, but again may be readily controlled by addition of an inert salt
solution to the continuous monitoring detection system.
Finally, the b variable of equation 1 is of considerable concern in a
monitoring system. As indicated in Table 1, it is affected by temperature,
the nature of the ion selective electrode, and the charge of the ion being
sensed. For example, at 25°C the absolute value of b is typically 59 mv
for a univalent ion electrode and 29.5 mv for a divalent one (1). However,
b can vary, even when the system is selected and temperature is controlled,
again indicating a need for calibration and standardization.
An ion selective electrode system for monitoring cyanide in waste waters
has been described which increases the value of b and, therefore, results
IV
_ 30
-------
- 4 -
in a more sensitive system, since greater values of b cause a larger voltage
change for a given change in concentration (3). This technique is of
particular interest because it has been incorporated into a commercially
available monitoring system. As discussed by Riseman (3), the conventional
ion selective elebtrode system for cyanide at 25°C will produce a voltage
E such that
E(mv) = Ea - 59 log (CN-) (2)
with (CN~) referring to the cyanide concentration, assuming that ionic
strength is maintained constant. However, cyanide may also be detected
using an indicator technique with the silver-sulfide electrode whose potential
follows the relationship
E(mv) = Ea + 59 log (Ag+) (3)
by adding a known concentration, C, of silver cyanide complex to the
solution. The free silver ion concentration in solution will then be
controlled by the variable cyanide concentration through the mechanism of
the silver-cyanide complex ion equilibrium, with
Ag+ = C/g2(CN-)2 (4)
Substitution of equation 4 into equation 3 gives
E(mv) = Ea + 59 log (C/%(CN-)2) (5)
which may then be arranged into the form
E(mv) = E' - 2x59 log (CN~) (6)
3.
IV - 31
-------
- 5 -
A comparison of equation 6, resulting from this indicator electrode
technique, with equation 1 describing the conventional cyanide electrode
system, indicates that the former has a slope, b, twice as large. This
results in greater precision, but also, because the linear relationship
between E and log (CN~) continues at lower concentrations, the sensitivity
of the indicator technique is a few ppb of cyanide, compared to about
50 ppb by the conventional cyanide electrode method (3),
It was also pointed out that if the waste water contains nickel or
copper, it will be necessary to release the cyanide from their complexes
by acid treatment with EDTA, followed by the addition of sodium hydroxide,
with the formation of free cyanide and binding of the copper and nickel by
EDTA, removing them as interferences. This indicator method for cyanide,
incorporating the acid-heating pre-treatment and EDTA addition, followed
by alkali and indicator complex, has been incorporated into a commercially
available cyanide monitor (3). The block diagram of the chemical sensing
system of this monitor, manufactured by Orion Research Incorporated, is
shown in Figure 1 (3). The acidified EDTA (Reagent 1) is mixed with the
sample at point A, heated at B, air bubbles removed from the stream at C,
the stream then being pumped to D, where it is mixed with alkaline silver
cyanide indicator (Reagent 2), and introduced with dynamic mixing into a
thermostatted electrode chamber, F, containing the sensing and reference
electrodes. The system also has provision for periodic introduction of a
1 ppm standard cyanide solution at G. Also of interest is the fact that,
rather than using a conventional reference electrode, such as calomel, it
uses a sodium electrode which is exposed to a constant and high concentration
IV - 32
-------
- 6 -
of sodium from the sodium hydroxide addition, thus reducing the instability
that is often associated with reference electrode systems.
The techniques incorporated in this cyanide monitor are also represented
in a range of several ion selective electrode monitoring systems manufac-
tured and described by Orion Research Incorporated as their series 1000 (4).
As with the cyanide monitoring system, they use reagent addition techniques
to reduce electrode and method interferences and control ionic strength,
and incorporate ion selective electrodes like sodium as reference electrodes
to reduce potential drift. They also provide for filtration of the sample
stream to remove suspended solids, and have temperature control and automatic
periodic standardization. It is reported that drift of readings typically
does not exceed 2 to 3% per 24 hours, and precision is better than + 5%
over four decades of concentration (4). These monitor systems have been
designed to eliminate the variabilities and interferences that would
otherwise reduce the applicability of ion selective electrodes as continuous
monitors for water and waste water.
The systems currently available commercially as Orion series 1000
monitors are listed in Table 3 as examples of the types of monitors capable
of analyzing water or wastewater (4). The concentration range for each
monitor is also shown, as well as possible interferences and typical stan-
dardization concentrations. Typical specifications for the series as a
whole are shown in Table 4 (4). Although these monitors have become available
only recently, one early report on the use of the fluoride monitor of this
series indicates that the various components performed well during test,
the automatic calibration feature performed according to specifications, and
there was little baseline drift (5).
IV - 33
-------
- 7 -
There is little doubt that ion selective electrode technology will
continue to improve, and that an increasing number of ions will be capable
of being monitored by such electrodes. The monitors described above, as
well as those available from other manufacturers, attest to this. Also
new chemical techniques utilizing existing electrodes, such as the indicator
method, will also increase their versatility. Such monitoring systems will
provide a welcome addition to the range of techniques available for water
and waste water analysis.
REFERENCES
1. J.B. Andelman, "Ion selective electrodes -- theory and applications
in water analysis." J. Water Pollution Control Feder., 40, 1844
(1968).
2. J.E. Harwood, "The use of an ion-selective electrode for. routine
fluoride analyses on water samples." Water Research, 3, 273 (1969)
3. J. H. Riseman, "Electrode measuring techniques for measuring cyanide
in waste waters." Amer. Lab., 4_ (12), 63 (1972). '
4. Orion Research Incorporated, Newsletter -- Specific Ion Electrode
Technology, V, Number 1, 1973; also various monitor specifications.
5. T.F. Craft and R.S. Ingols, Wastewater Sampling and Testing Instru-
mentation., Georgia Institute of Technology, Technical Report
No. AFWL-TR-73-69, July 1973.
IV - 34
-------
thermos tatted
electrons
chamber
4-cNannet
proportional
pump
"1
LOU
1 ppm sample
standard
Figure 1. Block diagram of sensing system of
typical Orion Model 1100 monitor (4)
Table 1. Effects on Electrical Potential For
Ion Selective Electrode System
Variable
Major Effect on Variab1e
T, I, pH, complexing agents
T, I, reference electrode,
ions selective electrode
T, ion selective electrode,
charge on ions sensed
Y Y
x'x
Ion selective electrode, T,
nature of ions sensed
IV - 35
-------
Table 2. The Effect of EDTA and CDTA on Overcoming Aluminum Interference
with Fluoride Analyses of 1 mg F/13 Using Different Aluminum Concentrations (2)
Aluminum
con cent rat i on
(rag/1)
Complexing
agent
EDTA
CDTA— 1 g/1
0.2 0.4
Fluoride
0.94 0.79
0.99 0.99
0.7
result
0.66
0.98
1.0 2.0
(mg F/l)
0.62 0.42
0.98 0.96
3.0
0.35
0.95
Table 3. Currently Available (Sept. 1973) Orion Series 1000
Ion Selective Electrode Monitoring Systems
Species
NHJ/NH3
ci-
oci-/ci2
F-
Hardness
Range (mg/1)
1-100
3.5-3550
0.05-100
0.05-1000
0.05-100
0.1-10
0.01-100
i-io,ooo
Interferences
Volatile amines
None
Br-,I"
None
Au+3jAg+
None
S=
C104",C10 "
Typical
Standardization
Concentration (mg/1)
10
10
1
Variable
1
1
Variable
Variable
IV - 36
-------
Table 4. Some Typical Specifications of Orion Series 1000
Ion Selective Electrode Monitoring Systems
Accuracy:
Precision:
Response Time:
Sample Supply:
Sample Temperature:
Reagent Consumption:
Stand. Sol'n Consumption:
Frequency of Automatic
St andardi z at i on:
Size:
Weight:
Power Requirements:
+.10%
Better than +_5%
99% response to change at inlet:
8 minutes
90% response to change at inlet:
6 minutes
100-1000 ml/min
0°-60°C
0.22 ml/min
50 ml/stand, cycle
one/day
33 in high x 27 in wide x 14 in deep
220 Ib
115V, 60Hz, 100 watts
IV - 37
-------
A REGIONAL WATER QUALITY MONITORING SYSTEM
RUSSELL H. SUSAG, PH.D., P.E.
MANAGER OF QUALITY CONTROL
METROPOLITAN SEWER BOARD OF THE TWIN CITIES AREA
PRESENTED AT THE
U. S. -ENVIRONMENTAL PROTECTION AGENCY
SECOND CONFERENCE ON ENVIRONMENTAL QUALITY SENSORS
NATIONAL ENVIRONMENTAL RESEARCH CENTER
OCTOBER 10-11, 1973
INTRODUCTION
"I often say that if you can measure that of which you speak
or can express it by a number, you know something of your
subject; but if you cannot measure it, your knowledge is
meagre and unsatisfactory."
Lord William Thomson Kelvin, 1824-1907
"Maintenance of high water quality standards cannot be assured
unless there is a stringent operating program with constant
surveillance of river conditions and means to provide quick
and appropriate ameliorative action. Even the most skilled
and experienced sanitary engineer or hydrologist cannot
determine the quality of water just by looking at it. All
the latest scientific aids must be utilized to prevent
pollution."
Metropolitan Development Guide - Sanitary Sewer Section,
Metropolitan Council of the Twin Cities Area
January 22, 1970
The Metropolitan Sewer Board of the Twin Cities Area is a seven-
county regional agency created by the 1969 Minnesota State Legislature
" for the protection of the public health, safety, and welfare of
the area, for the preservation and best use of waters and other natural
resources of the state in the area, for the prevention, control and
abatement of water pollution in the area, and for the efficient and
economic collection, treatment and disposal of sewage " The
Metropolitan Sewer Board provides wastewater treatment for 1.7 million
residents of 90 communities that comprise 85% of the Minneapolis-St. Paul
IV - 38
-------
metropolitan area. Twenty-four wastewater treatment plants are operated,
ranging in size and complexity from a 70,000 gpd two-stage stabilization
pond-percolation system to a 220 mgd step aeration modification of the
activated sludge process. Further, the Metropolitan Sewer Board will
place into operation in November 1973 the first full-scale physical-chemical
treatment plant, a 600,000 gpd treatment facility providing phosphorus and
nitrogen removal in addition to BOD and SS removal.
When the Metropolitan Sewer Board began operation in 1970, the
agency took over the operation of 33 wastewater treatment plants from
individual communities and from sanitary sewer districts in the metropolitan
area. At that time, there was very little, if any, coordination between
the operation of these facilities. The primary consideration in location
of these treatment plants was political boundaries and water quality
suffered amid inefficiently operated facilities and counter charges of,
"It's their fault". The Metropolitan Sewer Board, demonstrating a regional
water management concept, is consolidating operations by phasing out over-
loaded and poorly located treatment facilities and diverting flow to
regional facilities. The prime consideration now is the protection of
the water resources of the area within the framework of efficient economic
wastewater treatment rather than merely the treatment of each specific
community's wastewater.
HATER QUALITY MONITORING
Uses for Data
There are three basic uses for water quality data obtained from a
monitoring program. These are:
IV - 39
-------
1. Planning. Water quality data is essential to the determination
of location and degree of treatment for wastewater treatment
facilities where the receiving water presents a water quality
limiting situation.
2. Surveillance. The effectiveness of water pollution control
facilities is measured by the determination of maintenance of
water quality standards. Use of this data by a regulatory
agency would provide the basis for enforcement action.
3. Operation Control. Wastewater treatment plant removal could
be varied to meet the varying conditions of incoming receiving
water quality in situations where removals in excess of
secondary treatment are required.
Planning
The monitoring of the quality of the water resources of the
Twin Cities Metropolitan Area on a routine basis dates back 47 years.
In 1926, a program of river sampling was begun by the Minnesota
State Department of Health to determine the quality of the Mississippi
River and the impact of wastewater discharges from the cities of
Minneapolis and St. Paul. This sampling program was taken over in
1927 by one of the predecessors to the Metropolitan Sewer Board, the
Metropolitan Drainage Commission. A program of sampling and analyses
at 16 stations over a 100 mile reach of the Mississippi River was
conducted to get background material for the purpose of designing
and locating a treatment plant for Minneapolis and St. Paul. Weekly
samples were analyzed for the routine pollution measures of
temperature, pH, turbidity, dissolved oxygen, BOD, and coliform
IV - 40
-------
organisms. The original Minneapolis-St. Paul (Metropolitan) treatment
plant was placed into operation in 1938 and the sampling program
initiated in 1926 was continued by the Minneapolis-St. Paul Sanitary
District in order to monitor the effect of the institution of
wastewater treatment on Mississippi River water quality.
Water quality standards have been changed twice since 1938 and
the data generated from this sampling program have served as the
design basis for plant expansion programs. The Minnesota Pollution
Control Agency is in the process of further revising water quality
standards and this background data is serving as a basis for load
allocation studies. The Metropolitan Sewer Board as a regional
agency can locate treatment works such as to provide the best
protection of the water resources of the area and thus the- Sewer
Board is less governed by political boundaries than individual
communities would be. Thus, a continuing water quality monitoring
program is essential to a continuing planning program.
Surveillance
The Metropolitan Sewer Board is charged with the responsibility
of maintaining the quality of the water resources of the area so as
to allow for beneficial uses by the citizens of the area. In order
to accomplish this, the Metropolitan Sewer Board is expected to
carry on a program of water quality monitoring to demonstrate that,
- in fact, water quality standards are being maintained. From the
standpoint of the operating agency, water quality monitoring from
a surveillance standpoint has further value in allowing for the
IV - 41
-------
evaluation of the effectiveness of a water pollution control project.
The requisite degree of treatment to meet water quality standards is
most often determined from mathematical models that are based upon
theoretical considerations or upon empirical relationships developed
elsewhere. A water quality monitoring program to provide surveillance
after the fact will allow for evaluating the tools used to predict
requirements. Necessary modifications can be made so that future
planning is effective.
The continuation of the sampling program, begun in 1926 on the
Mississippi River, is an example of a water quality monitoring program
begun for planning purposes that was continued for surveillance purposes.
In addition, the Minneapolis-St. Paul Sanitary District instituted a
program of automatic river water quality monitoring and telemetering of
data to a central processing point in conjunction with a demonstration
program for regulation of combined sewer overflows. Five automatic
monitors with capability for sensing six parameters (temperature, pH,
conductivity, dissolved oxygen, chloride, and oxidation-reduction
potential) were sited on the Mississippi River to evaluate the effective-
ness of a program of controlling combined sewer regulator overflows.
These monitors were operated for a period of three years during the
course of the demonstration program. Four of the five original monitors
are being relocated to new sites that will be more compatible with the
goals of an overall river monitoring program.
From the standpoint of a regulatory agency, water quality monitoring
for surveillance purposes becomes an enforcement tool. The U. S.
Environmental Protection Agency placed four automatic water quality
monitors on the Mississippi and Minnesota Rivers to measure the progress
IV - 42
-------
of the metropolitan area in complying with the recommendations of
an enforcement conference that was convened in 1965. The EPA has
continued to operate three of these monitors to the present date.
Most of the recommendations of the enforcement conference have since
been accomplished and the EPA has loaned two of these monitors to the
Metropolitan Sewer Board for use in the Sewer Board's water quality
monitoring program. These monitors will be an important part of the
Sewer Board's program because they were sited at points that are
significant from the standpoint of water pollution control surveillance.
Operational Control
A third use for water quality monitoring data is that of providing
information that can be used for decision making in operational control.
As has been the case with most states, the State of Minnesota is in
the process of revising water quality standards to conform with the
requirements of the 1972 Amendments to the Federal Water Pollution
Control Act. The upgrading of standards will, in many cases, necessitate
the installation of tertiary treatment units or will require the initiation
of operating practices that fall beyond the realm of what has been defined
as secondary treatment. Tertiary treatment measures will not be required
at all times to meet the upgraded water quality standards because seasonal
variations in flow, temperature, and dissolved oxygen levels will result
in variable receiving water assimilative capacities. During the low flow -
high temperature river water conditions of August, a high degree of
treatment may be required whereas secondary treatment would be adequate
during the remainder of the year. Because of the high energy and resource
consumptive nature of tertiary treatment practices, it would be desireable
IV - 43
-------
to limit their use to situations and conditions that warrant such
extraordinary measures.
The Hinnesota Pollution Control Agency has enacted, as a provision
of its standards, the requirement for operation of all primary and
secondary units of treatment works at their maximum capability.
Variable operation of tertiary treatment units is allowed provided
that water quality standards are met and that adequate monitoring
capability is provided by the wastewater treatment agency. Thus, if
chemical coagulation and filtration following secondary treatment
were needed to provide the level of treatment necessary to meet
standards during the infrequent low flow - high temperature critical
periods, these tertiary units could be operated on an interim basis
provided that monitoring is practiced such that the upstream flow and
quality characteristics could be determined'and transmitted to the
operating agency so as to predict the necessary level of treatment to
maintain the water quality standards. The use of a water quality
monitoring program will allow for maximization of treatment plant
capacity and optimization of the use of resources.
Any one of these uses of water quality data provides justification
for a water quality monitoring program. It is felt that there will be
continuing demands for data for all three uses. The requirements for
each use are not all the same but are compatible and can be incorporated
into a single monitoring program.
Practical Considerations in Siting Automatic Hater Quality Monitors
Although theoretically monitors should be located at sites that provide
the best information as to water quality and at the best points from the
IV - 44
-------
standpoint of monitoring impact of water pollution control facilities,
in point of fact several other considerations govern. These are:
accessibility, availability of utilities, security, and representative-
ness of sample. The following comments present Metropolitan Sewer Board
experiences in locating and operating water quality monitors:
1. Accessibility. In most cases pumping restrictions require the monitor
to be located near the water to be monitored. In some of the locations
in the Sewer Board's program, the monitor has been placed on or adjacent
to the flood plain, requiring periodic removal. One of our former
locations was on the river side of a flood levee (constraints by
political bodies and governmental agencies necessitated this location).
Removal of this trailer monitor during high water required the use of
a truck crane.
2. Availability of Utilities. Proximity to electric power and telephone
service is a necessity. At the present time, two monitors are sited
on dams in situations where it is not possible to telemeter data
because of the absence of telephone lines. The data is being punched
on paper tape by a digital recorder. (These monitors have as their
primary purpose data gathering for surveillance uses.)
3. Security. Vandalism has proved to be a problem at some of our
locations. At the EPA monitor on the Minnesota River, it was
necessary that a pile cluster be driven in the river so that the
intake could be located in the river. A navigational light on this
pile cluster is frequently shot out by vandals.
One of the Sewer Board's classic examples of vandalism occurred
at a site where a trailer monitor was located on an old bridge
abutment. This abutment was below the level of the existing bridge
IV - 45
-------
and vandals dropped rocks onto the top of the trailer monitor severing
the electrical connection and the intake line. The last straw was
when vandals broke into the trailer pulling all sensors, signal
conditioners, and telemetering equipment out of the monitor.
In contrast, the Sewer Board had one trailer monitor on the
Mississippi River sited along a very busy thoroughfare in plain view.
This trailer has not been touched. Here is a case where security was
gained not from isolation but rather from public exposure.
4. Representativeness of Sample. Location of the intake pump at a point
in the river that provides a representative sample is a primary consider-
ation. Fortunately, water intakes, bridges, docks, and dams generally
provide representative sites as well as meeting the previously mentioned
qualifications. Placement of pile clusters in rivers to allow for
sampling in the main channel of a river can be an impediment to
navigation and is frowned upon by the U. S. Corps of Engineers. One
site, presently under consideration, is at a water supply intake
structure. This site meets all qualifications but it will be necessary
to forego data gathering during spring ice breakup conditions when the
intake is closed.
METROPOLITAN SEWER BOARD HATER QUALITY MONITORING PROGRAM
The Metropolitan Sewer Board is in the process of developing a water
quality monitoring network that can be used not only to generate information
relative to maintenance of water quality standards but also to generate
flow and quality data that can be used to regulate treatment process
measures. In order to coordinate wastewater treatment operations with
IV - 46
-------
maintenance of water quality standards, it is necessary to obtain flow
data as well. The Metropolitan Sewer Board has entered into a cooperative
program with the U. S. Geological Survey to develop a water quantity-quality
monitoring network. The USGS maintains four gauging stations on the three
major rivers of the metropolitan area. Water quality monitors are being
sited at three of the four major river gauge stations in addition to six
other sites. The data will be transmitted over leased lines to a central
data acquisition computer (Honeywell Model H-316) operated by the Metropolitan
Sewer Board. In addition, the data will be tape punched on digital recorders
so that a continuing record of variations in water quality can be maintained.
Analog data generated by the on-site flow meter or water quality sensor
is converted at the site into a proportional time pulse signal of duration
0.2 to 13.33 seconds. The conversion is repeated every 15 seconds. Up
to 15 flow meters or water quality sensors are transmitted over a single
Bell System 3002 voice grade line to the central computer. Water quality
values- are converted to engineering units through the algorithm V = K] + K£
(T-.2)tN. K], «2 and N are assigned by the computer operator using a least
squares fit program on data collected during site calibration. Sixty- 15 sec
samples are collected every 15 min, screened for extreme values, reduced to
a 15 min average value, and stored on a disc. The 15 min values are stored
for the current operating day plus 3 days previous. An operator-initiated
program reduces the 15 min data to quality mean, minimum value and time it
occurred, maximum value and time it occurred. These reduced summaries are
stored on the second or history disc for retrieval at any time. The history
disc has capacity to store one year's data in the reduced form.
The automatic monitoring system will be composed of a mixture of new
IV - 47
-------
and existing equipment. The EPA has transferred, on loan, the operation
of two of their water quality monitoring stations. These have been
interfaced with the Metropolitan Sewer Board's data acquisition system
and the EPA still maintains contact with these through a recorder.
Weekly reports are sent to the local Water Quality Office of the EPA.
Additionally, the Metropolitan Sewer Board has two operating water quality
stations on the Mississippi River and one has been sited on the Minnesota
River in cooperation with the USGS. Thus, at the present time the
Metropolitan Sewer Board has a network of five water quality monitor
stations, two on the Minnesota River defining a 36 mile river reach and
three on the Mississippi River over a 22 mile reach. During this year,
two additional monitors will be located upstream on the Mississippi River
and a monitor will be placed on the St. Croix River. This will provide
the Sewer Board with a network of eight monitor stations spanning 110 miles
of the three major river systems in the Twin Cities Metropolitan Area. The
EPA still maintains one monitor at the downstream extremity of the metro-
politan area.
The original monitors of the Metropolitan Sewer Board are Fairchild
monitors, which company was purchased by Automated Environmental Systems
which then passed into the hands of Raytheon Corp. The EPA monitors are
Schneider monitors and the new monitors that are being sited on a
cooperative basis by the USGS and the Metropolitan Sewer Board are
Ionics, Inc. monitors. Maintenance of these monitors is the responsibility
of the Metropolitan Sewer Board. Technical support is provided for mainten-
ance both by staff of the Metropolitan Sewer Board and by USGS staff.
The frequency of servicing of monitors varies by location. Those river
IV _ 48
-------
locations that have the highest level of organic material require the
most frequent servicing (up to twice a week). At some locations, during
portions of the year, the water is of such outstanding clarity that only
infrequent servicing is necessary. As a general rule, the monitors are
serviced once a week. The pumping system is the principal continuing
problem. As yet no substitute has been found for the Peerless submersible
pump.
In addition to operating this monitoring program, the Metropolitan
Sewer Board will be working with the USGS in developing procedures for
data presentation such that the data will be meaningful to the general
public as well as to the operating and regulatory agencies. The intent
is to use this system to control operations such that water quality
standards are maintained and such that this information can be disseminated
and understood by all concerned. Automatic monitoring systems will provide
the means by which there can be a maximum utilization of the financial,
energy, and natural resources such that the desired water quality standards
are maintained compatible with the designated beneficial uses of the water
resources in the true sense of water pollution control.
IV - 49
-------
FOREST UKE ^1 I M*sr»'l(.i*
'OB(S1L««I I
OLOSCMO
J IKOtfCWltKCf Ifk "t0"" X* 1 2
I _______ L_
L.c?»»A»r r^"
LWOCOHI. -
f CARVER CO.
I fo I
U»if!tlB6 \
-.T
j 1 HtWr**6Ulj I '-fcLKO I I I
METROPOLITAN SEWER BOARD
Wastewater Treatment Plants
>Treatment Plant in Operation During 1973
1 Anoka 9 Lakeville
2 Apple Valley 10 -Long Lake
3 Bayport 11 Maple Plain
4 Blue Lake 12 Medina
5 Chaska 13 Metropolitan
6 Cottage Grove 14 Hound
7 Farmington 15 Newport
8 Hastings
17 Orono
18 Prior Lake
19 Rosemount
20 St. Paul Park
21 Savage
22 Seneca
23 South St. Paul
16 Oak'Park Heights 24 Stillwater
25 Victoria
IV - 50
-------
METROPOLITAN SEWER BOARD
WASTEWATER TREATMENT PLANTS
(LOCATED ON MAJOR RIVERS)
H
<
I
Ul
NEWPORT
ST. PAUL PARK
BLUE LAKE
•
MINNESOTA
-------
WATER QUALITY SAMPLING STATIONS
1927-1973
H
ANOKA BRIDGE
MINNEAPOLIS WATERWORKS
WASHINGTON AVE. BRIDGE
LAMBERT LANDING
ABOVE METRO PLANT
ST. PAUL TERMINAL
SOUTH ST. PAUL
1-494 BRIDGE
INVER GROVE BRIDGE
FORD DAM
FORT SWELLING BRIDGE
FORT SMELLING PARK
MENDOTA
BRIDGE
SAMPLE LOCATIONS
DOWNSTREAM ON
MISSISSIPPI RIVER
SMITH LANDING
DIAMOND BLUFF
RED WING DAM
REDWING
FRONTEAIAC
LAKE CITY
READ'S LANDING
WABASHA
-------
PRESENT METROPOLITAN SEWER BOARD
WATER QUALITY SAMPLING STATIONS
H
<
U1
OJ
ANOKA BRIDGE
LAMBERT LANDING
FORD DAM
FORT SWELLING PARK
INVER GROVE BRIDGE
-------
EXCERPTS
FROM
STATE OF MINNESOTA POLLUTION CONTROL AGENCY'S
REGULATION WPC 15
CRITERIA FOR THE CLASSIFICATION OF THE INTERSTATE WATERS
OF THE STATE AND THE ESTABLISHMENT OF STANDARDS OF QUALITY AND PURITY
Section (c)(4)
"The highest levels of water quality, ,
which are attainable in the interstate waters by
continuous operation at their maximum capability
of all primary and secondary units of treatment works
or their equivalent shall be
maintained in order to enhance conditions for the
specified uses."
Section (c){8)
" If treatment works are
designed and constructed to meet the specified
limits given above (BOD5 = 5 mg/l, SS = 5 mg/l} for
a continuous discharge, at the discretion of the
Agency the operation of such works may allow for
the effluent quality to vary between the limits
specified above and in section (c)(6) (BODg = 25 mg/l,
SS - SO mg/l), provided the water quality standards
and all other requirements of the Agency and the
U. S. Environmental Protection Agency are being met/
Such variability of operation must be based on
adequate monitoring of"the treatment works and the
effluent and receiving waters as specified by the'
Agency.
Emphasis added.
IV - 54
-------
US. GEOLOGICAL SURVEY
RIVER FLOW GAGING STATIONS
tn
tn
-------
PROPOSED LOCATIONS OF
CONTINUOUS AUTOMATIC WATER
QUALITY MONITORING STATIONS
CTv
-------
. WATER QUALITY HISTORY
DATE HR MO NAME TEST
. 09/26 '00:00 QR06 DO
09/26 Oil00 QROS
...09/26 02:00 QR06
09/26 03:00 QR06
. 09/26 04:00 QR06
09/26 05*00 QR06
_Q9/26 06:00 QROS
09/26 07:00 QROS
.09/26 08:00 QR06
09/26 09:00 QROS
09/26 lOsOO QROS
09/26 11:00 QROS
09/26 12:00 BROS
09/26 13:00 QROS
09/26 14:00 QROS
09/26 15$00 QROS
. 09/26 16:00 QROS
.7 1 V" •
EST
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
; DO
VALUE
7.3
7.2
7.2
7.1
7.0
6.9
6.9
6.9
6.9
6.9
7.0
7.0
7.0
6.9
6.8
8.8
ANNUAL QUALITY SUMMARY FOR ,973 PROOUCED ^^
03
08
5.2
LOCATION QR06 TEST DO
MONTHLY MEAN VALUES:
01 02
06 9.0 07 7.1
11 12
.MINIMUM VALUE 0.1 ON 05/02 AT 11:00
MAXIMUM VALUE 10.2 ON 07/16 AT 1'5:00
WEAN FOR YEAR 7.9
STANDARD WAS EXCEEDED FOR 0 D«v"
04
09
-5.5
05
10..
12.7
-»,a£
-*•; M/25
09 J
Q 09'*•
cs
c« —
a*°6 M-
«Vl
"*S8
_QR05 ^
a*06- TP-
"tP
CURRENT QUALITY READINGS
' DATE HR UN NAME TEST VALUE
09/26 16:00 QR02 DO 4.6
' 09/26 16:00 SR03 .DO 8.1
09/26 16{00 QR06 DO
2
.A
r
ci'
DO
PH
TP
CH
DO
PH
rp
CJV
"••C4
«7.g
80s
-?-2
-*
* * * /
ffos
8.^
«7.S
- 804
o7*'
8.20
MONTHLY RIVER QUALITY REPORT FOR 08/73 PRODUCED 09/26/73
fi'sr MINIMUM 'DATE"HRMH MAXIMUM" DATE" HR MN " MEAN" IRDQ'XSTD
NAME TEST
QR02 CH
QR03 CN
QR06 CH
QR02 . ,DO._
QROS DO
QROS DO
QR02 PH
QROS PH
QR06 PH
nona TP.
322
261
708
0.4
2.0
3.4
5.03
5.46
7.41
70.9.
08/23
08/18
08/13
08/05
08/30
08/13
08/14
08/20
08/15
08/25
08/18
11:00
21:45
09:15
12:15
22:00
08:30
00:00
17:00
09:30
06:00
21:45
514
496
883
. 9.4
9.3
9.3
8.10
8.86
8.20
84.6
81.5
84.9
08/07
08/20
08/28
08/14
08/03
08/29
08/07
08/29
08/02
08/07
08/10
08/08
13:30
16:30
09:00
07:15
17:30
18:00
13:30
08:45
17:00
13:30
04:00
16:00
09/26/7J
05
10
25
3
30
EXAMPLES OF COMPUTER PRINT-OUT FORMATS
OF
AUTOMATIC WATER QUALITY MONITORING DATA
IV - 57
.4 .r
74;f
386
350
777
5.2
6.9
5.2
7.79
7.99
7.67
77.7
76.3
79.5
31
27
31
30
28
28
31
28
31
31
28
31
-------
WATER QUALITY MONITORING IN SOME EASTERN EUROPEAN
COUNTRIES
by
Peter A. Krenkel* and Vladimir Novotny**
* Professor Environmental and Water Resources Engineering,
Vanderbilt University, Nashville, Tennessee
** Assistant Professor of Civil Engineering, Marquette University,
Milwaukee, Wisconsin
IV - 58
-------
Introduction
With the new interest in water pollution control, it has become
obvious that an adequate continuous and dependable water quality monitoring
system will be mandatory in the future. The United States has probably been
the leader in automatic water quality monitoring as exemplified by the
ORSANCO system. In fact, other countries have looked towards the ORSANCO
system as a model in order to formulate their own monitoring systems.
The problems of pollution control are compounded when a water passes
an international boundary, thus making the need for continuous monitoring
of water quality a requisite part of pollution control activities. It is
interesting to note that while the United States has somewhat in the order
of five per cent of the world's population, it uses, on an annual basis,
more than 30 per cent of the world's resources. Americans do not recognize
the fact that conditions throughout the world are not similar to those
existing in the United States. For example, while most Americans do not
concern themselves with the bacteriological safety of a drinking water supply,
many countries in the world still have this fundamental problem, i.e., the
possibility of contacting a disease through drinking water.
Because of the world wide implications of maintaining adequate water
quality, the World Health Organization has instigated a program of support
for certain countries through its United Nations Development Program (Special
Fund endeavors). Eastern European countries involved in WHO proj ects of
this type include Poland, Czechoslovakia, Romania and Hungary.
It is interesting to note that Hungary has a particular interest in
water pollution control activities, inasmuch as over 90 per cent of their
IV - 59
-------
surface water supply crosses their borders from another country. Hungary is
interested in a project that has been contemplated by WHO which would
involve a comprehensive water management plan for the Danube River. The
complexities of a water basin management plan in the United States are
magnified on the Danube River, inasmuch as eight countries border the
river and would have to cooperate in developing such a plan, should it be
promulgated. These countries are Czechoslovakia, Romania, Hungary, Bulgaria,
Russia, Austria, West Germany and Yugoslavia.
Most countries, upon learning of the success of the ORSANCO system
in the United States, have indicated an interest in developing automatic
water quality monitoring systems. Table I, which was taken from a study
by the Economic Commission for Europe, shows the existing monitoring stations
and plans for future stations of some of the countries involved. The country
having the most experience with the installation and operation of automatic
water quality monitoring stations at the present time is Poland, inasmuch
as they have had some six years of support in this endeavor by the World
Health Organization.
Because of the avid interest demonstrated in automatic water quality
monitoring systems, the World Health Organization held a seminar on this
topic in Cracow, Poland in the spring of 1971. This conference delineated
the objectives, status and plans for automatic water quality monitoring
systems by each country represented. The proceedings of this conference was
published by Vanderbilt University in 1972 and contains much information that
is not readily available in the English language (2).
Subsequent to several days of deliberation, the conferees arrived at
the following objectives for automatic water quality monitoring stations.
IV _ 60
-------
Rfgion inclu*d
n UNDP ..POLAND "26"
divines ) \
GENERAL LOCATION OF AWOMS
FLORCZYK H.-.RESWRCH AND STUDIES ON THE LOCATION OF AUTOMATIC WATER QUALITY KONITOBNG
STATIONS -AWOMS
FIGURE 1
Compilation of actual and planned automatic water
quality monitoring stations in various countries.
No
Country
Number of Stations
existing or in the
construction stage
planned for the
future
1
2
3
4
S
6
7
8
9
USA
Sweden
Poland
GFR
Great Britain
Austria
Spain
Hungary
Holland
205
12
7
6
4
1
-
1
_
1200
30
S3
75
240
10
40 - 50
2
TABLE I
IV - 61
-------
1. Identification of compliance and non-compliance with water
quality standards
2. The establishment of water quality baselines and trends
3. The ascertainment of improvements in water quality resulting
from abatement measures
4. The implementation of flow augmentation
5, The detection of emergency water quality problems
6. The determination of the "offender" in cases of waste spills
7. The detection of natural causes of abrupt water quality changes
8. The establishment of water quality relationships that would
allow prediction of downstream quality conditions
9. The early warning of downstream users that adverse water quality
conditions are approaching their water intakes
It is interesting to note that the common objectives as delineated
by this meeting comprised of many European countries were quite similar
to those of the United States, which only serves to demonstrate the
world-wide need for common goals in terms of maintaining adequate water
quality.
The Polish Automatic Quality Monitoring System
In order to demonstrate the procedures which have been instigated
through World Health Organization projects, the Polish studies will be
utilized. Through the WHO projects, Poland has installed seven automatic
water quality monitoring stations throughout their country. These stations
have been located on the two major river basins in Poland as shown on
Figure 1. As will be shown subsequently, considerable care was given by
the Polish workers in establishing the location of these stations and
their procedures should be used as a guide for future endeavors of this
type.
IV - 62
-------
The stations established were equipped with Honeywell W-20 monitors,
each measuring temperatures in the range from -5 to 40°C., conductivity
in the range from 100 to 1,000 mhos, dissolved oxygen in the range from
0 to 15 parts per million, pH in the range from 1 to 14, oxidation reduction
potential in the range +500 mv , turbidity in the range from 0 to 500 ppm,
chlorides in the range from 1 to 1,000 ppm, water level and solar radiation.
The two psime considerations with respeet to location of the available
number of automatic water quality monitors were to locate where the water
was most important to the country's economy and to locate where irregularity
of water quality parameters precluded manual sampling techniques. The
general location was based on the state of pollution, the pollution character-
istics, water utilization and the need and existence of waste water treatment
plants.
In order to demonstrate the methodology that was used by the Polish
workers in choosing the sites designated, the studies on the Odra River will
be used, as described by Florcyk (3).
The first step in establishing the general location of the stations
along the Odra River was to establish profiles of water quality parameters
of interest. The detailed profiles developed are shown in Figures 2 through 7
which depict the changes in certain water quality parameters imposed on a
profile of the Odra River. The parameters used on the Odra River were BOD,
suspended solids, chlorides, phenols, sulphates, and dissolved solids. In
addition to the profiles of the parameters, the location and designation of
each waste load along the river profile is presented along with tributary
rivers. It can be seen that four stations were chosen on the basis of these
profiles as follows, Chalupki, Kozle-januszkowice, Opole-Wroblin and
Wroclaw.
IV - 63
-------
kJotwdy 1*>
tot
LOKALI2ACJA OC&NA ASPAVNA TIE PRZEBIEGU KRZYWEJ BZT5 WZOKUZ BIEGU RZEKI ODRY.
GENERAL LOCATION OF AWQMS ON TtC GROUND OF COURSE OF B005 CURVE ALONG ODRA
RIVER WATER COURSE
1
I
»
4
let
-^
y
r
)
1
4
-P
(
— r
.4
sli
" I*
II
no
•*».
|
<
4
i'
4
1
ISO
f
g
2
30
I
1
1?
f
f
250
J
yt
>
•«.
33)
too
1
r : 1
t
j
••-• — .
i
£50 SCO
J
550
ECO
650 TO
75(
LOKAUZACJA OGOLNA ASPJWNA TIE PR2EBEGU KRZYWEJ ILOSd ZAWE5IN WZK.UZ BEGU RZEKI ODRY
GENERAL UXATION OF «NQMS ON THE GROUM3 OF COURSE OF SUSPENDED SOCJDS QUANT1TES CURVE
AUDNG ODRA RIVER WATER COURSE.
FIGURES 2 & 3
LOKAUZACJA OGOLNA ASRWNA TLE PR2EBEGU KRZYWEJ STE.ZEN FENOLI WZDtUZ SEGU RZEKI OORY.
GENERAL LOCATION OF AWQMS ON THE GROUND OF COURSE OF PHENOLS CONCENTRATIONS CURVE
ALONS ODRA RIVER WATER COURSE.
b y
e ji ttO
55
Uon»*y n»l
1
1
o
J
s
4
1
s
>
_
*
i
j
K
j[s
ll
BO
1
1!
<1
0
S
MB
»
1
_
ii
a
30!
360
100
«MW
*ao
..
900
•
B
0
on
680 OT
WO
LOKAUZACJA OC&NA ASRWNA TLE 'PBZEBEGU KRZVWEJ 5f^2Eft CHLORKdW WZMUZ BEGU RZEKI OORY
GOCRAL LOCATION OF AWQMS ON THE GROLHJ OF COURSE OF CHLORIDES CONCENTRATIONS CURVE AU3NS
CORA RIVER WTER COURSE.
FIGURES 4 & 5
IV - 64
-------
f 1
3 Z
M
c a
II
a
LOKALBACJA OG6LNA ASHWNA TIE PRZEBEGU KRZYWEJ STEZEN SIARCZANOW WZDtUZ BEGU RZEKI CORY.
GENERAL LOCATION OF AWQMS ON Tt€ GROUND OF COURSE OF SULPHATES CONCENTRATIONS CURVE ALONG
OORA RIVER WATER COURSE.
klonvtin of ffatr
LOKAUZACJA CGCiNA ASfiW NA TIE PRZEBEGU KRZYWEJ STEZEN ZWIAZKoV ROZFUSZCZONKH WZOUJZ
BEGU R2EN CORY.
GENERAL LOCATION OF WQMS ON THE GROUND OF COURSE OF DISSOLVED SOLIDS CONCENTRATIONS
CURVE ALONG OORA RIVER WATER CORSE.
FIGURES 6 & 1
Przekrdj pomiarowo - kontrolny
Monitoring cross - section
CHAtUPKI
Rzeka
River
ODRA
0.41S,
NT piorxJw badawczych
No of vtrtlcals In cross-
section
ROZKkAD PReDKOSCI PRZEPkYWU WODY W PRZEKROJU CHAkUPK! NA RZECE ODRZE
DISTRIBUTION OF WATER FLOW VELOCITY IN CHAIUPKI CROSS-SECTION, N OORA RIVER
FIGURE 8
IV - 65
-------
The station at Chalupki was chosen because it shows the sources
originating from Czechoslovakia, and in addition, demonstrates wide varia-
tion in water quality parameter characteristics. The station at
Kozle-Januszkowice was chosen because of salinity problems introduced from
Olza, Ruda, Bierawka, Klodnica, the Upper Silesia industrial region and the
Rybnicki. In addition, a nitrogen plant in Kedzierzyn, the discharges from
the city of Raciborz, and the location of the minimum point on the oxygen
sag occurs in this location. It is anticipated that this station will
control salinity discharges from various storage basins in the area in order
to minimize chloride concentrations in the Odra River.
The station at Opole-Wroblin was chosen because of the upstream
existence of the two largest sources of pollution on the Odra River, a paper
mill in Krapowice and the city of Opole, other sources from the Opolski
Silesia and the control of saline discharges from storage basins "in the area.
The station at Wroclaw was chosen because it is the only city on the
Odra River using Odra River water as a potable water supply, it has the
largest industrial users of water, it was thought that the composition of these
waters could be controlled, an upstream tannery exists in Brzey and
finally, the monitoring station will allow insight into weir leveling in some
20 weir sites in 157 kilometers of impounded river.
Subsequent to choosing the general location of the monitoring station,
specific sites were chosen using the criteria of reasonable field conditions,
easy access by car, power connections, land implements and the possibility of
providing constant supervision from a nearby existing facility.
The chosen sites were then subjected to comprehensive cross-sectional
analyses which included velocity, dissolved oxygen, clarified oxygen consumed,
IV - 66
-------
mixed oxygen consumed, chlorides and dissolved solids. If these profiles
demonstrated uniform mixing, such that a representative sample would be
obtained from the site, the site was designated as a good one. Cross-
sections indicating non-homogeneity were rejected and studies made to
determine where complete mixing occurred. Examples of the cross-sectional data
taken on the Odra River are given in Figures 8 through 13, from which it
can be observed that the parameters measured were reasonably homogeneous
across the cross section, thus indicating a good monitoring site.
The data collection system consists of four basic sub-assemblies as
shown on Figure 14. These include the measuring module, telemechanical
equipment, transmission lines and terminal and registering equipment.
As previously indicated, the measuring module was a Honeywell W-20
instrument, the water being delivered by a screw-type pump. The major
problem with this unit was with regards to maintaining the intake velocity
the same as the river velocity. Temperatures are measured with thermistors,
conductivity is measured with a measuring potentiometer with an internal
temperature compensator, the dissolved oxygen is measured by a polaragraphic-
type probe with a gold cathode and a silver anode and includes a temperature
compensator, pH and ORP are measured with reference and measuring electrodes,
the turbidity utilizes a reflectant methodology, the chlorides are measured
similarly to ORP and pH, the flow level device uses a bourdon gauge and
and solar radiation is measured using a eppley-type pyranometer.
Transmission is comprised of two systems, a radio link and a radio
link plus telephone lines. The operating frequencies of the radio stations
are each different in order to avoid interference.
With regards to terminal and registering equipment, three methods
were considered as follows: inscription on paper tape or on punched tape, periodical
IV - '57
-------
Przekrdj pomlarowo - kontrolny
Monitoring cross - section
CHAtUPKI
Rzeka
River
ODRA
NT ptemJw badawaych
Mo at wrtlcoli In cro» -
Mctkm
ROZKfcAD ZAWARTOS'CI TLENU ROZPUSZCZONEGO W PRZEKROJU CHAtUPKt NA RZECE OORZE
DISTRIBUTION OF DISSOLVED OXYGEN CONCENTRATIONS IN OHAWJPKI CROSS- SECTION,
IN ODRA RIVER
FIGURE 9
Przekroj pomiarowo - konUolny
Monitoring cross - section
CHAtUPKI
Nrpfcndw badawczyeh
No of vtftkait In crtm-
MCtkMI
Rzeka
River
ODRA
ROZKtAD VmRTOSCI UTLENIALNOSCI Z PR6BEK SKLAROWANYCH W PRZEKROJU CHAKUPKI
NA RZECE ODRZE
DISTRIBUTION OF PERMANGANATE OXYGEN DEMAND VALUES FROM CLARIFIED SAMPLES
IN CHAtUPKI CROSS-SECTION. IN OORA RIVER
FIGURE 10
IV - 68
-------
Przekrdj pomlarosvo - kontrolny
Monitoring cross - section
CHAfcUPKt
Nr ptondw bodawaycti
Ho of v*rllcali In cnu-
iictlon
Rzeka
River
ODRA
ROZKkAD WARTOSCI UTLENIALNOS'CI Z PR(iBEK SKKiCONVCH W PRZEKROJU CHAkUPKI
NA RZECE ODRZE
DEMAND VAUJES ^ MIXED
FIGURE 11
Przekrdj pomiarowo - kontrolny
Monitoring cross - section
CHAtUPKI
M- pkxxJv. badawczych
No of v«f ttcais in cros> •
MCtion
Rzeka
River
ODRA
R02KLAD STE2EN CHLORKbW W PRZEKROJU CHAtUPKI NA RZECE ODRZE
DISTRIBUTION OF CHLORIDES CONCENTRATIONS IN CHAtUPKI CROSS-SECTION,
IN ODRA RIVER
FIGURE 12
IV - 69
-------
Przekrdj pomlarowo - kontrolny
Monitoring cross - section
CHAtUPKI
Hr plooA* bMawwych
No el v»H(cal» In em> •
MCtton
Rzeko
River
OORA
ROZKKAD STEZEti ZWIAZKdW ROZPUSZC20NYCH W PRZEKROJU CHAfcUPKI NA RZECE ODRZE
DISTRIBUTION OF DISSOLVED SOLIDS CONCENTRATIONS IN CHAkUPKI CROSS - SECTION,
IN ODRA RIVER
FIGURE 13
KRAK6W
WROCtAW
BLOCK DIAGRAM OF AWQMS SYSTEM
radio tronsmitttf
t*legraphic unit
TO teltmtliic tquipcnnK
yd data converiion unit
SI pertormanct unH
3 (nantorW-20
FIGURE 14
IV - 70
-------
inquiry and recording by teletype, and a complete data processing set which
would compute the necessary routines with the aid of a computer. The
option chosen was periodical inquiry and recording by teletype.
The data processing equipment performs the following operations:
1. Receiving 102 messages and 89 measured values from the memory
circuits of the telemetering systems and processing them
according to a given program. The reading time may be adjusted
to 5, 10 or 60 minutes.
2. Preparing an operational record which consists of the delivery
by the teletype of processed data from individual stations.
3. Preparing a disturbance record which is written by a separate
teletype.
4. Checking the incoming measured values by comparison with
preselected admissible limiting values.
5. Receiving via teletype the values of limiting intervals or
their changes.
The digital form of measured quantity may- be made available by wiring
a digital to analogue converter. The central processing center serves
presently for two purposes, one is to determine whether or not the concentration
of any parameter exceeds the permiss.ible level and the other is to gather
summary characteristics concerning certain periods of time (e.g., one day)
for long-range studies on the degree of water pollution in a given basin
area.
The four basic units comprising the telemechanical equipment are a
transmitter for measured values and messages, a receiver for measured values
and messages, a transmitter of commands and a receiver of commands.
The first W-20 instruments were placed into operation during December
of 1968. Table 2 demonstrates the results of the operating performance of
the difference sensors. The average reliability of the measuring module was
found to be in the order of 80 percent.
IV - 71
-------
TABLE 2. Analysis of the Performance of Monitor W-20 Sensors from 1.12.68 to 15.08.70
H
<
1
^J
ro
Sensors
1. Number of
hours in
the period
2 . Operating
hours
3. Operating
hours %
Oxygen
68 69 70
432 8760 5448
422 6916 4361
97.6 78.9 80.0
Redox potential
68 69 70
Water Level
68 69 70
432
432
100
8760
7934
90.5
5448
4909
90.1
Temperature
68
69
70
Conductivity
68 69 70
432
269
62.2
8760 5448
7311 4859
83.4 89.1
Chlorides
(from 1.07.69)
68
69 70
Turbidity
68 69
432
426
98.6
8760
6358
72.5
70
5448
4779
87.7
PH
68 69 70
432 8760 5448
422 7306 4844
97.6 83.4 88.9
Solar radiation
(from 18.04.70)
68
69
70
68 69 70
1. Number of
hours in
the period
2. Operating
hours
3. Operating
hours %
434 8760 5448 434 8760 5448
404 7457 4859 327 7535 4899
93.5 85.1 89.1 75.7 86.0 89.9
4416 5448
3051 4716
69.0 86.4
2504
2504
100
-------
The interpretation of data collected by the monitoring stations is
subjected to somewhat different analyses than is usually performed else-
where as described by Manczak (4). the Polish workers consider three types
of flow-concentration curves to exist. These three curves, along with
their equations, are presented on Figure 15. The Type 1 curve represents
Q
a heavily polluted river and the value of -r is considered to be the amount
of wastes discharged and b represents the natural pollution ,of the river.
The Type 2 curve represents a clean river and the intercept at Q = 0 is
said to be natural pollution, increased flows representing pollution from
sludge resuspension and runoff accumulation. Finally, the Type 3 curve is for
pollution intermediate between the Type 1 and Type 2 curves. During low
flows, dilution is said to be important while during high flows the
important factor is sludge resuspension and runoff accumulation.
The magnitude of the number of observations made by the Polish workers
allows the development of empirical equations on a statistical basis. For
example, Figure 16 demonstrates a correlation between the initial five-day
BOD and the value of the river deoxygenation coefficient. Figure 17 shows
the flow concentration relationship developed, taking into account the
temperature effects on BOD and Figure 18 illustrates the curves developed
from previous considerations. Figure 19 through 22 demonstrate the types
of flow-concentration curves utilized by the Polish workers in. their water
pollution control studies and regulation. Finally, Figure 23'demonstrates
the results obtained using the W-20 monitoring station for a period of
almost two years.
The Polish workers then take each parameter of interest and statistically
analyze the data collected as shown in Figures 24 through 28. These analyses
IV - 73
-------
Basle types of curves re-
presenting concentration
of pollutants and rate of
flow
type I - for heaviily polluted
rivers
type II - for clean rivers
type III - for intermediately pol-
luted rivers.
FIGURE 15
k. (r}20°C=QOOl% BZlP50
0/4 0?
kt(r)20°C
Correlation between initial BOD- and the value
of
k1/r/ 20°C
FIGURE 16
IV - 74
-------
70
60 I
50
II
20
10
RZEKA OORA W CHAtUPKACH
RIVER ODRA AT CHAKUPK!
1067
20 40 60 80 100 120 140 160
tSO 200 220
Przoptyw m3/sek
Flow m3/sec
240
FIGURE 17
10
15
20 25 30
Q rrf/seN
Relationship between
MD,j and flow taking
into consideration
temperature effect
FIGURE 18
IV - 75
-------
14
12
|8
sJU
eg
« 8
£5 .
RZEKA ODRA W CHAKUPKACH
RIVER OEXUf AT CHAKUPK1
«*
CC< 0,001
1967
=-0,750
OC<0,001
20 40 60 80 KX) 120 140
160 180 200 220 240
Prztptyw m3/t»k
Row m3/wc
FIGURE 19
RZEKA ODRA W CHAtUPKACH
RIVER ODRA AT CHAKUPKI
v .10.43
*11-15«C
= 0,553
OC<0.001
=0,116
a>o,i
^«. ^^V *A _ J •
20 40 60 80 100 120 140 1
180 200 220 240
Przcpkyw nr/sek
Row m3/s«c
FIGURE 20
IV - 76
-------
300
I O
3_,
Si
66
x»
OOO-
- 1000
?_
Ifeoo
xo
RZEKA OORA W CHAkUPKACH
RIVER ODRA AT CHAtUPKI
?T
II "1
V> tf>
100
RZEKA ODRA W CHAtUPKACH .....
. RIWR OORA AT CHAtUPKI '*•'
.' A v
20 40 8) 80 100 t20 160 160
Pr»ptyw m^/iek
Row m3/»«
20 <0 60 80 W)> 00 KO WO
FVttptyw^/
Row m3/s»c
RZEKA ODRA W CHAtUPKACH
RIVER ODRA AT CHAtUPKI
196?
noo-
1000
600
SS
200
2040 W 80 DO BO WO 160
Przcptyw m^Hk
Flow m3/nc
. RZEKA AIREWBEAL
o RIVER AIRE AT BEAL lsw>/»7
SUCHA
TOTAL SOLIDS
02380
X
0.818
396
200 MO 600 800 TOO UOO MOO ROO
Pre«p»yw m3/t»k
Flow m^lstc
FIGURE 21
RZEKA ODRA W CMACUPKACN
RIVER OORA AT CHAtUPKI
dta xalcrMW tamptratur > 1S*C
for t«mp«ratur* rangt > 1S*C
diet zakrtsu ttmptratur 0-1S*C
• for t»mp«ratur« rang* 0-1S'C
Praptyw m3/
Flow m«/s«c
FIGURE 22
IV - 77
-------
, f
0r Br B
1 1
IJlfjl
ll
WYNIK1 POMiARdW W-20
RESULTS OF W-20 MEASUREMENTS
FIGURE 23
cNorW mg/l Q"
chlodd** mgA Cl"
Hi
6"
c*
H-
O
a
a
o
& §
*I
VO
FIGURE 24
IV - 78
-------
xas/6 pooi
FIGURE 25
PrzcNrdj poirtarowo-kontrolny
Morttortng ero*»-*«ctlon
25900
Rztka
River
OORA
Rok
Ytar
1966
M 28000
X
"* ***
•fi-Stcfuvt.
fcadunck
Chloride*
2
2
2
3
J
f,
,
OfauKtr zmicnnoiei
tadunku CT
1
//////////////////M
M
«
jjf
II
$/////////Ate
20 40 60 60 100 120 140 160 160 200 220
dni
days
>najdtuzci trwaj^cy tadunek
most frequently occuring values
" chloride* load
FIGURE 26
IV - 79
-------
Przekr6j pomiarowo-kontrotny
Monitoring cross-section
Rzeka
River
5°* 1968
Year
S
8
21
29
72
fa
21
b
5rJl
•«*
Pi
.031 _
I!
i!
nojdlruiej trwajqce steienle
100 most frequently occuring values
—of chlorides concentration
20 40 60 80 100 dn!
ctoys
Prequency of occurrence ctirve of chlorides con-
centrations in Odra river at Chaiuplci, 1968.
FIGURE 27
Przekrxij pomiarowo-kontrolny_,,,,.._,,.
Monitoring cross-section CHAUIPKI
Rzeka
Rivzr
ODRA
600i
1*00
^
e
J
*-
•5
2
,200
3
4
a
N
K
'c
u
•N
C
r 300
b
01
e
c200
O
1
'c
c
O
u
chlorides
1
|
JC
0
u
V
1
•N
cr
_^
Ul
CT
je
-^•j
O
(A
•D
0
U
1
1
5
•5
jt
»
•o
S
hrzywa czasow trwania stezzrt cMorkdw
time duration curve of chlorides concentration
b. krzywa czp.sbw trwanio rtdunhu chlorkow
time duration curve of chlorides load
_„ \
-------
then form the bases for establishing standards and water quality goals in
the river in question. While it may be observed that the Polish workers
are able to collect much more data than is usually possible in the United
States, their procedures appear to be quite rational and form a sound
basis for the rational establishment of water quality goals.
Automatic Water Quality Monitoring in Czechoslovakia
In 1966, three experimental monitoring stations were installed on
the Vltava River, downstream from Prague, on the Ohre River near Carlsbad
and in Terezin. At these stations, Czechoslovakian made analyzers were
tested for the monitoring of temperature, pH, DO, conductivity, chlorides,
turbidity, color and different type pumps. New devices have been developed
in Czechoslovakia for the automatic determination of pH, ORP, conductivity,
DO and temperature. In addition, the central unit of the monitoring station
has been developed so that it is able to transform signals coming from
various analyzers into digital form for printing, magnetic tape or punched
tape registration, in combination with telemetry, if needed. It is planned
to place 50 automatic water quality monitoring stations throughout
Czechoslovakia in the future.
Automatic Water Quality Monitoring in Hungary
A World Health Organization water quality management project has
snabledthe Hungarian workers to plan for five continuously operating auto-
matic water quality monitoring stations in that country. The instruments
installed are or will be Hungarian made and will measure temperature, DO,
conductivity, pH, turbidity and solar radiation.
IV - 81
-------
The data will be recorded graphically using a five point recorder
and will also be connected to a telex system. This will allow interrogation
of the station, resulting in printout of the water quality results on an
hourly basis. Interrogation will be made from a center which will be located
in the Hungarian Research Institute for Water Resources Development in
Budapest where computer facilities are available. Statistical analysis of the
data as well as data storage will be accomplished by the computer.
Three monitors are foreseen for the Sajo River, the first station,
located at the Hungarian-Czechoslovakian border, was placed into operation
in April of 1973. A second station on the Sajo will be placed into operation
during December of 1973 and is approximately 69 kilometers from the first
one. The third station will be a mobile one and will be mounted to a trailer.
This will be used as water quality conditions require. The fourth station
will be placed in the vicinity of the border between Hungary and Czechoslovakia
and the location of the fifth station is not yet decided.
Summary and Conclusions
The status of water quality monitoring in Eastern Europe has been
presented and the role of the World Health Organization delineated. It was
shown that the most experience in the design and operation of these stations
has been gained in Poland thus far, however, other countries are rapidly
planning for comprehensive automatic water quality monitoring systems.
The procedures for the establishment of station locations in Poland
were examined and it was seen that a considerable amount of care was exercised
by the Polish workers. In addition, the vast array of data collected by the
Polish authorities has resulted in some rather interesting relationships
regarding river flow and concentration of certain pollutants.
IV - 82
-------
It may be concluded that some substantial advances have been made
in automatic water quality monitoring in Europe and that cooperation
between European countries and the United States should be encouraged.
IV - 83
-------
REFERENCES
1. Economic Commission for Europe, Body on Water Resources and
Water Pollution Control Problems, Experience in Setting up and operating
Automatic Water Quality Monitoring Stations, Jan. 1969.
2. Krenkel, P. A., Editor, Proceedings of the Specialty Conference on
Automatic Water Quality Monitoring in Europe, Department of Environmental
and Water Resources Engineering, Vanderbilt University, Nashville,
Tennessee, Technical Report No. 28.
3. Florcyk, H., Polish Studies on the Location of Automatic Water Quality
Monitoring Stations, [Proc. Specialty Conference on Automatic
Water Quality Monitoring in Europe, Dept. of Environmental and Water
Resources Engineering, Vanderbilt University, Nashville, Tennessee,
Tech. Report 28, Edited by P. A. Krenkel.]
4. Manczak, H., A Statistical Model of the Interdependence of River Flow
Rate and Pollution Concentrations, [Proc. Specialty Conference on
Automatic Water Quality Monitoring in Europe, Dept. of Environmental
and Water Resources Engineering, Vanderbilt University, Nashville,
Tennessee, Tech. Report 28, Edited by P. A. Krenkel.]
IV - 84
-------
AN AIRBORNE LASER FLUOROSEN.SOR
FOR THE
DETECTION OF OIL ON WATER
By
H. H. Kim
Wallops Station
Wallops Island, VA 23337
and
G. D. Hickman
Sparcom Inc.
Alexandria, VA 22304
IV - 85
-------
An Airborne Laser Fluorosensor for the
Detection of Oil on Water
A remote active sensor system designed to detect laser induced fluorescence
from organic and biological materials in water has been suggested by a
number of investigators. (1,2) Several different laser airborne systems are
in the process of being developed in both the U.S. and Canada. (3,4)
In this presentation, we would like to report our successful operation of
an airborne laser fluorosensor system which is designed to detect and map
surface oil, .either natural seepage or spills, in large bodies of water.
The test flights were conducted in daylight preliminary results indicate
that the sensitivity of the instrument exceeds that of conventional passive
remote sensors which are available for the detection of an oil spill today.
The package was jointly developed by NASA Wallops Station and Sparcom Inc.
of Alexandria, Va. The salient features of the system consist of a pulsed
nitrogen laser, a f/1 28 cm diameter Cassegranian telescope and a high
gain photomultiplier tube (RCA 8575) filtered by a U.V. blocking filter
(0.01% and 0.3% transmission at 337 ran and 390 run respectively). The laser
produces a nominal 1 m joule pulse of 10 nsec duration at 337 nm contained
in a retangular beam having a half angle divergence of approximately 30 by
2 mradians. The repetition rate 100 pulses per second affords one good
spatial resolution when operated from an aircraft flying at 300 km/hr..
Figure 1 is a photograph showing the laser equipment installed in NASA DC-4
aircraft.
IV - 86
-------
H
w
ro
0
n
0
M
(D
H D
< en
o
CO H
~J 3
W
ft
CO
W
n
H
HI
rt
-------
The laser induced fluorescence of the oil in the 450-500 nm spectral region
was monitored. Each return pulse-was fed into a range gated multi-mode
analog to digital (ADC) conversion unit which recorded the peak amplitude of
fluorescence. Even though the pulse width of the return fluorescence did
not exceed 10 nsec, the width of the input gate to the ADC was considerably
wider. This was to insure signal detection as fluctuations occured in the
laser/oil distance which were produced by aircraft motion: roll, pitch and
changes in altitude.
A 35 mm frame aerial camera equipped with a wide angle lens viewed-the same
area on the water surface as seen by the fluorosensor. Our experiences
gained through previous NASA aircraft photo surveillance missions have shown
us that the color photographic image technique is still one of consistently
reliable positive indicators of the presence, position, and extent of the
oil slicks.^
The first series of flight tests were conducted in conjunction with a con-
trolled oil spill off Norfolk, Virginia in May 1973. This spill consisted
of 400 gallons of No. 4 grade heating oil. The field experiments were
managed by the U.S. Coast Guard. The NASA aircraft containing both the oil
fluorosensor and a dual channel microwave radiometer,C6) flew over the spill
site at altitudes ranging from 100-1000 feet". Figure 2 illustrates typical
return signals which were obtained at the airborne receiver from the surface
oil as the plane passed over the slick. The data shown in this figure was
obtained from an aircraft altitude of 400 feet. This figure shows a large
but fairly constant background previous to (< 4 seconds) and after (> 12
seconds) the plane's passage over the spill. There is a marked increase
IV - 88
-------
50
600 m
Figure 2:
Typical Fluorescent Signals Observed with
Laser Fluorosensor-Aircraft Altitude 400 Ft
IV - 89
-------
in the amplitude of the detected signal during the period of time that the
aircraft was over the oil slick. Detection of the oil was recorded by the
dual channel microwave radiometer during the time period of 6-8 seconds,
which is close to the center of the spill. In all probability this re-
presented the thickest layer of oil. This single qualitative^experiment
dramatically showed that while the microwave radiometer was able to detect
the-central portion of the spill, the increased sensitivity of the laser
fluorosensor. permitted detection'.of approximately the entire visual extent
of the slick. Although the thickness of the oil changes as the oil spreads
on the sui.-Tace of the water, the amplitude of the fluorescent signal remained
essentially constant (Figure 2). Since oil exhibits extreme absorption in
the UV region of the spectrum one would expect the amplitude of the fluores-
cence to be relatively independent of thickness. This is in agreement with
the flight test results. Confirmation of the dependence of oil thickness
on fluorescence has been made in the laboratory.
A second set of flight tests consisting of six separate flights was made in
August, 1973 to detect ambient oil on the Delaware River. Figure 3 shows
the results of one of these flights from a 48 km section of the river
between the Chesapeake and Delaware Bay Canal to the Delaware - Pennsylvania
state line. The observed fluorescent intensity was approximately 5 times
higher in the upper section of the Delaware River as in the lower section
of the river. The background noise was substantially reduced over that
recorded in the initial flight test. This was accomplished by narrowing
the gate width of the digitizer input from 250 to 50 nsecs. The system
was calibrated to register a value of 50 on the ADC unit against a thin
oil film target in full view of the receiver at an altitude of 500 feet
IV - 90
-------
H
H-
-•-i
f ;
hi
CD
CO
v£) P- 3
-Jrt &
UJ ^ H-
CD
rf 3
H-'Cr' n-
O fD
" K. °
^ > H-
O H" h-1
O
(D
ro
w
l_l.
b
ci-
d o
o cu
H M
o &
CO -'
fD £U
3 fl
cn o>
O
H jo
• H-
<
05
to
10
(D
O.
MEMORIAL BRIDGES
C.and O. CANAL
48
km
O 50
AMBIENT^OIL LEVELS IN THE DELAWARE RIVER ( AUG. 24 *73 10:40am
-------
and a value of zero against ambient noise in the open sea. This was
accomplished by adjusting the gains of the phototube and the threshold
levels of the input discriminator to the digitizer. Therefore, our
calibration procedure assured us that the signal observed in the lower
section of the river was a real fluorescence and not background noise.
Figure 4 shows a bar chart of the morning flight results, previously
shown in Figure 3, along with the return afternoon flight made the same
day. Each block in the figure represents an-average value of'3000 return
pulses. -This -figure shows dramatically the change in the intensity of
the oil in the lower section of the river in a fairly short time.
Images from the aerial photography showed the presence of oil when a
scale reading of 50 or greater was reached on the ADC output. There-
fore, photography did not show the presence of oil in the lower section
of the river during the morning flight, although detection of the oilj was
made with the laser fluorosensor. This is significant, in that it shows
the tremendous sensitivity of the laser fluorosensor in detecting traces
of oil that can not be detected by other remote sensors.
IV - 92
-------
OIL SPREADING
10:40am
>. 50
4J
•H
fi
-------
REFERENCES
1. G. D. Hickman and R. B. Moore, "Laser Induced Fluoresin Rhodamine
B and Algae". Proc. 13th Conf., Great Lakes Res. 1970.
2. -J. F. Fantasia, T. M. Hard, H. C. Ingrao, "An Investigation of Oil
Fluorescence as a Technique for the Remote Sensing of Oil Spills".
DOT-TSC Report 71-7.
3. H. H. Kim, "New Algae Mapping Technique by the Use of an Airborne
Laser Fluorosensor". Applied Optics, vol. 12, p. 454 - 62 July 1973.
4. The following papers were presented at Hydrographic
Lidar Conference held at Wallops Island, VA Sept. 1973.
R. A. O'Neil, A. R. Davis, H. G. Gross, J. Kruus, "A Remote Sensing
Laser Fluorometer".
M. Bristow, "Development of A Laser Fluorosensor for Airborne
Surveying of the Aquatip Environment".
P. B. Mumola, Olin Jarrett, Jr. and C. A. Brown, Jr, "Multicolor
Lidar for Remote Sensing of Algae and Phytoplankton".
5. J. C. Munday Jr., W. G. Mclntyre, M. E. Penney and J. D. Oberholtzer
"Oil slick Studies uning Photographic and Multi-scanner DATA".
Proc. of the 7th TNT Sym. of Remote Sensing of the Environment.
Willow Renlab, University of Michigan, Ann Arbor, 1971, p. - 1027
IV - 94
-------
6. J. P. Hollinger and R. A. Mennella, "Oil Spills: Measurements of
Their. Distribution and Volumes by Multifrequency Microwave Radiometry,
Science, Vol. 191, p. 54, July 1973.
IV - 95
-------
SESSION V.
ENVIRONMENTAL THEMATIC MA.PPING
CHAIRMAN
MR. JOHN D. KOUTSANDREAS
OFFICE OF MONITORING SYSTEMS, OR&D
-------
THE U.S. GEOLOGICAL SURVEY AND LAND USE MAPPING *
James R. Anderson
Chief Geographer
U.S. Geological Survey
Washington, D.C. 20244
BACKGROUND
A modern nation, as a modern business, must have adequate informa-
tion of many complex, interrelated aspects of its activities in order
to make decisions. Land use is only one such aspect, but it is one
that has become increasingly important as the Nation seeks to grapple
with the problems of. suburban expansion, demand for outdoor recreation,
highway and transportation planning, environmental quality, and use of
energy resources. The land area of the United States is a finite
quantity that has not changed very much for more than a century and is
not likely to change in the future. However, the uses made of the
Nation's land and water resources have changed greatly.
Urbanization has been absorbing land at the rate of about 730,000
acres per year during the 1960's and another 130,000 acres per year
were being transferred to transportation uses from other uses. About
1 million acres per year have in part been going into some kind of
recreational use during the past decade. In the future, the possible
use of strip mining to increase the exploitation of coal resources
could bring significant land use changes to those areas of the Nation
where strippable coal deposits exist. To date about 1.5 million acres
of land have been disturbed by strip mining of coal but as much as
45 million acres with strippable coal deposits exist in the United States.
* For oral presentation and possible publication in the Proceedings of
the U.S. Environmental Protection Agency's Second Environmental Quality
Sensors Conference, Las Vegas, October 11, 1973.
V - 1
-------
Therefore, the growing population of this country coupled with
a widening horizon of demands being made on land resources has brought
an expanding array of pressures on the available resource base. These
pressures have brought conflicts in many parts of the Nation that
urgently need attention. Some examples include agricultural production
in conflict with real estate development and resulting urbanization;
environmental protection versus production of energy to meet increasing
demands for power; recreational development versus the use of land for
forestry, grazing, and extractive uses; conservation of coastal areas
for recreational uses in the face of needs for more port facilities and
shoreline industrial sites; preservation of wetlands for natural wild-
life and fisheries habitat in the face of new demands for development
of such wetlands for urban uses, agricultural production, and other
uses,
In recent years Americans in general have become more and more
concerned about resources and their use, about the quality of the
environment, about urbanization of productive agricultural land, need
for recreational development closer to where they live, and many other
local, state, and national resource issues. However, there is a
widespread lack of understanding of resource use and environmentally
related problems. For example, the current furor over the high cost
of food, particularly meat, might be more rational if more Americans
realized that an average acre of land is hard pressed to produce
500,000 calories of food per year if used for beef production. When
used to produce wheat or rice that average acre can produce about
2 million calories of food annually. Each American consumes about
a million calories a year. The Chinese are not vegetarians by choice
V - 2
-------
but by necessity. Peanut butter has as much protein per pound as
beefsteak but obviously most Americans are still insisting on eating
beef. I well remember when I was a boy in the 1930's how disgraceful
my mother always considered it was to serve oleomargarine. Today
many of us use it as an acceptable substitute for butter - that is
until the recent rise in the price of soybeans caused a significant
narrowing of the price spread between oleomargarine and butter.
Another example of American's general lack of understanding of
resources and their use relates to the present scarcity of energy.
When I was growing up on a farm in the White River Valley of Indiana,
we used only 30 kwh of electricity per month. Recently when I paid
my last electric bill in moving from Florida to Reston, Virginia,
our family had used 1,543 kwh in the month of December, a good month
for Florida. This great expansion in the consumption of energy by
Americans has brought us rather abruptly to an agonizing reappraisal
of priorities and options necessary to bring about a solution to the
current energy shortages compatible with the need to maintain and
improve 'the quality of the environment.
NEED FOR LAND USE DATA
The increasing number and complexity of land use conflicts
indicate a need for positive private and public planning efforts to
resolve these acute problems and to prevent or reduce future conflicts,
The severe strains being placed upon the natural environment in many
parts of the country must be reduced and the stresses on social,
political, and economic institutions must be relieved. Improvement
in the land use decision making processes at local, state, and Federal
levels is one way of reducing these strains and solving these problems,
V -
-------
Many responsible public officials and prominant authorities on
land resource planning, decision making, and management have stressed
the need for more information about existing land use. For example,
Marion Clawson, former Director of the Bureau of Land Management, author
of numerous books dealing with land resources, and for several years
with Resources for the Future, makes the following statement in the
Foreword to a report published in 1965 on Land Use Information: A
Critical Survey of U.S. Statistics Including Possibilities for
Greater Uniformity:
"In this dynamic situation, accurate, meaningful, current
data on land use are essential. If public agencies and private
organizations are to know what is happening, and are to make
sound plans for their own future action, then reliable informa-
tion is critical."(2)
Pending legislation in the 93rd Congress recognizes the need for
Federal participation in the collection of land use data. In Senate
Bill 268, Title II, Section 202, the Secretary of the Interior,
"Acting through the Office (of Land Use Policy Administration), shall:
A) Maintain a continuing study of the land resources of
the United States and their user
B) Cooperate with the States in the development of standard
methods and classifications for the collection of land use
data and in the establishment of effective procedures for
the exchange and dissemination of land use data: ..."
H.R. 4862 is also quite specific on the need for adequate informa-
tion on land use. In Section 101 of Title I, the following statements
are made:
"The Congress finds that adequate data and information
V - 4
-------
on land use and systematic methods of collection, classifi-
cation, and utilization thereof are either lacking or not
readily available to public and private land use decision
makers; and that a national land use policy must place a
high priority on the procurement and dissemination of
useful land use data."™'
One of the prime requisites for better use of land is information
on existing land use and changes in land use over time. The present
distribution of agricultural, recreational and urban land; knowledge
of how and where urbanization and other development has been occurring;
and data on the proportions of a given area recently devoted to dif-
ferent uses are used by the legislators and state officials to
determine land use policy, and by planners to project transportation
demand, identify areas where future development pressure will be
greatest, estimate future infrastructure requirements, and develop
more effective plans for regional development.
Another possible use of current land use data is in equalization
of tax assessment procedures among counties. A representative of a
State revenue office indicated that current statistical data on dif-
ferent land uses in each county would be invaluable to the State in
reviewing assessment reports submitted by county assessors.
Information on existing land use and changes in land use is also
of significance for water resource planning. As land is changed from
agricultural or forestry uses to urban uses surface water runoff
increases in magnitude, flood peaks become sharper, surface and ground
water quality deteriorates, and water use increases thereby reducing
V - 5
-------
water availability. By monitoring and projecting land use trends
it will be possible to develop more effective plans for flood control,
water supply, and waste water treatment.
Federal users also need current land use information. The assess-
ment of recreational needs and opportunities requires knowledge of the
location and extent of urban areas and potential recreational lands.
This information is used to forecast demand, identify potential solu-
tions, and develop recreation plans. Comprehensive inventories of
existing uses of public lands plus the existing and changing uses of
adjacent private lands can improve the management of public lands.
Other Federal uses of land use data include assessing the impact of
energy resource development, water resource and river basin planning,
managing wildlife resources and studying changes in the use of lands
in the migratory bird flyways, and preparing national overviews of
changes in the use of land for national policy formulation.
Another application of land use data is in assessing the impact
of natural disasters such as floods. Statistics on the acres of agri-
cultural, urban and other types of land inundated by flood waters
would be invaluable in estimating damages, future crop losses, and
consequent econitnic impacts.
Presently, there is no systematic compilation of information on
existing land use and its changes on a national basis. For detailed
planning at the local level, ground surveys, occasionally supplemented
by aerial photographs^ are used. In some cases, land use information
is hypothesized on the basis of data on utility hookups, school popula-
tion, building permits, and similar information. Transportation
planners collect the necessary information using similar techniques.
V - 6
-------
Some States such as Connecticut, 5 jjew York, ^ and Minnesota; '
have land use information available on maps at scales ranging from
1:24,000 to 1:500,000, but in most cases these States have not been
able to update the land use maps, therefore, they have decreasing
utility. Some Federal agencies, such as the Forest Service, Soil
Conservation Service, and Bureau of Land Management, collect some
land use information, but it is for a specific need and is difficult
to adapt to other uses. In 1958, and again in 1967, a National
Inventory of Soil and Water Conservation Needs was carried out by
/ o \
the U.S. Soil Conservation Service. The inventories have pro-
vided much useful general information about land uses by counties,
but since the inventory was based on a two percent sampling of the
total area of the United States it is deficient with respect to
specific geographic distributions of various land uses.
Some of the major problems with these existing data sources are
the lack of consistency, the age of the data, spotty coverage, and
the use of incompatible classification systems. The data have been
collected to meet specific limited needs using definitions of use
classes which are appropriate for only that need. They have often
been collected on a one-time basis so the data are of marginal utility
for other applications. Furthermore, it is nearly impossible to aggre-
gate the available data because of the differing classification system
used.
DEVELOPMENT OF A LAND USE DATA PROGRAM
A step to develop a framework for the meaningful classification
of land use on a nationwide basis has been taken by the U.S. Geological
Survey. In Geological Survey Circular 671 entitled "A Land Use
V - 7
-------
Classification System for Use with Remote Sensor Data," published in
October 1972, a land use classification system is proposed for testing
and review. This classification system has been developed to meet the
needs of Federal and State agencies for an up-to-date overview of land
use throughout the country on a basis that is uniform in date, scale,
and categorization at the more generalized first and second levels.
Remote sensor data will definitely be the most cost-effective method
of acquiring such land use information. Data from ERTS-1, from high
altitude aircraft platforms, and from the Special Mapping Center of
USGS at Reston, Virginia are available for obtaining land use informa-
tion. The classification system utilizes the best features of existing
widely used classification systems to the extent that they are amenable
to use with remote sensing, and it is open-ended so that regional,
state, and local agencies may develop more detailed land use classifica-
tion systems, at third and fourth levels, to meet their particular
needs and at the same time remain compatible with the national system.
There has been longstanding use of remote sensing technology in
land resource inventory and mapping. In the 1930's the Tennessee
Valley Authority made extensive use of air photographs in land and
water resource surveys. The U.S. Soil Conservation Service has found
aerial photography invaluable for soil surveying and the U.S. Geological
Survey has for many years also made extensive use of air photographs
in its Topographic Mapping Program.
However, only in the mid-19601s was there a growing recognition
of the utility of the technology of remote sensing for extensive
resource-oriented land use inventory and mapping. We now have the
remote sensor capability to map effectively the land uses of the entire
V - 8
-------
United States within a reasonable time frame of approximately three
years. We also have been working to provide an effective and efficient
approach to the periodic updating of land use data. While it is not
possible to collect all of the information about land use from remote
sensor sources that is needed for planning, management, regulatory,
and other purposes, experiments presently being carried out in the
Geological Survey and elsewhere indicate the desirability of using
remote sensing technology as much as possible if cost effective inven-
tory and mapping of land use is to be achieved.
Because land use data are highly perishable and need to be updated
frequently it is very necessary that the compilation and dissemination
of such data be in accord with a schedule compatible with the need for
updating at regular and frequent intervals. Some areas have a much
more dynamic land use situation than do others, therefore the need
for updating land use data is not uniform throughout the country.
For example, rangeland and tundra areas of the Western States and
Alaska respectively do not generally need surveying nearly as frequently
on the whole as do areas around major cities, many coastal areas, areas
of critical environmental concern, and areas of recreational impact.
In order to accommodate the need for promptness and regularity
in land use inventory and mapping, it will be necessary to revise
traditional approaches to the compilation and dissemination of land
use data. Traditionally land use data have been compiled on attractive
multi-colored lithographed maps that were generally out-of-date before
publication. The United Kingdom has had two land use surveys - one
in the 1930's and the other in the 1960's that utilized this approach.
V -9
-------
More recently Japan, is publishing a series of land use maps that show
great detail about the uses being made of land in that country.
Land use data has been presented at a great variety of scales
ranging from detailed scales of 1:10,000 or even larger scale for
metropolitan planning to very gross scales such as 1:2,500,000 and
1:5,000,000 used for the very attractive land use maps of the World
Atlas of Agriculture presently being published in Italy, In the
United States it seems appropriate for nationwide coverage in land
use mapping and inventory to use a scale of 1:250,000 for which an
existing base map is available for presentation, with scales of
1:24,000 or 1:50,000 utilized for selected areas where greater detail
in land use data is needed.
In the future, careful considerable.will need to be given to the
use of computer-generated graphic and tabular displays of land use
data in order to achieve the timeliness and regularity of presentation
needed by many of the users of land use information. Use of computer
technology is also very important from the standpoint of developing
the capability of making rapid and varied comparisons of land use data
with other data sets pertaining to land resources such as slope, soil
type, access to various kinds of transportation, population density, etc,
The approach to land use classification being proposed by the
U.S. Geological Survey is resource-oriented. In contrast to the
people-oriented system developed inthe mid-1960's by the Urban Renewal
Administration and the Bureau.of Public Roads and published as the
Standard Land Use Coding Manual. • The people-oriented system assigns
7 of the 9 more generalized first level categories to urban uses of
V - 10
-------
land which account for less than 5 percent of the total area of the
United States. This 5 percent of the area has, however, about 95 percent
of the total population of the United States - thus a real need exists
for an urban-oriented land use classification system.
However, it has become increasingly apparent that a resource-
oriented land use classification system is also needed. The USGS
classification system has been developed to meet that need. Eight
of the proposed nine Level I categories are associated with the
95 percent of the total area of the United States not in urban and
built-up uses. A considerable degree of compatiblity between these
two systems of classification can be achieved at the more generalized
levels. However, complete compatiblity can probably not be achieved
between land use data collected from ground observation and enumeration
and that compiled from remote sensor sources, particularly at the more
detailed levels of categorization.
In developing a land use data program with a resource orientation
several basic general assumptions, needs, and requirements have been
recognized.
In the first place, there is much evidence that there is a need
for an effective national and interstate regional perspective of the
major uses of land in the United States. Federal and State agencies
engaged in land management and planning activities need to have a
comparable set of land use data in order to carry out various Federal
and State interagency activities effectively,
In the second place, it has been assumed that no one ideal land
use classification will ever be developed. Therefore provision has
been made to provide for flexibility but at the same time make it
possible to .summarize and generalize on a reasonably uniform basis
V - 11
-------
Thus each Federal agency and each State will be able to develop an
extension of the USGS classification system appropriate for their
specific needs. In other words, no attempt will be made to develop
or seek adoption of a common approach to the categorization of land
use at the more detailed Levels III and IV.
Thirdly, it has been clearly demonstrated that remote sensor data
offer the most efficient and least costly means of compiling land use
data over extensive areas at the more generalized Levels I and. II in
the USGS land use classification system. In using remote sensor data
at the main data base, it is necessary to use a "cover" approach to
land use classification rather than an "activity" approach. Farming,
grazing, and forestry are activities: cropland, rangeland, and
forestland are cover-oriented categories.
In the development of the USGS land, use data program provision
is being made to accommodate the need for additional items of land
use data that cannot be provided directly from remote sensor sources.
For example, the outdoor recreational uses of land are often very
difficult to determine from remote sensor data. Hunting may be
associated with several different uses of land. Data on hunting
activity gathered by enumeration and ground observation can be
included as an overlay or additional data set in the Geographic Informa-
tion System being developed for the handling and dissemination of land
use data in the U.S. Geological Survey.
Demonstration of a reasonable level of accuracy in compiling land
use data from remote sensor data sources must be made if there is to
be widespread acceptance of the land use data compiled from such
sources. At the more generalized levels of categorization coirxect
V - 12
-------
interpretation of land use 85 to 90 percent of the time is considered
to be an adequate level of accuracy. Often overlooked is the fact
that data collected by enumeration and ground observation techniques
is generally not infallible. For example, in the Census of Agriculture
taken every five years by the U.S. Bureau of the Census, there is
generally an under-enumeration of about 7 to 8 percent with incorrect
enumeration of land use items on an additional 5 to 10 percent of
the farmland being enumerated.
In the Land Use Data Program of the U.S. Geological Survey, atten-
tion is being given only to the compilation of present or current land
use data. No attempt is being made in this program to rate or evaluate
the capability or suitability of land for residential, agricultural,
recreational, or other uses.
EXPERIMENTAL PROGRAM PRODUCTS
Specific products being provided experimentally and operationally
by the USGS Land Use Data Program are:
0 Transparent overlays depicting current land use classified
by the 34 categories in Level II of the USGS land use
classification system formatted to the standard 1:250,000
scale topographic maps and showing the political units,
major classes of public land ownership, river basins and
sub-basins, and statistical recording areas such as census
tracts.
° Transparent overlays formatted to a photo-image base at 1:50,000
scale showing existing Level II land use for selected Standard
Metropolitan Statistical Areas (SMSA), and other areas where a
larger scale is needed.
V - 13
-------
0 Computer-generated graphic displays (maps) at 1:250,000 scale.
0 Computer tapes of current land use data (for those users who
desire the data in this form and who have appropriate computer
capability).
0 Computer-generated overlays showing the changes in land use
for intervals of from three to ten years, depending on the
area.
0 Statistical data on land use patterns and changes for various
political, statistical, and administrative units.
SUMMARY
Clearly stressed in the 1972 Report of the Citizen's Committee on
Environmental Quality is the fact that "Of all the factors that determine
the quality of our environment, the most fundamental is the use we make
of our land." This is true not only in the big urban centers of our
country where such problems as air and noise pollution plague us and
often go unresolved, but land use is also a key factor in determining
environmental quality in the most remote rural and relatively untouched
parts of the Nation. Each situation bears careful study and evaluation
as a basis for action before steps are taken to resolve conflicts and
relieve pressures relating to conservation and development of resources.
Such evaluations should be based on the best available facts about
current, past, and possible future land use patterns and the many
varied conditions that determine those patterns. To date many of the
basic facts have been lacking over extensive parts of the Nation.
In summary, land inventory and resource analysis should always
include and recognize the need for:
1. Timely and cost-effective inventory procedures including
current land use'with geographic location capability;
V - 14
-------
2. Regular monitoring of change in quality and use of land
and water resources;
3. Standardization of classification systems;
4, Capability for effective integration and synthesis of
information;
5. Systematic and regular analysis of changing patterns
of resource use.
* * >v
V - 15
-------
REFERENCES
U.S. Dept. of Agriculture, 1972, Farmland: are we running out?
The Farm Index, Dec. 1972, p. 8-10.
Clawson, Marion and Charles L. Stewart, 1965, Land use informa-
tion: a critical survey of U.S. statistics including
possibilities for greater uniformity: Resources for the
Future, Inc., 402 p., index.
U.S. Congress, Senate, 1973, Senate Bill 268, 93d Congress,
1st sess., 55 p.
U.S. Congress, House, 1973, H.R. 4862, 93d Congress, 1st
sess., 33 p.
Connecticut, Dept. of Finance and Control, Office of State
Planning, 1970, 1970 Land use study for Connecticut,
unpub. report,
New York, Office of Planning Services, 1972, Land use and
natural resource inventory of New York State, LUNR
classification manual: Albany, June 1972, 23 p.
Minnesota, Univ. of, Minnesota Land Management Information
System Study, 1971, State of Minnesota land use, 1969:
Minnesota Land Management Information System Study map
prepared under contract with the Minnesota State
Planning Agency, scale: 1:500,000.
U.S. Dept. of Agriculture, Basic statistics: national inventory
of soil and water conservation needs, 1967: Statistical
Bulletin No. 461., Washington, D.C., 211 p.
U.S. Urban Renewal Admin., Housing and Home Finance Agency
and Bureau of Public Roads, Dept. of Commerce, 1965,
Standard land use coding manual: Washington, D.C.,
1st ed., 1965, 111 p.
V - 16
-------
REMOTE SENSING DATA, A BASIS FOR
MONITORING SYSTEMS DESIGN
James V. Taranlk
Chief of Remote Sensing
Samuel J. Tuthill
Director and State Geologist
Iowa Geological Survey
16 West Jefferson Street
Iowa City, Iowa 52240
(319)338-1173
A paper for the U.S.Environmental Protection Agency's Second
Conference on Environmental Quality Sensors. EPA National
Environmental Research Center, Las Vegas, Nevada, October
10-11, 1973.
1 October 1973
V - 17
-------
REMOTE SENSING DATA, A BASIS FOR MONITORING SYSTEMS DESIGN
James V. Taranik
Chief of Remote Sensing
Samuel J. Tuthtll
Director, and State Geologist
Iowa Geological Survey
16 West Jefferson Street
Iowa City, Iowa 52240
ABSTRACT
Recent remote sensing investigations of large river systems bordering Iowa have
demonstrated that rapidly acquired remotely sensed data can effectively provide
information for monitoring systems design. A thermal remote sensing survey of 180
miles of the Mississippi River was conducted by the Iowa Geological Survey, Remote
Sensing Laboratory prior to the operation of a nuclear power plant in the Davenport-
Moline area of Iowa-Illinois. Thermal mapper data provided information on locations
of sources of thermal effluent, surface current distribution patterns in the river, and
relationships between thermal regimen of slough areas and river channel. Interpre-
tation of this data can provide information for the placement of monitoring sensors
and their synoptic nature can provide the insight into regional environmental patterns.
An aerial photographic flood mapping mission was flown over 70 miles of the Nishnabotna
River in western Iowa. The purpose of this mission was to accurately map distribution
of flood high water several days after waters recede. This remote sensing Investigation
provided data on the extent of floodpfain inundation following a 100-year recurrence
interval flood. Interpretation of this data can provide information on the nature and
areal extent of the floodplain. Delineation of the floodplain can provide information
for solid and liquid waste management, and for development of environmental disaster
plans.
THERMAL MAPPING OF THE MISSISSIPPI RIVER IN IOWA
Along the Missouri and Mississippi rivers bordering Iowa power companies, industries
and municipalities contribute thermal effluent to the thermal regimen of these river
systems. The Atomic Energy Commission and power companies are pressing for the
development of Atomic Power plants which would use river water for cooling purposes
and return theJieated water to the main channel. Warm effluents, nutrients from
runoff over agricultural land, and effluent from waste treatment plants provide the
ingredients for a major pollution problem along these river systems. Alternative
solutions to the cooling problem (use of groundwater, cooling towers, and cooling
ponds) are expensive solutions, and in some cases unsatisfactory solutions for the
cooling requirements.
V - 18
-------
Development of an environmental water quality surveillance system requires
knowledge of the natural and manmade characteristics of the hydrologic environment.
For a large hydrologic system, like the Mississippi River, gathering data by
conventional ground techniques can be an expensive task, both in terms of time and
money. Granting permits to operate requires analysis of environmental impact and
usually little time or money is available for extensive data collection. Use of the
remote sensing approach can provide a synoptic overview of entire systems, and
can provide data for the location of ground instrumentation.
In the spring of 1971 Commonwealth Edison Company sought to establish a diffuser
pipe in the Mississippi River near Cordova, Illinois. At that time little was known of
the thermal regimen of the river surrounding Cordova, and pressure from environmentally
oriented groups in Iowa demanded that thermal data be gathered on a regional basis.
In cooperation with the Iowa Conservation Commission, Commonwealth Edison, and
several other Iowa state agencies, Iowa Geological Survey decided to acquire remotely
sensed thermal mapper data over a large portion of the Mississippi River as it borders
Iowa. The decision to acquire remotely sensed thermal data was based upon:
1. The limited amount of time available to determine the thermal characteristics
of the river around the proposed diffuser pipe at Cordova;
2. The considerable expense of extensive point data collection studies over several
hundred miles of river;
3. The uncertainty of point data acquired by probe and thermometer instrumentation.
Generally point data acquired by these instruments do not measure the thermal flux out
of the skin of the river and thus the heat budget of the system is difficult to assess.
Also, large variations in temperature can occur within a few inches of the river
surface, and unless considerable care is taken with conventional measurements, large
uncertainties can be introduced by inadvertantly measuring temperature at varying
depths.
4. The uncertainty introduced by interpreting point data and contouring these data.
5. The contemporaneous nature of thermal mapper data. Contemporaneous point
data requires extensive ground instrumentation.
In the spring of 1971 Commonwealth Edison Company agreed to fund a data acquisition
remote sensing mission over the Mississippi River from Clinton to Keokuk, and to supply
Iowa Geological Survey with one copy of all imagery for unrestricted public use. Pre-flight
planning and the mission were coordinated by Dr. Samuel J. Tuthill, State Geologist, of
Iowa, Donald B. McDonald, Civil Engineering Department, University of Iowa, and
by Victor I. Myers, director of the Remote Sensing Institute at Brookings, South Dakota.
The Remote Sensing Institute at Brookings acted as technical consultants for mission design
and data analysis. In September 1971 the Remote Sensing Laboratory was formed within
Iowa Geological Survey to coordinate remote sensing activities in the State of Iowa. Remote
Sensing Laboratory staff within IGS analyzed the imagery acquired in June 1971 and
developed thermal maps for portions of the river from Clinton to Keokuk, Iowa.
V - 19
-------
An uncalibrated scanner, mounted in the Twin Beechcraft from the Remote Sensing
Institute, was used for the June mission. The unclibrared scanner from the Remote Sensing
Institute utilized the 4.5 to 5.5 micron portion of the infrared spectrum and thermal data
was recorded, in flight, directly on film. A Barnes PRT-5 radiometer tracked nadir below
the aircraft but was not roll compensated with the scanner. The June mission was flown at
or near noon and thus both the heating effects of the sun and reflective effects of radiation
at near infrared wavelengths were detected. Although the mission design was not optimal
for thermal mapping purposes, analysis of thermal imagery revealed that valuable information
was derivable from the imagery, particularly for purposes of monitoring systems design.
The Iowa State Conservation Commission, State Hygienic Laboratory, Iowa Attorney
General's Office, and U.S.Geological Survey, Water Resources Division, were informed
of the mission and they provided assistance in the form of manpower, boats, and equip-
ment. The U.S.Army Corps of Engineers, Rock Island, were informed and they agreed to
maintain the stage of the river as constant as possible during the mission.
Thirteen ground data acquisition teams, on the June 1971 mission, obtained records
of ground conditions as they existed at the time of the overflight. This data is required
to identify and quantify the images obtained by the multispectral camera and infrared
scanner. These teams were assigned cross-channel traverses spaced along the 180-mile
length of river from the Commonwealth plant site to Keokuk. The three-man teams
consisted of personnel from the Iowa State Conservation Commission, the Iowa Geolo-
gical Survey, and U.S.Geological Survey, and representatives of Commonwealth Edison.
Boats were provided by the Conservation Commission and water-sampling kits were
distributed through the State Hygienic Laboratory. Mobilization of this large ground
data collection effort was accomplished in a short period of time, with no direct project
expense involved.
The thirteen ground information stations were located to sample a variety of river
conditions, such as single-channel areas, multiple channel areas, backwater sloughs,
urban areas, proximity to locks and dams, and incoming tributary drainage. Along
the route of each open river traverse, a minimum of five temperature-sampling points
were designated. These points were established in relation to some recognizable island
or landmark. Each team was on the river from 0900 hours until 1600 hours making
traverses across the river channel. Teams had certified thermometers for recording
surface water temperatures. Thermometers were held completely immersed, parallel to
and fust beneath the water surface. Air temperatures were also taken at the beginning
and end of each traverse. Ground teams recorded 874 temperature readings from 117
sample points along the 13 traverses and the ground data represents the most comprehensive
compilation of water temperatures and climatological information ever amassed in a given
time period on the Mississippi River.
On 4 June 1971 arrangements were made for two coverages of the study area at
10,000 feet above ground level and one at 3,000 feet. When the mission was actually flown
V - 20
-------
a third level at 6,000 feet was added. Imagery from the first flight at 10,000 feet
was not satisfactory due to cloud cover. The second flight was flown at 2,850 feet
from Keokuk to Clinton. Some clouds appear on all imagery. The 6,000 foot imagery
was flown when the cloud problem abated and this imagery was the best acquired
during the June 1971 mission.
Imagery was analyzed by both visual means and automated electronic density
slicing. The visual evaluation of the thermal imagery was conducted using a Richards
light table and microscope. Thermal outfalls were noted and traced to their respective
sources. Analysis showed that during June, rivers flowing into the Mississippi are
generally hotter than the main channel water by several degrees. The thermal outfalls
from the Muscatine and Riverside electric power generating plants were at least 10°F
hotter than the channel water and their plumes extended at least one quarter mile
downstream. Slough areas were particularly hot owing to their relatively static condition.
Islands in the river were relatively hot and long sinuous plumes from these islands often
extended as much as one mile downstream. The thermal technique mapped surface
current distribution patterns in the river with remarkable accuracy. These patterns
indicate areas of slow moving water, the thalweg of the main channel, and areas of
rapid moving water. Areas of slow moving water are not subjected to the turbulent
flow present in the main channel and the outside loops of meanders. Therefore slow
moving water areas appear warmer because the water heated by the sun is not mixed
with colder water stratified below. By identifying areas of rapid flowing water, slow
moving water, and almost static water, and relating these areas to the geometry of the
river system, distribution patterns in the river are identified. The streaming effect from
islands, tributary drainage, and shallow areas is analogous to the streaming effect observed
by placing dye in a river system model. Although the original mission was designed to
evaluate thermal discharges in the river and to sample the thermal regimen of the system,
the major information derived from this study was information on surface current distri-
bution. This type of information could be essential for the development of spill removal
plans, but even more significantly it could provide information for the placement of
monitoring instrumentation.
Automated analysis was performed on the thermal imagery to produce an isothermal
map. The original film transparencies, obtained during the overflight, were used for
this analysis. Film density is directly related to the energy received by the scanner,
and equal film densities represent equal temperature levels on thermal imagery. By
correlating known surface water temperatures with film densities, point data can be
extrapolated and isotherms established without the element of human interpretation involved.
Film sections were analyzed on an electronic television density analysis device. This
apparatus consists of a light source, film holder, television camera, density digitizing
and coloring electronics, and a television camera. Color encoded isodensity mosaics
were produced covering 74 miles of river. The relationship between temperature and
density was not well defined. The discrepancy in relating temperature to density may
be a function of the thermal scanning equipment itself. The scanner used on the June
mission was not internally calibrated and utilized the 3 to 5 micron band. No reference
V - 21
-------
ISOTHERMAL MAP
OF
MISSISSIPPI RIVER
Cordova Slough —Lock 8 Dam I4||||
4 June, 1971 111
;;•;•$!:; RAPIDS'
*m CITY :
ISOTHERMAL MAP OF MISSISSIPPI
4 JUNE 1971
V -22
-------
temperature was built Into the system so that It might recalibrate itself. With time and
possible associated internal changes, the scanner response may change slowly.
Atmospheric conditions may also account for the imaging differences and particularly
atmospheric humidity may vary downstream. The scanner operated in the 4.5 to 5.5
micron range of the electromagnetic spectrum. This range is particularly susceptible
to atmospheric effects and even partially recordes reflected solar radiation.
A serious problem resulted because most of the recorded temperatures were obtained
within a small temperature range. Few known points were available at higher temperatures
to determine the temperature-density relationships.
While the most important results of this study are qualitative, the semiquantitative
analysis does show that an accurate quantitative map of river temperature is possible
with some refinements of technique and with use of proper equipment. Imagery
was obtained with a 4.5 to 5.5 micron scanner, close to noon, and emitted radiation
plus reflected radiation was detected and mapped. The thermal scanner was not calibrated
and thus instrument drift could not be assessed. Ther was no way of obtaining an exact
trace of the PRT-5 radiometer track. Ground temperature measurements did not correspond
to the time of the actual overflight. Ground temperatures concentrated almost
exclusively on main channel temperatures, thus creating a density of calibration points
near one temperature but few points from which to construct a relationship between
temperature and density. Water temperature measurements were made close to shore,
dams, or bridges and thus the instantaneous field of view of the scanner incorportated
temperatures of these features as well as that of the river water.
In spite of the numerous problems identified this study, it was established that airborne
remote thermal mapping of a large river system is a considerable economy, in terms of
time and money, over the acquisition of conventional point data. Remotely sensed thermal
data can be acquired over hundreds of miles of river in a single day, using only a
small number of ground personnel and a limited amount of equipment. Quantitative
results can be obtained if proper mission design Is executed. Mission design should
include a blackbody reference calibrated scanner operating in the 8 to 14 micron
range of the electromagnetic spectrum, preferably with the two references set at
the high and low ends of the temperature scale, and with an aircraft mounted PRT-5
radiometer trace electronically located on the scanner imagery. The scanner, airborne
PRT-5 radiometer, and ground PRT-5 radiometers should be calibrated simultaneously
before and after the mission. Good analytical results are obtained for missions flown
over river systems at an altitude of 3,000 feet to 6,000 feet above river level. To
properly evaluate thermal regimen of a river system, the mission should be flown
twice during the day, at noon to assess the heating effects of the sun, and after
sunset but before fog sets in, to assess the heating effects due primarily to manmade
sources. Ground information teams should measure river temperatures with portable
radiation thermometers (which measure radiant spectral emittance 8-14 microns), and
with vertical probe strings inserted at varying depths in the river. Air temperature,
humidity, wind velocity and direction, and climatological phenomena should be
recorded at the instant the aircraft passes overhead, and these variables should be
assessed during the entire mission. General weather conditions should be observed for
one week prior to the mission.
V - 23
-------
VISUAL MAPPING TO IDENTIFY SOURCES OF THERMAL EFFLUENT ON THE MISSISSIPPI
A low-afHfude reconnaissance flight was flown on 14 February 1972 to make a
visual and photographic record of all open water areas on the Mississippi River between
Keokuk, Iowa and fhe Iowa-Minnesota border. The purpose of the survey was to
identify points on the river at which warm effluents are being discharged. The behavior
of water during icmg is such that open water can result from three significantly
different conditions:
I. First is the obvious condition of imputs of wafer sufficiently above the temperature
of 0°C that loss of heat to the atmosphere is not at a rate high enough to cause ice to
form at the point of imput and/or downstream for significant distance. This type of
imput could have three types of sources in fhe Mississippi River in Iowa;
a. Industrial, municipal and/or domestic outfalls,
b. Discharges of groundwater, and
c. Upwelling of warmer bottom waters below dams.
2. The second condition that account for open water is turbulence of flow. The
occurrence of these conditions can be definitely recognized in fhe Mississippi River
bordering Iowa only below dams that are set as weirs. Open water between dams
that has no physically traceable source to hot-water outfall may be the result of
upwelling warmer bottom flow or purely the result of turbulence where elements of
the current converge. More likely, they result from a combination of both turbulence
and melting.
3. The third circumstance that accounts for open water during icing conditions is
physical disturbance of the ice layer, like that caused by an icebreaker.
The visual and photographic reconnaissance was accomplished by utilizing fhe
Iowa State Conservation Commission aircraft and a hand-held 35 mm camera. The
date in February was selected because prior to that time temperatures in the area
had been continuously below freezing. Twenty-seven major open water stretches
were detected on the Mississippi River as it borders Iowa, even though temperatures
ranged below zero for almost 20 days prior to fhe overflight. Ten sources were directly
related to open water below dams, while seventeen sources were traced directly to
industrial and/or municipal outfalls. Four of these seventeen sources emanated from
the Wisconsin side, one from the Illinois side exclusively and one from the Illinois
side in part. Examination of atmospheric data indicates that the amounts of heated
effluent being discharged at 12 of the sources must have been very significant and at
fhe balance significant in any evaluation of the unnatural elements of the thermal
regimen of fhe reach of the Mississippi River bordering Iowa.
V - 24
-------
Patch of thin ice
Open wafer
Distribution of Open Water and Ice Cover
on the Reach of the Mississippi River
bordering Iowa
14 February 1972
Muscatine Co.
Louisa Co.
Patches of thin Ice
LOCK a DAM 17
V - 25
-------
USE OF REMOTE SENSOR DATA FOR THERMAL MONITORING SYSTEMS DESIGN
Both the June 1971 thermal mapping overflight and the February 1972 visual and
photographic overflight to detect thermal discharges were relatively inexpensive data
collection efforts which provided extensive information on the thermal regimen of the
Mississippi River as it borders Iowa. The thermal mapping mission cost $2,800 for data
acquisition and the visual and photographic overflight cost $290.00 for aircraft and
film. By utilizing state agency personnel, equipment, and boats a large ground infor-
mation collection team was mobilized, on short notice, at no expense to the projects.
Even though mission design was not optima I, satisfactory data was obtained in a short
period of time over an extensive portion of the Mississippi River. The synoptic coverage
of the thermal mapper data allowed generalizations to be made on the thermal regimen
of the Mississippi River. Thermal mapper data indicated where and how ground mea-
surements should be taken to quantify future thermal imaging missions. Further, analysis
of June thermal imagery and photographic imagery obtained in February indicated areas
where ground monitoring instrumentation could be placed to monitor existing outfalls
and measure the effects of any new sources. The dynamic thermal equilibrium of a
major river can thus be efficiently monitored by strategically locating monitoring
instrumentation. When anomalous thermal levels are detected by ground monitoring
instrumentation then another thermal overflight could be undertaken to accurately map
new thermal effluents to their sources. Thus highly repetitive, and therefore expensive,
thermal mapping investigations would not be required for a monitoring system located on
major rivers.
FLOOD MAPPING AND FLOODPLAIN ANALYSIS IN IOWA
Floodplain management-planning in the midwest has major implications in the area
of environmental control and water quality monitoring. Probably one of the most dif-
ficult questions to answer is, "What constitutes the floodplain for purposes of landuse
planning?" Floods are one of the worst natural hazards in the Midwest. Each year the
Midwest experiences flooding caused by the spring thaw and frontal rainstorm activity.
Annually millions of dollars' worth of damage results to homes, businesses, public
works, and crops. Much time and energy is devoted by government agencies to study
floods, to assess immediate damages, and also to gain a perspective from which to base
future decisions concerning floodplain management. Iowa Geological Survey, Remote
Sensing Laboratory (IGSRSL) and the U.S.Geological Survey, Water Resources Divi-
sion, entered into a cooperative program to develop aerial methods for mapping mid-
western flood inundation on a seasonal basis.
Three major floods were analyzed by IGSRSL staff to develop a seasonal flood
mapping technique for Iowa. Late summer flooding was studied by Hoyer and Taranik,
1972, on the West and East Nishnabotna Rivers in Western Iowa. Records indicated
that the Nishnabotna River flood had a statistical probability of recurring once every
100 years. The time of imagery acquisition ranged from three days after flood crest in
September 1972 to during flood crest. ERTS imagery was acquired seven days after
V - 26
-------
'Essex
I Ongmol Scale:
1 I; 250,000
Inundation Map —
Nishnabotna Basin Flood,
September, 1972
Agreement of Low-Altitude and
ERTS Imagery Boundaries
^ Area of Agreement
Low-Altitude Imagery Boundary
ERTS Imagery Boundary
Indefinite Low-Altitude Imagery Boundary
Indefinite ERTS Imagery Boundary
tiRiverton 0
10
ZOmiles
•"¥•"
*agga*
B.E.Hoyer
G.R.Hailberg
Hamburg
0 5 10 15 20 25 30kilometers
Iowa Geological Survey February, 1973
v - 27
-------
flood crest. Maps of the Inundated area were prepared by Interpretation of the various
types of imagery and all known ground data agree with this mapping, in early January,
1973, an ice Jam developed at the lower end of the Iowa River and artifically raised
the river stage until the water flowed onto the adjacent floodplain. The flooded area
was readily apparent to airborne observers as It remained as smooth ice surrounded by
rough Ice and snow. The last flood analyzed was a spring flood along the Skunk River,
which occurred on 23 April 1973. The Skunk River experienced the greatest flood on
record following heavy rains on saturated ground.
Iowa Geological Survey, Remote Sensing Laboratory staff utilized a multiband
camera, metric camera, and hand-held cameras with a variety of film filter combinations
to map these three floods. The following conclusions summarize the results of these
studies.
1. Photographic infrared radiation is by far the most important and useful spectral
region for mapping floods either after flood cessation or concurrent with flood crest. The
characteristics of infrared radiation, including the absorption of these wavelengths by
water, the reduced infrared reflectance of wet soils and stressed plant species and dif-
ferences in reflectance between snow and ice, make the infrared band the most useful
for flood inundation mapping.
2. The addition of color aids the interpretation in both the visible and near-visible
wavelengths. Black-and-white panchromatic films are generally unacceptable for flood
mapping and color films are generally better in the visible portions of the spectrum.
Black-and-white infrared film may produce a usable product for mapping flood inundation,
but color infrared film is superior. For multiseasonal flood mapping, especially for late
summer flood mapping, color infrared film (Kodak 2443) seems to be the best available
film. Multispecrral Imagery, combined with color additive viewing techniques, is
actually the best approach for flood inundation mapping. However data handling problems,
smaller areal coverages of the multiband camera, and the information needs of agencies
involved in floodplain management make multiband imagery less desirable than color
infrared imagery.
3. Stereoscopic viewing aids flood mapping especially in areas in which interpretation
based on color or tone Is difficult. Proper overlap should be provided to allow stereoscopic
viewing and to allow photogrammetric operations such as topographic map production. This
could provide data useful for engineers and persons in floodplain management and planning.
4. ERTS satellite data supports the conclusions of the low-altitude studies. Evaluation
of ERTS-1 imagery indicates that it is useful for small-scale, regional flood inundation map-
ping. The synoptic coverage of ERTS imagery allows rapid regional appraisal of flooding.
This may be particularly useful for agencies involved in regional flood control, disaster
relief, agricultural crop prediction, and flood insurance. Its mapping capability may pro-
vide a first approximation of flood inundation especially useful for large areas, and in
sparsely settled, poorly mapped, or inaccessible areas.
V - 28
-------
5. The detailed mapping of inundation from large-scale Imagery indicates a close
correlation between large magnitude floods and particular soil series. Further analysis
may allow the quantification of soils in terms of the frequency, magnitude, and dis-
tribution of floods. This, in turn, may possibly provide the capability for the detailed
flood inundation prediction required for rational landuse planning.
6. This aerial technique allows flood inundation to be mapped a minimum of seven
days after flood crest in late summer, and at least five days after flood recession in spring,
Winter flooding could be mapped in at least these time periods.
7. The recommended techniques indicated by these studies should provide more
detailed data for flood inundation mapping in a more cost-effective manner. Generally,
flood inundgtion mapping is best accomplished with color infrared film (Kodak 2443)
filtered to eliminate blue wavelengths (Wratten 12 filter), and exposed in a metric
aerial camera. Imagery should be acquired as soon as weather conditions permit, after
flood recession. Sufficient overlap should be included to provide complete stereoscopic
imagery both for interpretation and for photogrammetric purposes. The scale of the
imagery must be carefully determined from the potential imagery uses and the extend of
flooding.
USE OF REMOTE SENSING DATA FOR FLOODPLAIN MANAGEMENT-PLANNING
AND FOR MONITORING SYSTEMS DESIGN
Low altitude multispectral imagery of flood inundation in three seasons has indicated
a close correlation between particular soil series and inundation boundaries of major
floods. Detailed analysis of low altitude imagery coupled with comprehensive soils
mapping appears to yield information on the nature and extent of the floodplain. Pre-
liminary analysis has revealed that it may be possible to quantify particular soils in terms
of the frequency, magnitude, and distribution of floods which affect them. Analysis of
this kind could produce an operational definition of what constitutes the floodplain for
landuse planning purposes. An accurate definition of the floodplain is required, par-
ticularly for locating sanitary landfills, granting of permits for feedlots and sewage
lagoons, and for liquid and solid waste management.
Another aspect of low altitude multispectral photography of the floodplain was the
synoptic coverage of existing petroleum, chemical, and other industrial liquids in tank
farms and storage lagoons on the floodplain. Runoff over feed lot operations could be
traced into small drainages, and eventually into some major rivers. Therefore it seems
possible to designate the location of ground monitoring instrumentation from a single
photographic overflight. Further, along the Mississippi River, analysis of photography
not only located major liquid storage facilities but also yielded information on the
nature of berms and dikes surrounding these facilities. Should major flooding, fire, or
other disasters damage tanks or ponds, and thereby allow pollutants to enter hydrologic
systems, strategically placed monitoring instrumentation could detect pollutants early
enough to perhaps allow their efficient removal. When pollutants are detected by
V - 29
-------
strategically located ground instrumentation, a remote sensing overflight could be
arranged to map the geometry of the spill. This information could be utilized to
coordinate emergency removal procedures and document possible violations. Again,
highly repetitive, and thus expensive, airborne remote sensing overflights would not
be required for the development of a Hver-floodplain monitoring system.
CONCLUSIONS
An understanding of the surface current flow characteristics, channel geometry,
natural thermal regimen, extent and characteristics of the floodplain surrounding a
major river, coupled with information on the location of thermal sources, location of
major outfalls from treatment facilities, industries and agricultural land, and location
of liquid storage facilities could provide information for the development of a com-
prehensive floodplain and river monitoring system. A single remote sensing thermal
and photographic overflight can provide information for the strategic location of
ground instrumentation. This ground instrumentation can detect anomalous concen-
trations of pollutants and alert monitoring personnel so additional remote sensing
missions could be arranged to trace pollutants to their sources. In this manner ground
instrumentation could be efficiently located and redundancy in placement avoided.
At the same time, highly repetitive, redundant, and thus expensive, airborne remote
sensing overflights can be minimized.
REFERENCES
Parker, M.C., Ed., 1972, Proceedings, Seminar in Applied Remote Sensing, May 8-12,
1972: Iowa Geol.Survey, Pub.Info.Cire.No.3, 176p.
Tuthill, S.J., Taranik, J.V,, and Hoyer, B.E., 1973, Thermal Remote Sensing on the
Mississippi River in towa: AlChE Symposium Series No. 12^,vol .69, p.391 -400.
Tuthill, S.J., and Taranik, J.V., 1972, Remote Sensing, A Tool for Sj-ate Pianning-
Management in Iowa: Proc.Sth Internet. Symposium on Remote Sensing of the
Environment, Env.Res.lnst.of Michigan, vol.l, p. 11-20.
Tuthill, S.J., Hoyer, B.E., and Prior, J.C., 1972, The Mississippi River Overflight to
Identify Sources of Warm Effluent: Iowa Geol.Survey, Prelim.Report, 29p.
Hoyer, B.E., and Taranik, J.V., 1972, Aerial Flood Mapping in Southwestern Iowa,
A Preliminary Report: Iowa Geol .Survey, Prelim.Report No. 1, lip.
Hallberg, G.R., and Hoyer, B.E., 1973, Flood Inundation Mapping in Southwestern
fowa; A Preliminary Report, Analysis of ERTS-1 Imagery: Iowa Geol.Survey,
Prelim,Report No.2, 15p. ~
V - 30
-------
Hoyer, B.E., Hallberg, G.R., and Taranik, J.V., 1973, Seasonal Multispectrai
Flood Inundation Mapping in Iowa: Amer.Soc.Photogram., Sioux Falls, South
Dakota, Symposium, October 1973.
Hallberg, G.R., Hoyer, B.E., and Range, A., 1973, Application of ERTS-1 Imagery
to Flood Inundation Mapping, in Symposium on Significant Results Obtained from
ERTS-J, Goddard Space FHght'C'enter, NASA.
, 1973, Application of ERTS-1 Imagery to Flood Inundation Mapping: in
Significant Papers oh ERTS-1, Plenary Session, Goddard Space Flight Center, NA~§A,
in color.
Hoyer, B.E., Wiitala, S., Hallberg, G.R., Steinhilber, W.L., Taranik, J.V., and
Tuthill, S.J., 1973, Flood Inundation Mapping and Remote Sensing in Iowa; Iowa
Geol.Survey Public Info.Clre.No.6, in final preparation.
Taranik, J.V.(in preparation), Thermal Mapping of Hydrologic Systems in Iowa; Iowa
Geol.Survey, Public lnfo.Circ.No.8.
Cooper, R.I., Taranik, J.V., and Tuthill, S.J.(in preparation), Water and'Land
Planning in South-Central Iowa Using Remotely Sensed Data from ERTS-1: Iowa
Geol .Survey, Public lnfo.Circ.No.9. r
V - 31
-------
TECHNIQUES AND PROCEDURES FOR
QUANTITATIVE WATER SURFACE TEMPERATURE
SURVEYS USING AIRBORNE SENSORS
By
E. Lee Tilton, III,
Kenneth R. Daughtrey
and
Earth Resources Laboratory Staff
NASA/Earth Resources Laboratory
Lyndon B. Johnson Space Center
Mississippi Test Facility
Bay St. Louis, Mississippi 39520
V - 32
-------
I, INTRODUCTION
The NASA Johnson Space Center's Earth Resources Laboratory (ERL)
located at the Mississippi Test Facility has as part of its responsibi-
lity the development and demonstration of remote sensing techniques and
procedures that lend themselves to describing, evaluating or quantifying
our natural resources and the state or condition of our environment.
Water characteristics are basic parameters of our environment and a key
measurement related to the condition of the water is temperature. Air-
borne remote sensing techniques for water surveys, with their ability to
cover large areas quickly, have a number of potential applications. An
obvious use is the location and measurement of outflows that are at a
different temperature than the receiving body of water (rivers, lakes,
etc.). Temperature surveys are also a routine part of practically all
oceanographic studies and provide useful information for marine resource
habitat assessment in coastal waters and estuaries.
In the past, large area temperature surveys have been conducted using
portable and stationary insitu instruments. However, the data collected
is usually not synoptic and the cost is great. During the past several
years remote sensing techniques have made it possible to obtain large area
surface temperature data on a qualitative basis at relatively small expense.
Recently, these remote techniques have been improved to allow the quantita-
tive measurement of water surface temperature on a routine basis with con-
sistent reliability and accuracy. The purpose of this paper is to describe
the procedures, and the techniques on which they are based, for a user to
properly plan and conduct a quantitative survey of surface water temperature,
Procedures and techniques described here for remote sensing thermal
properties of water are those used at ERL and are based primarily upon a
series of experiments (Ref. 1) performed in the Mississippi Sound during
1972 using a Texas Instruments RS-18 airborne thermal scanning radiometer,
hereafter referred to as the scanner or thermal scanner, mounted in a
light aircraft. (Ref. 7) However, the procedures described should be
adaptable to other data acquisition and processing equipments and may be
used with high altitude aircraft and satellite data.
This paper is a condensed version of a detailed ERL report (Ref. 2) on
techniques and procedures in which each step is supported by discussion and/
or appendices which present hardware and/or software descriptions, specific
procedures, or descriptions and examples of forms. It is assumed here that
the hardware and software required to carry out the steps in the survey are
available and checked out. The major elements of the hardware and software
systems are shown in Figure 1. The remainder of the paper describes in
general the steps, as illustrated in Figure 2, which are required to plan
and conduct the temperature survey.
V - 33
-------
DATA ACQUISITION
THERMAL SCANNING RADIOMETER
LABORATORY AND AIRBORNE CALIBRATION SOURCES
SCANNER PLATFORM (AIRCRAFT OR SATELLITE) AND
DATA RECORDING SYSTEM
DATA VERIFICATION/
QUALITATIVE PRODUCT
TAPEDECK AND OSCILLOGRAPH
TAPE TO" FILM CONVERTER
PHOTO PROCESSING FACILITY
QUANTITATIVE PRODUCT
ANALOG TO DIGITAL CONVERSION-
SDS930 COMPUTER AND SOFTWARE
DATA PROCESSING-UNIVAC 1108
COMPUTER AND SOFTWARE
DISPLAY PREPARATION-SC4020
PLOTTER AND SOFTWARE '
FIGURE 1. MAJOR HARDWARE AND SOFTWARE ELEMENTS
V - 34
-------
WORK DAYS TO
COMPLETE FUNCTION
0
13
15
FLIGHT AND SURFACE
MEASUREMENTS REQUEST
PREPARATION OF FLIGHT
SURFACE MEASUREMENTS
PLANS/READY SYSTEM
IMPLEMENTATION OF
INFLIGHT AND'SURFACE
OPERATIONS
QUICK LOOK/DATA
VERIFICATION AND
USER REVIEW
QUALITATIVE/ANALOG
PRODUCT
QUANTITATIVE/DIGITAL
PRODUCT
FINAL PRODUCT
AND
USER REVIEW
COMPLETE
DATA
ACQUISITION
REQUIREMENTS
SECTION II*
DATA
ACQUISITION
SECTION III*
\
\
DATA
PROCESSING
SECTION IV*
* SECTION OF DOCUMENT WHERE FUNCTIONS ARE LOCATED.
TEMPERATURE SURVEY SCHEDULE AND FLOW DIAGRAM
FIGURE 2
V - 35
-------
II. DATA ACQUISITION REQUIREMENTS
A. General
Development of data acquisition requirements is the first step to be
performed in planning for a surface water temperature survey as indicated
in Fig. 2. The following discussions comment on the methods used by ERL
in the planning process to establish requirements and determine most of the
factors to be considered in order to optimize the acquired data for the in-
tended use.
B. Flight and Surface Measurement Requirements Documents
The Mission or Survey Flight and Surface Measurement Requests are
documents (forms) initiated and filled in by the data user. They serve
as documents that delineate the objectives and all requirements ot the
survey to all participants. These request forms include: project name,
user name, data acquisition date, flight and surface requirements and
constraints, communications, sensor and calibrations requirements.
C. Data Optimization through Choice of Flight Conditions
During preparation for a data acquisition flight with the thermal scan-
ner several conditions under which the data is taken should be considered
to optimize the data as much as possible. The conditions to be chosen for
the flight are somewhat dependent on the purpose for which the data is being
taken as noted in the following discussion.
1. Time of Day - Since the signal detected and recorded by the scan-
ner system is a function of emitted and not reflected solar energy,
thermal measurements may be taken any time of the day or night. How-
ever, two factors should be considered in the choice of flight time; one
is cloud cover and the other is land/water differentiation.
Cloud cover above the aircraft normally is not a direct considera-
tion when recording remote thermal data. Indirectly it can be a con-
sideration if the purpose of the data is to study short term water
surface characteristics such as insolation. However, in most cases a
more serious concern is cloud cover below the aircraft. Since clouds
are opaque to radiation in the thermal region those surface areas
covered by clouds cannot be surveyed. Therefore, the choice of flight
time should be made by considering the normal weather patterns of the
area. For example, weather conditions in the coastal areas along the
Gulf of Mexico are such that the waters just off-shore are usually free
of clouds in the morning in the summer time. In addition, the amount
of cloud cover is important. If one wishes to measure gross features
of surface temperature distribution, 50% cloud cover below the air-
craft is probably an acceptable condition. However, if one is looking
for specific detailed thermal patterns or requires the measurements in
a specific area cloud cover requirements should be 0 - 107.. Cloud cover
in excess of 50% is usually not acceptable because of the difficulty
in making atmospheric corrections to the data and because the scanner
V - 36
-------
detector has a time constant which does not allow an adequate signal
recovery when the field of view passes from the cold cloud top tempera-
tures to the water surface temperatures in the small cloud free areas.
The second area of consideration is the differentiation of land/
water interface. In areas such as marsh where the land and water areas
form complicated patterns one may rely on the thermal imagery to de-
termine the location of data, i.e. in cases where aerial photography
is not acquired with the thermal imagery for data location and no navi-
gation information is available. At certain periods of the day and
night land and water surface radiometric temperatures approach each other
and may be indistinguishable in the thermal imagery. This can occur
during a time when lighting conditions do not make aerial photography
possible such as twilight or early morning and, therefore, such times
should be avoided to the extent possible.
In some cases time of day for the flight may be fixed by other
considerations such as surface conditions, requirements for thermal
data correlation with other sources of data or by operational con-
straints. For example, in coastal waters, flights are sometimes
scheduled^ for particular tidal conditions. Unplanned events such as
fish kills may be the object of an investigation requiring data col-
lection under non-optimum conditions. Another consideration may be the
availability of surface measurements for scanner calibration. In larger
bodies of water high sea state may sometimes limit this availability
if small boats are involved. However, this is usually a minor con-
straint for remote thermal measurements since only one boat is required
for relatively small areas.
2. Flight Line Determination - Once the time of day is chosen the
flight line determination must be made. If'the survey area can be
covered by one flight line the altitude of the aircraft is usually
set such that the scanner field of view covers the area with some addi-
tional coverage on the sides of the scan to account for navigation
error. If the area is large, one must cover the area with parallel
flight lines again allowing an overlap with the adjacent flight line to
account for navigation error. Typical overlap is 20%. In this case
the aircraft altitude is usually kept as high as practical to maximize
coverage and reduce the number of flight line miles. This is done for
both economy reasons and to make the entire data set as synoptic as
possible. Synoptic data is particularly important when studying dyna-
mic surface conditions.
A significant consideration in flight line planning is that of navi-
gation. For large geographic areas requiring multiple flight lines an
automatic navigation system is most desirable. However, this type of
navigation is frequently not available and one must resort to visual
navigation using surface landmarks. Under these conditions one must
pay attention to flight line layout to make the navigation task as
simple as possible for the pilot. For large bodies of water, for in-
stance, the preferred technique is to lay out parallel flight lines
perpendicular to the shore so that shoreline landmarks may be used for
navigation and in addition to serve as an aid in location identifica-
tion using photography and scanner imagery during data verification
V - 37
-------
and analysis. In areas such as marsh where no distinguishable land-
marks are available one may resort to deploying artificial landmarks
for navigation. If navigation aids such as radio beacons are available
one should lay out the flight lines to take maximum advantage of these.
Unless lighting conditions prohibit, high resolution aerial photography
is usually taken simultaneously and time correlated with the scanner
data to provide a means of data location during data reduction and
analysis. Sun angle (time of day) may then become a factor in choice
of flight ime.
Prevailing wind direction and velocity may also be a factor in
flight planning for two reasons. First, a cross wind at flight altitude
can cause an aircraft crab angle resulting in skew in the scanner
imagery. This may be corrected during digital processing but is diffi-
cult because the crab angle usually varies during flight. Crab angle
of less than 5° is usually acceptable without correction. Cross winds
also make navigation difficult when no or only simple navigation aids
are available.
The second reason for considering wind during flight planning is
the necessity for maintaining constant ground speed when flying flight
lines. Changes in ground speed cause changes in scale in the scanner
imagery and introduce an additional complication in data analysis,
especially when data acquired on parallel, adjacent flight lines in
different flight directions are to be compared or a large area thermal
contour map is to be constructed from the data.
V - 38
-------
III. DATA ACQUISITION
The ERL thermal survey data acquisition equipment consists of a thermal
scanner and data recording system mounted in a Beech Model E-18 twin engine
light aircraft. A block diagram of the system is presented in Figure 3.
A. Data Acquisition Preparation
1. Mission or Survey Flight Plan
The mission or survey flight plan parallels the same general out-
line as the mission flight and surface measurement requests. It is
developed to provide additional information for the guidance of the
flight crew presenting priorities and details relative to flight lines.
The information specifically covered in the document includes:
o Flight objective
o Sensor requirements
o Flight requirements
o Flight line priority
o Communications (radio)
o Mission constraints
o Flight schedule (timeline)
o Flight line requirements
o Flight line map.
The surface measurements request for boat operations is used also
as the surface measurements plan; therefore, no parallel document to the
flight plan is needed for most surveys.
2. Calibration and Checkout of System
The scanner equipment manufacturer includes system calibration and
checkout procedures in the operation and maintenance manual. (Ref. 5)
However, it is usually more desirable for the user to develop his own
procedures for a particular system using the operation and maintenance
manual as a guide. Such a procedure is utilized by ERL on a periodic
basis to assure proper performance of the equipment. After the lab
procedures are performed, the system is installed in the aircraft and a
pre-survey checkout is performed before each flight.
3. Other Preparation Activities
Prior to performing a survey the pilot must file an appropriate
flight plan that meets requirements of the FAA officials in the survey
area. The plan must be in accordance with flight lines to be flown
during the survey. A weather forecast or status should be obtained from
the National Weather Service for the survey area prior to flight time
to assure that meteorological conditions are within constraints speci-
fied in the Survey Flight Request and Surface Measurements Reques.t.
The camera system selected for the survey should be installed in
the aircraft with the proper film filters, framing rate (overlap), and
V - 39
-------
IRIG A Time Code
Generator
±
Remote Display
Fixed Data Inserter
Time
Mission
Line
Run
PCM
Camera
System
Voice Intercom
Aircraft Radio
Long Range Radio
Satellite Radio
Camera Control Unit
A/C
Power Control Panel
System
Patch
Panel
Intercom
Control
System
Voltoeter
Oscilliscope
Power
Supply
Control
Panel
Thermal Scanner
Tape Speed Compensation
Signal Processor
AR - 700
Tape Recorder
1
14x2 Switch Panel
AC Calibration
DC Calibration
-AC or DC Power to all Units
Block Diagram
ERL AIRCRAFT THERMAL SCANNER AND DATA RECORDING SYSTEM
FIGURE 3.
V - 40
-------
boresighted to the scanner field of view to meet the photographic
documentation requirements and lighting conditions during the survey.
B. Data Acquisition Implementation
During the implementation phase of data acquisition the primary ob-
jective is to acquire usable data that meets the requirements of the flight
and surface measurement plans. The following paragraphs discuss the acti-
vities which take place during the flight and surface operation.
1. Communications required for a survey depend on the complexity of the
acquisition activities and the ability of the boat and flight crews to
follow plans developed for the survey. Generally, a radio communica-
tions link is not required between the surface measurement boat and air-
craft crew unless problems are encountered during the flight. If real
time communications are used planning should include the designation
of the radio frequency on which the crews will be operating.
2. Navigation
Scanner data spatial quality is influenced by how well the aircraft
follows the designated flight line. That is to say, the desired goal
is to go from point "A" to point "B" in a straight line with a minimum
of aircraft maneuvering. The success one can expect in achieving the
goal is a direct function of the sophistication in aircraft navigation
systems. The three levels of navigation one can encounter are:
a. Dead reckoning
b. Homing beacon - automatic direction finding
c. Inertial navigation system - auto pilot control
Dead reckoning requires that the pilot be familiar with flight
line requirements (lines on maps or photos) and be able to translate
these requirements into a real time flight path. Results may vary to a
minor extent depending on the pilot's experience. A homing beacon
is an electronic assist to the pilot since it allows him to direct
the aircraft to or from a known fixed point. Accuracy in flying straight
lines is improved with the assistance of this equipment, but the homing
beacon is most useful for single flight lines. The inertial navigation
system is the ultimate in flight avionics but as such is the most ex-
pensive. Using this automatic navigation equipment, position accuracy
greatly increases and the aircraft is auto-piloted in a straight line
(large maneuvers are suppressed) and also the craft can be steered
accurately an successive parallel lines.
Because of the inherent cost of the inertial navigation system it
is seldom encountered except in the largest of flight programs. Most
activities use dead reckoning or homing devices and as such live with
the degree of compromise these methods require.
V - 41
-------
3. Camera Operation
Photographic documentation should be gathered whenever the thermal
scanner is gathering data on a flight line for later documentation and
data location purposes. For altitudes below 10,000 feet a 70mm film
format with 40mm lens has been found adequate for most purposes. For
open water work where the location of small objects such as buoys is
necessary for data locations a larger film format or longer focal length
lens may be necessary. Film type may be dependent on other user re-
quirements but color infrared has been used primarily by ERL because
of its insensitivity to atmospheric haze and its enhancement of sur-
face features. However, color infrared film may not be acceptable when
poor lighting conditions prevail because of its relatively slow speed.
4. Thermal Scanner Operation
Inflight operation of the thermal scanner requires that the system
be allowed at least a ten minute warm up period prior to acquiring data
on the first flight line. The roll correction gyro must be erect to
prevent 'skewing of the imagery and the thermal scanner gain control
must be adjusted over water in the surver area rather than land to obtain
maximum signal range of surface water radiometric temperature. Pro-
cedures for the operation of the thermal scanner are given in Ref. 2.
5. Flight Logs
The flight logs are a necessary form for record keeping, so that
important flight information is preserved for future reference. Para-
meters recorded on the logs figure prominently in subsequent data pro-
cessing for determination of start and stop times, for recording gain
settings, calibration, voltage times and levels, air and ground speed,
altitude and anomalies experienced. An example of flight logs used by
ERL is contained in Reference 2. These include the thermal scanner log,
magnetic tape calibration log, photographic log and aircraft log.
6. Voice Annotation
Voice annotation of the magnetic tape is important for two reasons:
a. Recording pertinent information necessary to interpret tape
recorder calibration data.
b. Placing flight parametric data onto tape to assist subsequent
data tape handling and data processing.
Voice annotation is usually placed on one tape track and records all
voice communication via the sensor crew intercom and the air-to-ground
radio links. Because the voice annotation is activated while data is
being taken, the information on this tape track complements and augments
the data on the airborne operator's logs.
V - 42
-------
7. Surface Measurements
Advances in the state-of-the-art of thermal remote measurement
from aircraft have reduced surface measurements for calibration pur-
poses to a minimum. For small geographic areas where atmospheric con-
ditions are uniform only one boat is required. Reference 2 gives the
procedure for taking surface temperature measurements from a boat. In
addition to the water temperature measurements, some meteorological data
such as wind speed and direction, air temperature, humidity and cloud
cover distribution and altitude are sometimes required or at least help-
ful for atmospheric corrections.
Other than the usual logistic problems with boats the only consi-
deration relative to remote thermal measurement calibration is that of
coordination with the aircraft. To^assure the maximum validity of the
surface measurement it should be acquired at a location that can be
easily identified in the remote data and it should be acquired at the
same time that the aircraft is at the same location. Alternately, a
series of measurements around the time of aircraft flyover would allow
the construction of the time history of surface temperature at the
desired'location. Very often the circulation characteristics in a tar-
get area are dynamic enough to require a measurement at the time of fly-
over. However, in all cases where a boat is used it is necessary to
identify the location and time of the surface measurement in the remote
data because of the methods used for making atmospheric corrections.
8. Post Acquisition Activities
After completion of flight and surface measurement activities the
photographic film, magnetic tapes, flight logs and surface measurement
logs are returned for data processing, cataloguing and indexing.
Photographic film is submitted immediately to a photo lab for processing
and production of a positive transparency product. Magnetic tapes and
flight logs are returned to a quick-look processing facility. Surface
measurement logs are retained for later use in the data processing pro-
cedure.
At the present time at ERL the procedures for determining overall
thermal scanner measurement accuracy are being developed. (See
Section V). If the final procedure dictates a post-flight calibration
procedure this activity should take place immediately after landing as
part of the post acquisition procedure.
V - 43
-------
IV. DATA PROCESSING
Data processing includes all functions performed on the analog data
tapes and 70mm film from the time those items are removed from the aircraft
and brought to the data center or laboratory until the final product is
ready to be prepared. The following data processing discussions and pro-
cedures will be divided into three parts: Quick-Look Data Verification,
Qualitative Analog Data Product, and Quantitative Digital Data Product.
A. Quick-Look Data Verification
This activity allows the user to evaluate the validity of the data
acquired in a. short time frame after the survey is conducted. Many times
problems can be discovered early before much effort is expended in pro-
cessing the data.
The data tapes as recorded during the flight are the basis for all
thermal information acquired. Precaution should be used and standard pro-
cedures followed when handling and processing the tape. A duplicate should
be run and the processing done from the duplicate with the original stored
in a safe place.
The following basic steps are required for quick-look data processing.
1. An oscillograph recorder provides a hard copy record of the data
acquired on the analog tape. The 0'graph record is used to get a quick-
look at the calibration, sync and video signals, to assess system
operation and verify calibration and data acquisition times during the
survey. An alternate method of quick-look processing involves the use
of an oscilloscope. However, this method is more time consuming and
there is no hard copy record of the data produced except that photo-
graphs may be taken of the scope to document specific problem areas on
the tape.
2. The ground based RFR-70 film recorder used with the RS-18 thermal
scanner converts the analog signal (light energy) to a 70mm photographic
film via a cathode ray tube (CRT) trace, which is focused by a lens and
mirror onto the film that is moving across the face of the CRT tube.
This film must then be developed and copies run by standard processes.
The film image is a black and white representation of surface tempera-
tures where the different shades of gray represent different surface
temperatures. Usually a film positive is produced where the lighter
the shade of grey, the warmer the surface temperature. The light table
is used for review of this imagery. At this time in the procedure all
of the magnetic tape data is processed through the film recorder. Only
a small time slice is photographically processed into a black and white
transparency for data quality verification.
3. Simultaneously with the review of thermal scanner data on the
oscillograph and the tape to film conversion, the aerial photography
obtained from the camera boresighted with the scanner is processed.
It is reviewed during the quick-look process to aid in determining
whether the aircraft generally followed the planned flight lines and
V - 44
-------
whether landmarks, boats, buoys are locatable and identifiable for sub-
sequent determination of actual flight lines.
4. Utilizing data acquired from the quick-look data verification pro-
cessing a Quick-Look Report is prepared for the user to determine the
overall quality of the data. Inputs are made from the oscillograph,
oscilloscope, analog imagery and aerial photography. Reference '* gives
an example of a Quick-Look Report which consists of the following in-
formation:
a. Mission summary
b. Flight and surface measurement requirements
c. Aircraft flight plan
d. Operator's Logs
e. Magnetic tape quick-look report
f. Photographic quick-look report and flight map
g. Anomalies experienced.
In summary, the verification of the airborne data is accomplished by
review of an oscillograph record made from the analog data tape, a small
portion of t'hermal imagery processed from film obtained from the tape using
tape to film conversion equipment and the aerial photography obtained with
the boresighted camera.
B. Qualitative Analog Data Product
Assuming that the quick-look activity has verified the general quality
of the data, an interim product may be prepared which is suitable for pre-
liminary analysis of water surface thermal conditions. This imagery is
qualitative in nature since absolute temperature is not presented. Relative
surface temperatures may be inferred from the variations in grey tone. The
value of this product may be enhanced by the notation on the imagery at
appropriate locations of any surface measurements made during the survey or
mission. Supporting data consists of actual flight lines plotted on a ref-
erence map and derived from the photographic and/or thermal imagery.
The data product consists of a strip(s) of black and white imagery which
represents surface temperature in shades of grey covering a swath of area to
either side of the flight line, and an actual flight line map derived with
the aid of aerial imagery. An example of the product is shown in Figure 4.
It is produced from the data magnetic tape by using a film recorder as
described previously. It may be produced as a positive (lighter grey tones
represent warmer temperatures) or as a negative (lighter grey tones repre-
sent cooler temperatures) product depending on use or preference. For
example, clouds below the aircraft appear white in a negative and black in
a positive product. The product may also be produced as paper print or as
a transparency for use on a light table. Depending on use, the transparency
or print may then be annotated with the surface measurement temperatures
taken at the same time as the remote measurements. Location of the surface
measurement points in the analog product is accomplished using the geographic
features on the analog image itself aided by the boresighted aerial photo-
graphy. Relative surface temperature distribution can then be inferred from
the grey tone distribution. One problem with this product is that it is
difficult to match grey tones in areas of imagery overlap on parallel flight
V - 45
-------
S
c_>
ALTITUDE 3000 FEET
19 SEPTEMBER 1973
APPROXIMATE SCALE 1:32,000
ORIGINAL FORMAT FROM
TAPE TO FILM CONVERTER
J
E-
Q
U
E-i
K
ALTITUDE 6000 FEET
19 SEPTEMBER 1973
ORIGINAL FORMAT FROM
TAPE TO FILM CONVERTER
APPROXIMATE SCALE 1:74,000
11.5 SURFACE TEMPERATURES °C
• BOAT STATION
LIGHTER TONES INDICATE
WARMER TEMPERATURES
BILOXI BAY
INDEX MAP
QUALITATIVE ANALOG DATA PRODUCT
Figure 4
-------
lines during photographic processing and, therefore, it is difficult to
mosaic imagery from parallel flight lines. However, this has not proven
to be a serious constraint in the use of the data.
One particular advantage of this product is that it may be produced
almost immediately after acquisition, assuming the close availability of a
photo lab. At this time the user should review the entire set of data in-
cluding the aerial photography. The purpose of the review is to verify that
the quality of the thermal product meets user requirements and to select the
data to be further processed to produce the quantitative digital data pro-
duct. Detailed procedures for producing the analog product are given in
Reference 2.
C. Quantitative Digital Data Product
The quantitative digital data product obtained from the thermal scanner
is produced through computer processing from the analog magnetic tape and is
shown in Figure 5. The product may be left in digital format on tape for
entry into a data bank or may be produced as a grey scale image where each
grey level corresponds to a selected temperature interval. In addition,
numerical temperatures may be superimposed on the grey scale image giving
absolute temperature on a selected uniform pattern. This product is pre-
pared according to requirements specified by the user after review of the
analog product. The following discussion describes the techniques and pro-
cedures used in producing the digital product.
Data from the RS-18 thermal scanner is recorded on one track of an
Arapex AR-700 Analog Magnetic Tape Recorder along with housekeeping and cali-
bration data. Scan line synchronization data from the RS-18 is recorded
on a second track of the recorder. A data bandwidth of 50KHz requires that
the recorder, utilizing wide band Group II electronics, be run at 15 inches
per second or higher.
Recording the scanner data in an analog format requires that an analog
to digital (A/D) conversion be accomplished to synchronize the appropriate
pulses and time reference the calibration and video data for subsequent
computer processing. Certain parameters are required as inputs in order to
perform the analog to digital convarsion. These include:
1. The inflight tape calibration log for verifying calibration voltage
levels on the tape.
2. The Oscillograph quick-look record for verification of calibration
voltages and data time interval, and an indication of data quality.
3. The Quick-Look Report to determine whether anomalies exist on the
tapes and if so what correction factors could be applied.
Once these parameters are identified and the appropriate inputs made
ii'om this data the analog to digital process is ready to proceed. The A/D
conversion is accomplished on the SDS930 computer at the Slidell Computer
Complex (SCC). A description of the A/D software program is contained in
Reference 2.
V - 47
-------
<
I
Q
CJ
E-l
to
ALTITUDE 3000 FEET
19 SEPTEMBER 1973
\
SC 4020 MICROFILM ENLARGED
\
\
\
APPROXIMATE SCALE 1:32,000
r~
Q
O
ALTITUDE 6000 KEE1
19 SLI'TCMUEr-' l'J73
SC 4020 MICROFILM ENLARGED
APPROXIMATE SCALE 1:74,000
EACH OF EIGHT GREY LEVELS
REPRESENTS A TEMPERATURE
INTERVAL OF ONE °C
RANGE 25° - 32°C
TEMPERATURES > 32 - WHITE
TEMPERATURES < 25° - BLACK
BILOXI BAY
INDEX MAP
QUANTITATIVE DIGITAL DATA PRODUCT
Figure 5
-------
The output of program AP174 are digital data tapes containing counts
representative of the raw analog detector voltages and time. These digital
data tapes and the documents described below are required for the next pro-
cessing step.
1. Inflight tape calibration log for determining calibration voltage
levels.
2. RS-18 thermal scanner inflight log for sensor configuration and
time correlation.
3. Actual flight line map for determining approximate grid orientation.
4. Aircraft flight log for determining heading, estimated ground speed,
anomalies, altitude, cloud cover and haze conditions.
5. Photographic equipment log for verifying flight time to refine grid
orientation.
6. Analog film for determining locations and relative values of surface
temperature and to determine areas of isothermal water for atmospheric
correction techniques.
7. Analog to digital conversion report to provide tape number, time
slices, and data quality.
8. Surface measurements data to verify temperature values and meteo-
rological conditions.
The Univac 1108 computer and associated "off line" devices are used in
the next stage of data processing. The sensor data was previously pro-
cessed through A/D conversion routines and recorded on digital magnetic
tape. The digital tapes are stored in the computer area tape library and
respond to data processing commands. A digital program tape is also stored
in the tape library and responds to data processing commands. The program
contained on this tape is described in Reference 2. This program, using
the digital data tape containing raw data counts, averages the appropriate
number of scan lines, applies the blackbody calibration curves and also
applies the selected atmospheric correction.
The specific order of data processing is contained in the following
steps:
1. The data user defines the new type of final product desired. The
definition of the final product includes the number of copies, format,
content, etc. A typical data product is defined as follows:
a. Produce digital data product of the survey area formatted as
follows:
1) Shades of grey scale to change in .5°C increments over water.
V - 49
-------
2) Imagery to be. overlayed with numerical temperatures at
standard intervals (seven values across picture)
3) Apply the best possible atmospheric correction to the data.
2. A sample portion of the data is processed to determine overall
quality.
3. After the necessary data quality runs are made a determination of
the best method to apply atmospheric and instrument corrections are
made. Several options are available to the processor as follows
(refer to References 2, 3, 4 and 6 for additional details on these
options):
a. If an area of isothermal water is available the computer can be
instructed to determine a corrective factor by inputting time and
surface temperature. This technique corrects for both atmospheric
conditions and instrument error.
b. Atmospheric corrections based upon the RS-18 scan mechanics can
be inputted to the computer. This can be used both with or without
a known surface temperature. Instrument error can also be removed
by this technique.
c. Corrections based on radiosonde sounding can be applied to the
data which does not correct for instrument error. This technique
is still under development and is not recommended as a standard
procedure. However, the option of using it is available in the
software.
d. Data can also be corrected by an arbitrary mathematical method
which would be used as a last resort and is also described in Ref-
erence 2.
4. After testing the data and determining a method of applying atmos-
pheric and/or instrument correction all the data is processed to obtain
data tapes containing time and absolute temperature in counts.
Digital derived imagery overlayed with numerical temperature is produced
by processing the tape obtained in Step 4 above, through the Univac 1108
computer and an "off line" S-C 4020 plotter. The program used with the
plotter is described in Reference 2 and converts the corrected counts into
temperature units which are then used to produce the selected gray level
intervals and numberical overlay in the correct orientation. The output
from this process is 35mm film imagery and a supplementary paper output
which records the processing parameters used to achieve this film output.
The film can then be processed into paper products as shown in Figure 5,
In addition, the digital computer tape with suitable formatting may be used
for direct entry into a digital data bank. It should be noted that in the
quantitative digital product only the temperature values over water are
valid since the emission properties of the land are so variable and not well
known.
V - 50
-------
V. QUANTITATIVE DIGITAL DATA PRODUCT ACCURACY
The following discussions described approaches being considered
at ERL to define the digital product accuracy. Primary error sources
are considered to be system errors including instrument and processing
errors, atmospheric correction technique errors and geometric errors.
In addition, other sources of error not as significant as the above are
also being investigated. The results of these investigations will be
reported as completed. The present level of experience with the system
indicates that an absolute remote measurement accuracy of jK).5°C from a
light aircraft is a reasonable goal.
A. System Errors
The accuracy of a temperature determination employing thermal scanner
generated data is a function of the various processes involved in data
acquisition and subsequent data processing. What is of primary importance
is the total end-to-end system/data reduction accuracy. An approach to
determining .the accuracy consists of placing an extended area blackbody
source (with a temperature accuracy of ;f.l25°C) four feet beneath the scan
head and recording the video and sync wave forms on the actual data acqui-
sition system employed on the data collection aircraft. Data will be
recorded at three blackbody temperatures: a temperature near the set
temperature of scanner blackbody number one, a temperature midway between
blackbody temperatures numbers one and two, and a temperature near that
of blackbody number two. This data will then undergo the standard thermal
scanner data reduction routine, consisting of A to D processing and subse-
quent temperature profile printout through Univac 1108 processing. A com-
parison of the derived temperature with the true blackbody source temperature
yields- the overall end-to-end system accuracy. No atmospheric attenuation
correction is necessary in this case due to the extremely short (four feet
or less) path length.
Once the system accuracy is determined an assessment can then be made
of those processes within the system that can be considered the major
sources of error and candidates for further development effort selected.
The scanner by itself has a possible 0.87°C R.S.S. error in temperature
determination. This steins from two separate error sources. The first is
in the measurement made of the temperature of the scanner internal black-
bodies. A platinum resistance thermometer is imbedded in the center of each
of the blackbody sources. Each sensor is connected to a linearizing bridge
which feeds electronics for gating blackbody temperature information into
the video signal. The bridge outputs are accurate to better than 0.5°C,
typically 0.25°C. Neglecting further electronically induced error maximum
error of 0.5°C exists in blackbody temperature measurements.
The second source of error stems from the non-linear nature of black-
body radiation. In the data reduction process the video signal is compared
to the signal levels produced by the two blackbodies and the corresponding
V - 51
-------
video temperature is determined by assuming a linear relationship between
voltage out versus temperature. However, blackbody radiometric input
energy is not linear with temperature thus voltage out must also be non-
linear. Calculation has shown that if the blackbodies are operated with
15°C temperature (typical at ERL) and the video level falls midway between
radiometric BB #1 and BB //2, a .5°C error is introduced.
B. Geometric Errors
Positional errors or those errors that result from platform location
are difficult to define in some cases since they are a function of two
different recording scales - across flight and along flight. Scales in
direction across track are relatively constant. Scales in directions
parallel to the flight track may vary from the cross track scale in one
imagery strip due to velocity to height ratio of the aircraft. Ideally,
these scales are matched during processing, which requires exact applica-
tion of ground speeds. The scanner data is usually processed with the
ground speed as estimated during flight with no special corrections applied;
therefore, the scale of the infrared imagery may vary in the direction of
flight.
C. Atmospheric Correction and Other Errors
A significant amount of thermal scanner data has been digitally pro-
cessed at ERL using the several different methods of atmospheric correc-
tion described previously. An analysis of this data using comparison of
the several techniques and comparisons of remote and surface measurements
is underway to determine an accuracy estimate for each technique. In
addition, other sources of error such as accuracy of in situ instruments
and sea-air interface effects are also being investigated to determine
their contribution to the overall remote temperature measurement technique
error.
V - 52
-------
VI. PERSONNEL, SCHEDULE AND COST
A. Personnel
Staffing required for conducting a surface water temperature survey
depends upon the frequency of usage and turnaround time required for products.
However, assuming 3% flight line hours per week or about 175 flight line
hours per year of mission performance with no extremely tight schedule demands
for products, the following types and numbers of personnel should be able to
perform the task .with reasonable effectiveness.
Type No.
Electro-optical Engineer 1
Electronics Technician %
System computer programmer* %
Computer production
specialist 1
^Required for one year only or until software package is operational.
The elctro-optical engineer may be one with formal training in
electronics and experience in or association with the field of optics or the
reverse would be acceptable where the person had received formal training
in physics with experience in or association with electronics. Duties of the
engineer would include overall responsibility of the system, provide engi-
neering procedures for installation, checkout, modification, and calibration
of the system, fly with the system to assure proper operation, assist the
user in analyzing and interpreting the quick-look and final products.
An electronics technician with experience in solid state circuitry could
perform calibration and maintenance on the system by devoting one half (%)
work time to the system. In addition to calibration and maintenance the
technician could assist the electro-optical engineer and others with duties
as required.
The systems computer programmer's job consists of modification of exist-
ing software for adaptation to the particular computer system the user has
available for data processing. This effort may require more than one half
time for the first year but after the software package is modified, checked
out and in operation, the effort required for maintenance and modification
of the software would be practically zero. Familiarity with the sensor and
other components of the thermal scanner is required of the systems programmer.
Duties of the computer production specialist include coordination of
processing with the computer group performing the analog and digital data
reduction, displaying the data utilizing the program options and lead card
setup required by the user and coordination of all operational processing
efforts.
V - 53
-------
In summary, the equivalent of three people initially or 2% after the
software package is operational can effectively perform a temperature survey.
It should be recognized that personnel required for other functions such as
aircraft operation, photo and computer processing are not included.
B. Schedule
The Temperature Survey Schedule and Flow Diagram, Figure 2, shows that
under normal conditions a typical survey can be performed in about 15 work
days. Quick-look verification data can be produced in most cases in 1 day
or less and a qualitative product in about 3 days. Total work days required
can be reduced substantially with all supporting groups such as photo and
computer processing working on a high priority basis.
C. Cost
The basis for estimating the user cost of the data for a survey is the
flight line nautical mile. A flight line mile of data is acquired when the
overflight aircraft or data platform traverses one (1) nautical mile of a
prescribed flight course with the sensor system in the acquisition mode.
Computer processing costs for this report have been determined based on
experience gained at the JSC/Earth Resources Laboratory at MTF with the fol-
lowing rates for labor and computer CPU time:
1. Labor - $10 per hour
2. SDS930 - $97 per hour
3. Univac 1108 - $287 per hour
In addition, the CPU time to flight line time ratio was established at
three to one for RS-18 thermal scanner data processing on the Univac 1108
computer and four to one on the SDS930 for a speed of ISOmph and an altitude
of 10,000 ft. (Lower altitudes would increase cost somewhat because of the
increased number of scan lines to be processed.) Using the above rates the
costs for SDS930 processing is $2.60 per mile and Univac 1108 processing
is $5.74 per mile. Also the manhours associated with processing a 30 min,
or 75 mile mission for the SDS930 and Univac 1108 are 20 and 30 manhours,
respectively. This is illustrated in the following chart.
RS-18 Thermal Scanner Data Processing Costs ($) per Flight Line Mile
(Nautical)
Computer Labor Total Product
SDS930 Univac 1108
$2.60 $5.74 $6.70 $15.04 1) Grey Level Scales w/grid
2,00 2) Imagery (analog film)
0.50 3) Film positive photo (color)
V - 54
-------
At an altitude of 10,000 ft. the ground scan width is four nautical
miles. Therefore, the data processing cost would be $4.28 per square
nautical mile.
To obtain total cost of the survey per square nautical mile would
require the addition of such costs as salaries, equipment and rentals for
the other survey phases like calibration, maintenance and surface/flight
data acquisition. Costs associated with many of these other aspects are
discussed in Ref. 7.
V - 55
-------
VII. CONCLUDING COMMENTS
Techniques and procedures for quantitative surface water temperature
surveys described in this document have been developed using the Texas
Instrument RS-18 thermal scanner. It should be noted that other scanners
are available on the market, and that these procedures with minor modifi-
cations should be applicable for use with these sensors. A similar com-
ment is also applicable to the major processing equipments involved i.e.,
the UNIVAC 1108 and the SDS930 computers. In addition, the procedures
are applicable to higher altitude aircraft and satellite acquired data
provided the proper modifications are made to the atmospheric correction
procedures.
As more experience is gained in the1use of the scanner equipment,
processing techniques and in the applications of the products, further
refinements will be made in the procedures to reduce cost and improve
schedule. Some of these refinements that are in work include: 1) Software
simplification to minimize computer capacity and processing requirements;
2) The development of alternate display formats utilizing simpler proce-
dures; 3) Providing a system end-to-end accuracy/calibration procedure to
be performed as a routine survey activity.
In conclusion, a procedure is presented for the conduct of water tem-
perature surveys utilizing a commercially available thermal scanner, leased
light aircraft and data processing equipments and provides qualitative
and quantitative display products as well as digital computer tapes for
entry into a data bank.
V - 57
-------
REFERENCES
1. Mississippi Sound Remote Sensing Study, NASA, Earth Resources
Laboratory Report No. 048, April, 1973
2. Daughtrey, K.R,: Techniques and procedures for quantitative water
surface temperature surveys using airborne sensors. NASA, Earth
Resources Laboratory Report No. 080, August, 1973.
3. Boudreau, R.D., 1972a: A radiation model for calculating atmospheric
corrections to remotely sensed infrared measurements. NASA, Earth
Resources Laboratory Report No. 014, Mississippi Test Facility, 71 pp.
4. Boudreau, R.D., 1972b: Correcting airborne scanning infrared radio-
meter measurements for atmospheric effects. NASA, Earth Resources
Laboratory Report No. 029, Mississippi Test Facility, 35pp.
5. Operation and Maintenance Manual for Airborne Infrared Scanning
Radiometer, Document No. HB41-EG71, Texas Instruments, Inc.,
December 31, 1971.
6. Pressman, E.A., E.P. Elliott, R.J. Holyer, 1972: Personal communi-
cation, Lockheed Electronics Company, Inc., Mississippi Test Facility,
MS
7. Rhodes, O.L., E.F. Zetka: Methods, problems and costs associated
with outfitting light aircraft for remote sensing applications.
NASA, Earth Resources Laboratory Report No. 076, July 4, 1973.
-------
REMOTE SENSOR IMAGERY ANALYSIS FOR ENVIRONMENTAL IMPACT ASSESSMENT
by
C. P. Weatherspoon, J. N. Rinker, R. E. Frost, and T. E. Eastler
Photographic Interpretation Research Division
Geographic Sciences Laboratory
TJ. S. Army Engineer Topographic Laboratories
Fort Belvoir, Virginia 22060
ABSTRACT
A conceptual framework is presented for the application of remote sensing to
problems confronted by the Environmental Protection Agency. Essentially the
framework provides for the derivation of basic information necessary for an
understanding of complex environmental relationships, followed by the appli-
cation of this information to problems of immediate and long-range concern.
Both- basic and applied information is generated by an interactive interdis-
ciplinary team utilizing remote sensing as the focal point of its study. EPA
problem areas in which remote sensor imagery analysis can assist include
location and characterization of pollution sources, assessment of impact,
projection of trends and potential problem areas, determination of appropriate
control and preventative measures, and acquisition of necessary base line and
periodic data.
INTRODUCTION
The problems of concern to the Environmental Protection Agency are highly com-
plex, and for their solution often require an understanding of numerous inter-
acting facets of the environment. The study of these problems—how they arose,
how they can be solved, how they can be prevented—lies in the province of en-
vironmental analysis. Among the greatest potential aids to such study are the
tools and methods of remote sensing. Furthermore, an adequate unravelling of
environmental complexities is best accomplished through interactive study by a
diverse team of investigators, each of whom is highly qualified in a different
physical, biological, cultural/social, or engineering discipline. It follows,
then, that a most effective approach" to environmental analysis is via the study
of remote sensor imagery by such an interdisciplinary team.
It seems appropriate in a meeting such as this, in which new sensors and
techniques are being discussed, to include a reminder that existing remote
sensing technology and methods are available to address many of the problems
with which the EPA is confronted. The purpose of this paper is to present a
conceptual framework for the application of remote sensing to these problems.
This framework provides fundamentally for the derivation of basic information
essential to the understanding of environmental interactions and probable re-
actions to various stresses. This key function was mentioned earlier, and we
shall refer to it here as the "interdisciplinary environmental analysis." The
broad data base thus produced has application both to immediate problems and to
long range study and planning. In both the basic and applied phases, the frame-
work can utilize both the old and the new. Conventional, off-the-shelf imagery
and methodology are useful, as well as new sensors and techniques; well-
established knowledge can be exploited, in addition to current research into
the complexities of the causes and effects of environmental stresses. In any
-------
case, however, we feel that the emphasis in the environmental analysis should
be more on the ordering and exploitation of the intellectual processes of the
group members, both individually and as a team, than on remote sensing hardware
or systems.
The benefits accruing from interdisciplinary input to environmental problems
are widely acknowledged. The research program of EPA itself is strongly inter-
disciplinary in nature. The approach described here is not a departure from
this concept, but rather a tool, a format, a focal point for amplifying the
effectiveness of an interdisciplinary team. Remote sensor imagery is virtually
unequalled in its capability for giving a unified representation of an environ-
ment, free from the arbitrary compartmentation we tend to impose upon it. If
we add to this a carefully chosen, skilled team, functioning in a truly inter-
active, synergistic mode, and utilizing a proven framework for ordering their
thought processes, a great deal of reliable information can be generated
quickly and economically. Of prime importance here is the realization that
imagery gives us nothing: we must extract information from the imagery, as a
function of our knowledge and core of experience.
As indicated earlier, this paper presents a conceptual framework; One reason
for considering it so is that, while the cornerstone—the interdisciplinary en-
vironmental analysis—has been tested and proven repeatedly, many of the
specific applications to EPA problems are less well grounded. To date, our own
experience in these applications has been limited. We have been involved,
however, in a wide variety of other imagery analysis applications, including
several associated with natural and man-induced environmental stresses. Based
on this experience, we believe this approach to be basically sound, and look
forward to its being tested on EPA problems.
The interdisciplinary environmental analysis procedure will be summarized,
followed by a brief presentation of some of the ways we believe remote sensing
can be used to address problems of concern to EPA.
THE ENVIRONMENTAL ANALYSIS
Most remote sensing imagery interpretation procedures can be assigned to one of
two broad categories. These categories are perhaps better thought of as ex-
tremes of a continuum, for a line that precisely separates the two is difficult
to distinguish. At one end.are those techniques associated with direct detection
and identification of discrete objects. At the other end are the methods that
are best described as analytical or interpretive in that they rely on inductive
and deductive evaluation of the various environmental components in terms of the
knowledge, field experience and academic backgrounds of the individuals con-
cerned. These require a team effort in the true sense of the word.
In the context of EPA's problems, if the sole objective of a study is the
detection and/or monitoring of some forms of pollution, and if the technology
for the task is well developed, it would be a waste of energy and time to go
through the laborious analytical procedure. On the other hand, if the study is
more complex, involving such matters as assessment of short- and long-term im-
pact of certain stresses, projection of trends, and recommendations for correc-
tive or preventive measures, then the detection-oriented task will be inadequate.
V - 60
-------
The interpretive procedure, or environmental analysis, relies on associating
numerous threads of converging evidence. It considers any and all information
that can be obtained: all pertinent forms of remote sensing data, ground pho-
tography, field notes, maps, texts, reports, current research findings, as well
as the education and experience of the individuals involved. It is well to
remember that the imagery is not so much the source of information as it is the
focal point of attention for the knowledge, experience, and judgment possessed
by the team members. It is a well worn but true example that a person cannot
extract much, meaningful geological information from a photo if he knows nothing
about geology.
Man tends to divide the landscape or environmental elements into neat packages—
e.g., polar regions, tropics, etc.—which, like his educational units, are useful
but nevertheless somewhat arbitrary and often artificial. In fact, of course,
there is a transition from the polar regions to the tropical areas—it is not a
step function—and there is an equivalent transition in the elements of pattern
of the landscape as well as in the adaptive patterns of man and other biological
systems to these climatic landform complexes. Within any area, the delineation,
texture, chromaticity and configuration of the landscape elements form a char-
acteristic arrangement that is unique to that place at that time. Various
physical, biological, and cultural forces have interacted with each landscape
element in a given manner and in a given sequence. The complexity of detail
within this image represents the scene at one brief instant in the continual in-
terplay between the forces of nature and the materials of nature. To the extent
that these relations and complexities are understood, a team not only can deter-
mine the present, but also can deduce much of the past and predict future trends
in response to a given stress. Hence the rationale for our procedure of imagery
interpretation.
The diagram in Figure 1 is an organizational chart that shows the relationships
of the various phases of the analysis. Again, this is a conceptual outline
rather than a rigid operational one. People will find their own ways of work-
ing at a given task and, during an actual analysis, it might be difficult to
identify each of these steps except for the beginning and the end. It is beyond
the scope of this paper to give a detailed description of the entire procedure.
A few comments may be appropriate, however.
The type of imagery selected for the study, of course, depends on its purpose.
In many cases, however, we find that the "horse blanket approach"—i.e., using
standard off-the-shelf panchromatic aerial photography—provides most of the
information required in the basic environmental analysis. Much of the needed
photography is already in existence in federal, state, or commercial files, and
may be purchased for a nominal sum. A case in which existing aerial photography
was used to conduct an interdisciplinary environmental analysis on a large
scale quickly and economically was the involvement of our Division in Project
Sanguine. The purpose of our contribution to Sanguine was to provide the Navy
with environmental information about northern Wisconsin and bordering Upper
Michigan useful for developing engineering plans, cost estimates, and projec-
tions of environmental impact for an envisioned communications system. Several
terrain components—landforms and soils, surface drainage, bogs and other
organic wetlands, and land use and cultural involvement—were assessed and
mapped over an area of 22,000 square miles in a time period of fourteen months—
late 1968 through early 1970. Total cost was $250,000, or less than $12 per
square mile.
V - 61
-------
AIR PHOTOS
OR IMAGERY
AT LEAST TWO
SETS PLUS
MDSAICS
ANALYSIS
TEAM
COMPOSITION
DEPENDENT
ON TASK
AUXILIARY
INFORMATION
MAPS, TEXTS,
LOGS, OTHER
REMOTE SENSOR
DATA
II
GROUP STUDY - REGIONAL BASIS
REVIEW AVAILABLE MATERIAL - EVALUATE AND
DISCUSS MAJOR OR REGIONAL PATTERN ELEMENTS
AND LAY PLANS FOR INDIVIDUAL STUDY
III
INDIVIDUAL STUDY - REGIONAL AND SPECIFIC
INDIVIDUAL AND SMALL GROUPS - EVALUATION
OF STEREO IMAGERY, PREPARATION OF OVERLAYS, ETC.
GEOL i GEOG I ENG I BIOL I HEALTH I HYDROL I CULTURAL I ETC.
IV
GROUP DISCUSSION AND EVALUATION
INTERCOMPARISON OF ALL INDIVIDUAL LINES
OF REASONING - DRAWING FURTHER
_, INFERENCES AND CONCLUSIONS
REPORT PREPARATION
DATA BASE
VI
VII
APPLICATION
TO SPECIFIC
PROBLEMS
LONG RANGE
STUDY
AND PLANNING
FIELD CHECK
Figure 1.
analysis.
A simplified presentation of the major steps in an environmental
V - 62
-------
The use of existing photography to provide not only synoptic coverage, as in
Project Sanguine, but also sequential coverage of an area can be particularly
important for EPA studies seeking to establish patterns and trends of population
growth, industrial expansion, and agricultural practices, to name a few. To the
extent that additional imagery types are required, such a primary analysis serves
as an excellent focal point for planning for the most expeditious acquisition
and use of these other types. We feel this is preferable to the procedure of
initial "saturation bombing" of the area with many sensors, which not only is
expensive, but also can set one up for a painful case of data indigestion. This
certainly is not said to detract from the capabilities of the more sophisticated
sensors, but simply to urge their wise and discriminate use.
Regardless of the imagery type used, the study should start with an analysis of
the smallest scale available. EKTS images often serve this function very well.
Photo index sheets—usually larger in scale than ERTS images—are also useful
for this step. Next, individual photographs, perhaps at a scale of 1:15,000 to
1:20,000, are laid out to form a large mosaic of the area. On this, one can
delineate gross pattern features and note their interrelationships as well as
their mutual influence on particular areas of interest. Then comes a detailed
study of each set of stereo pairs. Pattern elements are visible on the very
small scale imagery that are not apparent at the larger scales, and the reverse
is also true.
Since the purpose of an environmental analysis is to obtain as much information
as possible about any given environment by interpretation of remote sensor data,
it follows that a team composed of members with different professional experi-
ence in the fields of science and engineering will accomplish more than a group
whose backgrounds , training, and viewpoints are all similar. How many people
and which backgrounds should be represented are a function of the assigned task
and of the resources and people available. Our contribution to Project Sanguine,
for example, utilized twenty professional people, working in four interdisci-
plinary teams, and representing the disciplines of geology, geophysics, hydrol-
ogy, soils, mathematics, physical science, botany, forestry, agronomy, ecology,
economics, geography, and civil engineering. For many studies a team of five
to eight is enough to provide sufficient breadth of background knowledge and
yet small enough to be manageable as a single group. The team probably should
include one or more members with expertise in the specific application areas to
be addressed after the basic environmental analysis is completed. For some
pollution studies, an industrial chemist, a fishery biologist, or an environ-
mental laywer might be considered. It should be pointed out again that the
emphasis in the analysis should be on the people—their backgrounds, training,
interest, ability to function effectively in a group, and capacity for creative
thinking—rather than on hardware.
Appropriate auxiliary data should be exploited to the full extent permitted by
the time and resources available. Remote sensing is not an end in itself, but
a very useful tool to be utilized in concert with other tools in seeking an im-
proved-understanding of the environment and a solution to specific problems.
In the case of EPA's problems, we are dealing with a host of complex interac-
tions, many of which are poorly understood. It is obvious that remote sensing
will not give us all the answers. Accordingly, past and present research work
should be employed freely. The remote sensing team analysis can comprise a
highly effective focal point for evaluating, integrating, extrapolating, and
applying these research findings in the context of the specific environment and
problems being addressed.
V - 63
-------
The group discussions indicated in Parts II and IV of Figure 1 are important
and provision must be made for these sessions. A group of five people working
separately (each in his own room working on photos and writing his section on
geology, cultural patterns, etc.) is not equivalent to five people around a
table with photography and stereoscopes, arguing and debating each point as it
develops. By the same token, time must be allowed for individual thought and
pondering.
The analysis is strongly interactive; it thrives on feedback. A point which
surfaces in group discussion may provide the stimulus for in-depth study by one
of the other team members. The reverse is also true. Similarly, as the analysis
progresses, it may become evident that certain types of auxiliary information,
perhaps including other remote sensing data, are now needed. A field check of
the interpretation, whenever this is possible, provides invaluable feedback, both
by improving the quality of the current work and by adding to the fund of
experience of the team members, thereby permitting them to do a better job the
next time.
Figure 2 shows an outline which has proven useful as a general guide for con-
sidering terrain elements in an environmental analysis. It provides for a
regional and local analysis, an interrelating of the various components of the
analysis, and application of the results to specific problems. Modification and
amplification of certain parts of the outline undoubtedly will be required in
.response to the different objectives of different studies. Whether this or
another outline, of none at all, is used, the important thing is that all sig-
nificant environmental parameters be considered rather than just isolated pieces
of the environment. Naturally, the degree of detail in which various aspects
are treated will be tailored to the objectives of the study. However, giving
some attention to all aspects helps to insure that potentially important factors
and interactions will not be ignored, and provides the framework into which more
detailed data, if required later, can logically and efficiently be fitted.
Having at our disposal the information derived from the environmental analysis,
we now can begin to address the more specific objectives of the study. This
may include any of a broad array of problem areas—basic ecological studies,
resource management, planning for rational development, site selection for con-
struction activities, military planning, or assessment of environmental impact,
to name a few. The preparation of environmental impact statements is a major
concern of several agencies, including the U. S. Army Corps of Engineers, of
which our own organization, the Engineer Topographic Laboratories, is a part.
The pollution-related problems for which EPA has primary responsbility, and
which we will now consider more specifically, may be considered a very large
subset of environmental impact problems. To all these problems, we believe, the
environmental analysis approach summarized here has much to contribute.
APPLICATIONS TO EPA PROBLEMS
The application categories which follow are presented in a more-or-less chrono-
logical order. The categories, as well as the chronology, are sometimes ill-
defined and overlapping. They are used here primarily to lend some degree of
order to the discussion. Strong interactions and feedback among all these cate-
gories can be anticipated.
As with the basic phase of the analysis, full use should be made of past and
current research. Furthermore, opportunities should be exploited for the
-------
OUTLINE GUIDE
FOR
ANALYSIS OF NATURAL FEATURES
USING AIRPHOTOS
I. Regional Aspects (Major Environment) - study of photo mosaic
A. Geography of Region
1. Location - physical aspects, space relationships, civil and political.
2. Economic Aspects - Land use and major apparent economic interest.
3. Transporation Aspects - All Types
B. Physiography of Region
1. Mountains
2. Hills
3. Plains
4. Es carpments
5. Basins
C. Geology of Region - Origin
1. Transported origin
a. Wind
b. Water
c. Ice
d. Gravity
2. Residual origin
a. Igneous
b. Sedimentary
c. Metamorphic
d. Combinations
D. Climate of Region
1. Arid
2. Semi-Arid
3. Sub-Humid
4. Humid
5. Tropic
6. Polar
In each, describe the physical expression
in terms of form, character, extent,
boundaries, degree of dissection. Look
for indicators of these. Describe fully -
In each, look for the indicators of
origin, movement, agent of movement and
deposition - The mechanisms responsible.
Describe fully.
Look for indicators of origin of those
deposits formed in place. Describe
fully.
Look for the indicators of climate in
erosion, vegetation and land use. Pay
careful attention to: Location and
distribution of vegetation; intensity
of erosion and redeposition; and
presence or absence of irrigation.
II. Local Aspects (Minor Features Pattern or Elements) - Stereo-study.
A. Land Forms
1. Description of physical characteristics
2. Space relationships
3. Arrangement
FIGURE 2
-------
B. Drainage Patterns
1. Non-p at te rn
2. Pattern type and Plan
C. Erosional Aspects
Bound, mark and describe fully all por-
tions of the drainage net in a watershed.
1. Wind related forms (blow-outs)
2. Water related forms (gully
characteristics)
3. Ice related forms
4. Gravity related forms
5. Chemical and thermal
6. Biological
In each, look for indicators
of mechanisms, and describe
fully. Draw sketches, showing
profiles, plan views and cross
sections of any erosional
forms.
D. Photo Tones
1. Tone arrangement
2. Pattern causes
E. Vegetation
1. General classes
a. Barren
b. Grass
c. Brush
d. Timber
F. Special Features
Describe the arrangement of tones in
terms of shades, contrasts, design, etc.
Describe in terms of location, extent,
density, slope, exposure, etc. I'f
possible describe the structural
character.
1. Unique/unusual features. Describe fully (unusual features) in
terms of space relations, physical aspects, forms. Look for
indicators of origin.
G. Cultural Aspects
1. Man's activities
a.
b.
Urban
Rural
Description of the type use and alteration
Df the landscape. Describe the intensity
af man's activities.
III'. Analysis - summations
A. Regional
1. Summation
2. Relationships
3. Interdependencies
B. Local
1. Summation
2. Relationships
3. Interdependencies
FIGURE 2 (Cont'd)
V - 66
-------
C. Local with respect to regional
IV. Interpretation of data
A. Summary
B. Evaluation, Significance
C. Interpretation of data with respect to problems
FIGURE 2 (Cont'd)
V - 67
-------
analysis team to combine imagery analysis with field activities. These points
are especially important in view of the meager state of knowledge about many
pollution problems. For similar reasons, detailed procedures for most of the
application portions of the analysis must await more practical experience in
the use of this approach.
Source Location and Characterization
Our first task probably will be that of locating significant sources of pollution
in a given area and characterizing them in terms of type and quantity of pollu-
tants . Some sources already will be known. The location of others may become
apparent after a rather cursory examination of aerial photography of the area—
e.g., most significant stationary air polluters, feedlots and other concentra-
tions of livestock, industries discharging substances which alter the color or
turbidity of a water body, and refuse dumps. Next, work can begin on a listing
and mapping of the location and nature of suspected or possible sources. In
addressing this task, one could pose a wide variety of questions to which imagery
analysis might contribute to the answers. For example, what have been the rate,
pattern, and distribution of population growth in the area (.the oise of sequential
imagery is implied)? What types of activities are the occupants engaged in—
i.e., how do they use the land? How are the land use practices distributed
spatially, and how have they changed with time? In what ways has industrial ex-
pansion occurred? What natural resources are available and how have they been
exploited? What can resource exploitation, transportation systems, location,
general appearance, and other indicators tell us about the nature of specific
industries? What is the distribution of specific crops or natural vegetation
types in relation to known uses of fertilizers and/or pesticides? Can we find
anomalous features which might be pollution-related? In some areas "natural
pollution"—e.g., saline springs or natural oil leaks—may cause significant
problems, in which case geological and hydrological information derived from the
environmental analysis could provide important clues to their localization. To
the extent that sedimentation can be considered a pollution problem, the analysis
can supply pertinent information about soil properties, topography, drainage
characteristics, and natural and man-caused soil disturbances. Considerations
such as these not only will help in determining likely trouble spots, but also
should permit the elimination of many areas and pollution types from further
consideration, at least for the present.
In this process , it may be possible to make a limited test of the validity of
some of the data, inferences, and projections derived from our analysis by com-
paring them with locations and types of already known pollution sources. Hope-
fully, this will permit an improvement in subsequent analyses and inferences.
To utilize such a test, of course, we would have to divorce ourselves initially
from explicit information about these known sources.
At this point our predictions of source locations must be refined and pin-
pointed. Possible sources must be confirmed or rejected. Furthermore, we need
rather precise data about the types and quantities of pollutants associated
with confirmed sources. Obviously these requirements demand more detailed and
specific sampling. The value of the foregoing step is that it suggests where
to concentrate these sampling efforts, and to some degree, what sensors to use.
Sampling efficiency is thereby greatly increased. Further substantial improve-
ments in sampling efficiency may result from our ability to group, or stratify,
areas which are relatively uniform with respect to likely locations and types of
pollution. The relative ease with which terrain components can be stratified on
imagery constitutes one of the most powerful assets of remote sensing.
-------
A multi-stage sample strategy may well be indicated for many types of problems.
The more generalized prediction stage can help provide a logical basis for the
effective choice, flight planning, and utilization of other remote sensing data.
Interpretation of tb_e data acquired at this sampling stage should provide further
refinements—rejection of some postulated sources, pinpointing and better char-
acterization of others. For some purposes, acquisition of multiple scales of a
particular type of imagery may be useful. For example, thermal imagery collected
at a high altitude could enable the interpreter to make a general assessment of
thermal pollution over a wide area, while lower flights could help to pinpoint,
quantify, and determine fine thermal structure for particular trouble spots.
Much recent work has gone into the development of new remote sensing systems and
the utilization of existing ones toward the end of pollution detection and char-
acterization. These range from cameras equipped with special films and filters
to airborne acoustical and ionizing radiation sensors. It is unnecessary to
elaborate on the nature and capabilities of these here, particularly since many
of them are being discussed in this meeting. Suffice it to say that they rep-
resent an impressive array of operational and potential tools with numerous
applications to EPA's problems.
The refinements made at the second sampling stage increase sampling efficiency
still more in terms of guiding locations for in situ sampling. Obviously, if
the data are to be sufficiently definitive for legal purposes and for many
scientific needs, sampling cannot stop short of the in situ measurements.
The entire process of location and identification of pollution sources will un-
doubtedly be iterative in nature. Data obtained on the ground may prompt a re-
evaluation of some portion of an earlier imagery analysis, and so on. Through-
out, auxiliary data will be used heavily.
One of the products of this phase of the work should be a map, or, preferably,
an airphoto mosaic, showing the precise location of each known source and a
coded description of its pertinent parameters. This will enjoy frequent use in
other categories of applications. It may be augmented as desired to indicate
such items as in situ monitoring sites, locations where control or corrective
measures are in effect, sites where particular adverse impacts are occurring,
and potential trouble spots.
Impact Assessment
The assessment of environmental impact from pollution is beset with a great many
unknowns. It would be either naive or deceptive to suggest that remote sensing
is the answer to these exceedingly complex problems. If we think of ultimate
impact in terms of chemical, physiological, psychological, or other "micro"
effects, then perhaps remote sensing can make few direct contributions.
Knowledge of impacts" at this level is of limited value, however, without infor-
mation about interactions at a more "macro" level. This involves determinations
of the distribution of pollution-related stresses in space and time and in re-
lation to the distribution of environmental components likely to be impacted by
these stresses. It also includes evaluations and projections of the "macro" re-
sponses of the impacted environmental components—e.g., population movements,
changes in land use or.occupational practices, eutrophication of water bodies, or
loss of vigor of vegetation. It is at this level that remote sensor imagery
analysis can make very significant contributions.
V - 69
-------
A useful first step would be the careful study of the previously prepared map
or mosaic in conjunction with overlays generated during the environmental analy-
sis. This will permit the analysis team to see and evaluate the distribution
and types of identified pollution sources with respect to watersheds and drain-
age patterns, airsheds (if this information is available), geology and physi-
ography, soil types, population distribution, land use distribution, crops and
natural vegetation, wildlife and fish habitats, etc. The clues and inferences
derived from this exercise can be further refined by careful, selective study
of the airphotos or other appropriate remote sensor imagery. As before, the
imagery study can suggest locations for remote sensor flights or ground obser-
vations to supply more definitive information at selected sites.
We need not include here a long list of possible considerations in assessing en-
vironmental impact using remote sensing techniques. Many of these will not sur-
face anyway until a specific area and its specific problems are addressed.
However, brief discussion of one class of considerations might be warranted as
an example—i.e., those associated with water-borne pollution. A detailed sur-
face drainage map or overlay is an early product of an environmental analysis.
Using this, it is rather straightforward, but important, to say whether a given
location will or will not be directly affected, in terms of surface drainage,
by contaminants from such diverse sources as industrial plants, mines, feedlots,
pesticide-treated fields, and refuse dumps. Subsurface drainage is much trick-
ier. Water in this system may make its way back to the surface at a lower ele-
vation or stop in some local water table. In either case it also is important
in terms of pollution hazards. To understand subsurface flow we need good in-
formation about geology and soils, much of which can come from the imagery
analysis. On the receiving end of the impact, we need to know about such things
as locations of drinking water sources, water-related recreational activities,
subsistence and commercial fishing, and the ecology of the affected water
bodies. Again, to varying degrees, remote sensing can contribute to the answers.
Controls and Monitoring
Based on the information previously compiled on locations, and types of sources
and their probable or certain impacts, controls and corrective measures will be
indicated. In some cases the imagery analysis may be a useful .tool even at
this step. This might especially be the case if there is a question of re-
locating a source.
At any rate, once the controls are imposed, means must be established for moni-
toring the offending site to assess changes indicative of the effectiveness of
the control measure and to gain legal evidence, if need be. This is a job pri-
marily for rn situ sensors and for some of the more specialized remote sensing
systems.
Projections of Trends and Potential Problem Areas
Environmental impact is a dynamic process, usually having both short- and long-
term components. Our ability to project potential trouble spots and future im-
pacts depends on how thoroughly we have grounded ourselves in an understanding
of fundamental environmental relationships. Many of the same considerations
used in predicting current sources and impacts will likewise be useful in
addressing future problems. One important ingredient of future impact projec-
tion is the assessment of key sociological trends. Of course this is a diffi-
cult task. However, the basic environmental analysis, including careful study
-------
of sequential imagery, should provide a sound basis for intelligent extrapola-
tion of past and present social patterns and changes and related environmental
stresses to the future.
To help in assessing and minimizing future pollution problems, it is necessary
to establish current base line data and to update these data periodically in
order to detect trends and rates of change. These base line/monitoring sites
might be placed in two general categories: (1) locations where problems already
exist, and where control measures may or may not have been applied; and (2) lo-
cations having no current pollution problems, but where likely or certain pol-
lution impacts will occur in the future. In either case, analysis of remote
sensor imagery can be of considerable benefit both in establishing optimum lo-
cations for these sites and in deriving much of the pertinent base line and
periodic data.
Preventive Measures
Insofar as possible, the best corrective measure for potential problem areas is
to prevent or minimize the problem in the first place. Technological advances
in pollution control, waste disposal, etc., should prove very important in this
regard. An additional important consideration is the location of anticipated
pollution sources with respect to environmental factors having a significant
bearing on the degree of impact. Very careful thought should precede the place-
ment of an actual or potentially important pollution source in a fragile envir-
onment or in a location where detrimental impacts could be far-reaching.
Planned refuse dumps should be located to minimize adverse health and aesthetic
impacts. In the location of dumps as well as a wide variety of other sources,
surface and subsurface drainage characteristics should be carefully considered.
Numerous other examples could be cited. All these might be placed in the gen-
eral categories of development planning and site selection. Remote sensor
imagery analysis is eminently amenable to such planning activities.
Not all significant pollution sources are chronic and thus semi-predictable.
Occasionally pollutants are unleashed catastrophically. An environmental analy-
sis of an area where the possibility of such a crisis exists, completed before-
hand, can provide a better basis for immediate assessment of potential impact
and appropriate emergency measures, should the crisis arise.
Research
Research needs are implicit in all the categories of applications mentioned
here. Remote sensing can serve as a highly effective tool in many of the re-
search programs in which EPA is and will be involved. This will have spinoff
benefits in improving the use of this tool for applications. Of course, the
need also exists for research efforts geared specifically to improving the
utility of remote sensing technology and methods in these applications.
CONCLUSIONS
The ideas presented here are not startlingly new. They are based largely on
common sense, and put together in a way which we hope will provide a useful
broad framework for the utilization of remote sensing in problems confronting
the EPA. Essentially the framework provides for the derivation of basic infor-
mation necessary for an understanding of complex environmental relationships,
using remote sensing as the focal point, followed by the application of this
V - 71
-------
information to problems of immediate and long-range concern. As straightforward
as this approach seems, it is not being used anywhere near its potential in
addressing environmental problems.
Our discussion of EPA applications in this paper has necessarily been somewhat
hypothetical. More specific and more useful procedures relevant to these appli-
cations are bound to emerge from the actual functioning of a skilled imagery
analysis team motivated to seek answers to real, urgent problems in a real en-
vironmental setting.
-------
AERIAL SPILL PREVENTION SURVEILLANCE DURING
SUB-OPTIMUM WEATHER
P. M. Maughan, R. I. Welch, and A. D. Marmelstein
Earth Satellite Corporation (EARTHSAT)
1747 Pennsylvania Avenue, N. W.
Washington, D. C. 20006
ABSTRACT
An oil and other hazardous materials spill prevention surveill-
ance system has been developed for use during sub-optimum aerial photographic
weather conditions. For sky cover and visibility conditions thought to be
representative of the nearly infinite range of weather possibilities under
which multi-band aerial photography was acquired, only a highly sensitive
color positive film provided consistently interpretable results.
Rapid access techniques were also evaluated resulting in recommen-
dations for a tactical data acquisition system for both real-and near real-
i
time information update during sub-optimum aerial photographic weather con-
ditions.
INTRODUCTION
It is widely recognized that aerial photography, flown to strict
specifications, can provide vital information quickly and economically on the
location, quality and quantity of various components of the natural environ-
ment. Welch, et al, (1972) showed that aerial photography could be used to
identify potential sources of environmental damage from accidental spills of
oil and hazardous materials, and that such information could be used in spill
prevention.
Welch demonstrated that under optimum aerial photographic weather
conditions, high altitude color infrared photographs taken at scales of 1/40,000
to 1/60,000 were useful in regional surveys for locating industrial activities
-------
that could be expected to product spills of oil and other
hazardous materials. After these areas were delineated on small scale
photographs, color aerial photography flown at larger scales of
1/5,000 to 1/10,000 could be taken of selected areas for more
detailed analysis in order to locate specific potential threats
and actual spills. Strategic information gained in this manner could
be used to prevent spills by early detection of careless practices
which could lead to undesirable releases of oil and hazardous materials
into waterways.
It is frequently the case that circumstances necessitating a
tactical approach to data gathering combine to present an update problem
at a time when weather conditions are not truly suitable for aerial
photography. For example, an extended heavy rainfall could produce
flooding and soil saturation, jeopardizing earthen revetments. Such
a protracted storm is also likely to result in residual cloud cover,
necessitating a delay in updating unless specific procedures for such
an eventuality have been previously defined.
Existing camera systems, films, filters, and processing and
interpretation techniques can be manipulated in a variety of ways to
provide acceptable photography under various weather conditions. In
investigating these methods for use in spill prevention, it was hypothesized
that even in critical situations where weather was not ideal, acceptable
aerial photography could be obtained.
This study was undertaken to provide a means for acquiring high
quality information for spill prevention under a variety of sub-optimum
aerial photographic weather conditions. Optimum conditions are defined
for this purpose to be clear skies with at least 15 miles horizontal
visibility. Thus, the term "sub-optimum weather conditions" applies to
V -74
-------
any situation resulting in sky cover (clouds) or reduced visibility
that may restrict aerial photography operations. In addition,
several rapid access techniques useful for minimizing the time
necessary for acquisition and dissemination of useful information
were investigated.
METHODS
Cloud and haze conditions under which aerial photographic
surveillance was to be attempted were chosen so as to be representative
of conditions which commonly occur over industrial areas. Weather
parameters considered were divided into two categories — clouds and
haze -- each of which offers different problems in acquisition of high
quality photography. For this study, haze is considered to describe a
general set of circumstances resulting in decreased horizontal visibility
due to smoke, dust, photo-chemical smog or other particulates, but differing
from, although frequently occurring with clouds. Further, interaction
between sun angle (angle above horizon) and haze was analyzed because the
effective path length of the illuminating source (a function of angle
above the horizon) is critical in selecting photographic parameters. In
heavy haze, the difference of a few hours in mission scheduling may allow
considerable more flexibility in photo acquisition as haze may dissipate.
Table 1 summarizes the cloud and visibility conditions chosen for
analysis. The conditions were considered to be representative increments
of weather conditions which vary continuously over a nearly infinite range.
The increments specified were utilized as data acquisition guidelines in
attempting to analyze the effects of combinations of the weather types
listed on interpretability of ground features.
V - 75
-------
TABLE'1
CLOUD AND VISIBILITY CONDITIONS UNDER WHICH PHOTOGRAPHS WERE ACQUIRED FOR THIS STUDY
Sky Cover (Amount In tenths)
Overcast (10)
Broken (5-9)
Scattered (1-4)
Cloud Base (Height in feet)
,000
X
X
X
5,000
X
X
X
10,000
X
X
X
20,000+
X
X
X
Haze Type (visibility in Miles)
Medium (2-4)
Heavy (<2)
Sun Angle (Degrees above horizon)
20° 40° 60°
XXX
X
V - 76
-------
Previous work by Welch, et. al. (1972), was used to specify three
film-filter combinations for testing: (color Kodak film type SO-397 or
equivalent) with a Wratten 1A filter; color infrared (Kodak Aerochrome Infrared,
film type 2443 or equivalent) with a Wratten 12 filter; and panchromatic
(Kodak Plus X film type 2402 and Panatomic-X film type 3400 or equivalent)
with no filter. A three camera vertically mounted array of 70mm Hasselblad
500EL cameras was used as the primary data acquisition system.
The requirements of detailed photo interpretation as well as cloud
ceilings imposed by sub-optimum weather conditions necessitated the acquisition
of photos at scales 1:5,000 or greater. Flight lines were located over three
test areas in the San Francisco Bay region where sufficient aerial photography
and ground truth data were available from previous work to eliminate the need
for detailed surface observations.
In the performance of the rapid access test, it was assumed that data
similar to that obtained during the sub-optimum weather condition tests were
needed quickly. Rapid film processing techniques for both color and black-and-
white were investigated.
Detailed interpretation was performed independently oy tnree trained
photo interpreters to determine the merits of each film-filter combination under
the weather conditions encountered for this test. Eight classes of features
were examined: counting of storage tanks; leaks or seepage; oil on water;
condition of dikes; trash and debris; effluents; water quality; and tracing
pipelines (Figures 1, 2, and 3). Ratings were recorded for the photographic
quality of the imagery and the interpretability of the preselected features.
_ 77
-------
FIGURE 1
Vertical aerial photo of a steel
salvage facility at Richmond, California.
Careless practices in handling waste
materials along shoreline are evident
in this panchrometic photo taken from
a 1,300 foot flight altitude under a
5,000 foot overcast cloud base. Note
absence of shadows and low relative
contrast typical of photos taken under
overcast sky conditions. (Plus-X 70mm,
no filter, Hasselblad 500 EL camera,
1/500 f 2.8}
V - 78
-------
V - 79
-------
FIGURE 2
Waste water and oil storage pond from
ship bilge pumping operations at Richmond,
California. San Francisco Bay Shoreline
is at lower right of photo. Note reflec-
tion of photographic aircraft on oil
slick at photo nadir. This situation
verifies that oil slicks reflect the sky
above. At this latitude (40°N) the sun
does not reach zenith, therefore, the
image is not the shadow of the aircraft.
It is possible to detect oil slicks on
an overcas^ day with relatively greater
ease than on a clear day because of the
reflectance of the clouds by the oil slick.
(Aircraft altitude 750 feet, overcast cloud
base at 800 feet, Plus-X 70mm, no filter,
Hasselblad 500 EL camera, 1/500 f 4).
V - 80
-------
V - 81
-------
FIGURE 3
Petroleum refinery at Avon, California.
Detailed photo interpretation of refinery
facilities that might be a spill threat
to a nearby waterway can be performed
on this image taken under an overcast sky.
Stereoscopic interpretation of this image
permits evaluating the integrity of
protective facilities - levees, barricades
and revetments - that are associated with
refinery storage tanks and pipelines.
(Aircraft altitude 1,300 feet, overcast
cloud base at 5,000 feet, Plus X 700mm no.
filter, Hasselblad 500EL camera, 1/500
f 2.8).
V - 82
-------
V - 83
-------
DISCUSSION AND CONCLUSIONS
In comparison to the other film/filter combinations tested, only positive
color film (exposed through a Wratten 1A filter) provided adequate image quality
regardless of time of coverage, weather conditions, or reasonable deviation
from optimum exposure settings. In fact, Kodak SO-397 aerial color film
apparently can be utilized under any daylight conditions where an aerial photo-
graphic mission can be safely attempted. The wide exposure latitude of this
film combined with comprehensive flight planning adds the necessary tactical
dimension needed to maintain a dynamic information file based on the strategic
surveillance system specified in Welch, et al. (1972). Color infrared and
black-and-white panchromatic films that have specific surveillance applications in
clear wfcather exhibited shortcomings which warrent recommendation against their
use under less than optimum weather.
Constraints on both the photographic system and on aircraft operations are
more severe under sub-optimum weather conditions. The pilot must consider
ceiling height and horizontal visibility as well as aircraft speed in order to
successfully obtain low altitude photography. As altitude decreases image
motion increases unless there is a compensating decrease in aircraft speed.
Aircraft safety must be a major consideration in mission planning, its impact
depending to a limited extent on the urgency of the mission.
Careful planning by the photographer is also required. To insure
correctly exposed film and maximum possible data content, his planning should be
based on a reliable source of weather information. Clouds, suspended particulates
precipitation, shadows and non-uniform illumination all require careful
consideration, along with the usual factors of target characteristics and
product requirements.
V - 84
-------
The importance of weather to the safety and performance of the photographic
mission necessitates a source of timely, accurate weather information. Both
forecasted and observed conditions for the target vicinity as well as the
routes of ingress and egress should be available. FAA flight service facilities
along the flight path provide a good source of enroute weather. For weather
over the target the ideal source is an on-site observer in radio contact with
the aircraft. Comprehensive weather information will help both pilot and
photographer to perform their respective jobs effectively.
The preferred surveillance system for use under sub-optimum weather
conditions is a single-engine, high-wing monoplane carrying in vertical mode
an aerial camera of format size suitable to the desired altitude, low speed
operation necessitated by operation under low clouds. The high-wing configuration
is preferred because it simplified the problem of visually aligning the
aircraft on a perdetermined flight line. An acquisition scale of 1:5,000 provides
adequate detail provided a high quality camera system is chosen.
In situations requiring real-time or near real-time assessment of
transient conditions, such as natural disasters or catastrophic spills a
properly employed rapid access photographic system can meet most information
requirements which could be addressed in a routine situation by aerial
photography. A rapid access operation requires coordination in all phases, but
can significantly reduce the elapsed time between exposure and information
dissemination. Major time reductions can be accomplished by collating the film
processing, interpretation, and dissemination facilities and if possible, by
overlapping processing and transport functions.
v - 85
-------
REFERENCES
Welch, R.I., Marmelstein, A.D., and Maughan, P.M., 1972. A Feasibility
Demonstration of An Aerial Surveillance Spill Prevention System, Water
Pollution Control Research Series, 15080H01 01/72, U.S. Environmental
Protection Agency, Washington, D. C.
ACKNOWLEDGEMENT
This work was sponsored by the Environmental Protection Agency
contract Number 68-01-1091. The authors wish to acknowledge the support
and guidance of Mr. J. Riley, EPA Project Officer. Other EarthSat
personnel who contributed to this project included Dr. R. Colwell and
Messers. R. Temple, and S. Daus.
V - 86
-------
SESSION VI
OIL AND HAZARDOUS MATERIALS SENSORS
CHAIRMAN
MR. DONALD R. JONES
DIVISION OF OIL & HAZARDOUS MATERIALS, OAWP
-------
Title:
Single Wavelength Fluorescence Excitation for On-Site
Oil Spill Identification
Abstract:
This communication discusses the use of improved fluores-
cent techniques to field-classify and fingerprint oils by employ-
ing a single, fixed, excitation wavelength (254 nm) to generate
characteristic oil fluorescence spectra. The effects of sample
preparation and weathering of the spilled oil on the fluorescent
fingerprint are presented. One of several oil spill incidents
to which this approach has been successfully applied is discussed
in detail. In light of these findings fluorescent spectroscopic
techniques are ideally suited for rapid field identifications of
spilled oils.
Disclaimer:
The contents of this report represent the findings and
views of the author who is responsible for the facts and
accuracy of the data presented. The paper does not necessarily
reflect the official views or policy of the U. S. Coast Guard.
Author:
J. Richard Jadamec
VI - 1
-------
INTRODUCTION
Present oil spill investigation procedures involve the collection
and transmittal of oil spill and suspect source samples to a laboratory
for analysis. This procedure results in a considerable time delay in
identifying the spill source. This could be avoided if field personnel
had the capability to classify or identify oils directly. Identifica-
tion of the source responsible for an oil spill can be accomplished by
either quantitative or qualitative methods. A qualitative method is
required to develop a field capability; a method in which no separation
or discrete determination of oil components is required. This approach
is known as a fingerprint method, wherein the oil is analyzed by a suita-
ble technique to obtain a characteristic fingerprint or signature with-
out any chemical or physical separations being employed other than freeing
the oil from water or other debris.
1 o
Previous investigations by Coakley , Fantasia, et.al. , Thurston,
o s 5
et.al. , Gruenfeld , McKay, et.al. , among others, have shown that
fluorescent spectroscopic techniques are directly applicable for the
detection, identification, and analysis of oils. These investigations
have demonstrated that fluorescence techniques are not only rapid,
but are very specific in their ability to fingerprint oils. The above
VI - 2
-------
mentioned studies have employed a variety of excitation wavelengths,
data processing, and experimental conditions which at first glance
are not: directly applicable for field deployable instrumentation.
This communication discusses the use of fluorescence techniques to
field classify and perhaps fingerprint oils by employing a single,
fixed, excitation wavelength, (254 nm) to generate characteristic
fluorescence spectra. Previous investigators, Parker, et.al. and
Levy' have shown that oils have a strong absorption in the short wave-
length ultraviolet region. Parker, et.al." observed that oils had
absorption spectra similar in response to their excitation spectra,
with the maximum absorption being at 238 nm and subsidiary maxima at
253 to 263 nm and 286 nm. These same areas were verified by Levy^
in a recent report employing ultraviolet absorption techniques to
identify petroleum oils. Initial spot screening of oil samples, by
ultraviolet absorption, obtained for this study confirmed the 238 nm
maximum absorption and also the presence of a secondary 255 nm abosrption
region for all oils screened. Little or no absorption was observed
at 286 nm or longer wavelength regions for some oils. Realizing that
a fluorescence signal is preceded by the absorption of radiant energy,
the common, but not excessive, absorption in the 254 nm region for all
oils was selected as a possible single excitation level for all
petroleum samples.
VT - 3
-------
EXPERIMENTAL:
Apparatus: All fluorescence emission spectra were obtained using a
Perkin-Elmer MPF-3 Fluorescence Spectrophotometer, equipped
with a 150 watt xenon lamp and Model QPD-33 recorder, All
spectra were recorded at a fixed excitation and emission
spectral band width of 34 ran and 1.5 nm, respectively.
Solvents: Spectroquality (MCB) cyclohexane was used throughout this
study.
Procedure: All spectra were recorded at a constant concentration of
1/10,000, weight to weight, petroleum to cyclohexane. All
samples were stored in low actinic glass volumetrics prior
to analysis. In the recording of all fluorescence spectra
the maximum fluorescence signal for each oil studied was
normalized to 95% of the recorder scale.
Results & Discussion
Figures 1 and 2 show the fluorescent emission spectra of five
different oils when excited at 254 nm and 290 nm respectively. It is
obvious that excitation at 254 nm produces a greater variability in the
character of the fluorescent emission spectra. Table 1 summarizes the
VT - 4
-------
fluorescent data for 16 of the crude and refined petroleum oils assembled
for this study. This cross section of oil types is tabulated listing
the wavelengths of the major and minor fluorescent responses for each
oil. The minor responses are listed, from left to right in decreasing
order of amplitude. Viewing this tabulated data it appears possible to
index oils as a function of fluorescent peak responses when a constant
excitation energy is used. In cases where oils have similar peak re-
sponses as shown in Figures 3 and 4, the use of peak ratios, a technique
used earlier by Thurston, et.al.3, may be of value.
Assuming that the field party had an oil index, based on fluorescent
spectral responses and peak ratios, the question then arises as to the
reproducibility of spectra with respect to sample preparation and oil
weathering. Investigations on the dependence of fluorescence spectra
with respect to sample preparation are shown in figures 5, 6, and 7.
Initial studies were made at a constant oil to solvent ratio of 1 to
10,000. Realizing that this type of control is not feasible for direct
field application, investigations were conducted on the dependence of
fluorescent signature with respect to variations in oil/solvent concen-
tration ratios. Figures 5, 6, and 7 show the variation of estimated
oil/solvent concentration with respect to controlled concentrations.
VI - 5
-------
Figures 5 and 6 indicate that the fluorescent signature is unchanged
for a marine lube and number 4 fuel oil, respectively, and it can
readily be seen that if the amplitude of each respective maximum was
normalized to 95% the spectra would be identical. Figure 7, a marine
diesel oil, shows a slight variation in spectral response, but still
retains its basic fluorescent signature in a reverse order. Similar
results were obtained in studies using a cross section of oil types
as listed in Table 1.
Preliminary studies on the weathering of thick and thin oil films
indicates that weathered oils may be correlated directly with unweathered
oils, in many cases. As observed earlier by Coakley-"- and Thurston, et.al.3,
the fluorescent spectral fingerprint of an oil is not significantly
affected by weathering processes, but the fluorescent signal strength
decreases as a function of weathering time. Present weathering studies
in progress involve the weathering of a thick, 0.15 mm, and thin, 0.03 mm,
oil layers at one month and two week exposures to the environment,
respectively. Figures 8 and 9 show, respectively, the effects of
weathering on the fluorescent fingerprint of a number 2 fuel oil, and
marine lube oil. Variations do exist in the fluorescent fingerprint of
some oils, but it is felt that these variations will not affect the rapid
field identification of oils. Other oils, for example a number 5 and 6
fuel oil, and a Venezuela crude had no significant changes in their
fluorescent signature. As stated previously, the fluorescent signal
VI - 6
-------
strength decreases with weathering time. In this study, the decrease in
signal strength has been minimized by normalizing the maximum fluorescent
signal to 95% of recorder scale before recording the spectrum of each
oil.
Another possible field application would be the identification of
a spill source. Initially field investigation units could identify the
type of oil present and indicate prime spill source candidates. Utiliz-
ing the fluorescent fingerprint or signature of the spilled oil they
could then match it with that of a suspect source. This approach has
been used successfully in real spill investigations. Figure 10, represents
one particular spill situation. One of eight vessels was suspected of
pumping its bilges and being responsible for an oil spill incident. A
total of twenty-three samples was collected and transferred to a labora-
tory for analysis. Nineteen of these collected samples had fluorescent
fingerprint drastically different from the spill fingerprint. Three of
the four fluorescent fingerprints shown in Figure 10 are nearly identical.
One suspect fingerprint is a direct overlay of the spill fingerprint. A
second suspect fingerprint differs slightly. If the field unit had been
equipped with portable instrumentation, the identification of the spill
source could have been made directly in the field, or it could have
resulted in the transmittal of only four samples to a laboratory for
-------
detailed analysis. The identification by fluorescent fingerprinting
in the above mentioned case, was confirmed by infrared, gas chromatography,
simulated distillation gas chromatography, and atomic absorption analysis.
CONCLUSIONS
Identification and fingerprinting of oils by fluorescent techniques
directly in the field, at the site of an oil spill, appears feasible.
Utilization of a single wavelength excitation source, such as a line
filtered mercury lamp, appears sufficient to generate characteristic
fluorescence spectra for a variety of oils. Sample preparation and
short term weathering of oils do not significantly affect the fluorescence
fingerprint or signature of an oil. Investigations are currently in
progress to develop these fluorescent techniques and instrumentation for
direct field application.
vT - 8
-------
REFERENCES
1. Coakley, W. A., 1973, Conference on Prevention and Control of Oil
Spills, Washington, D. C., pp. 215-222
2. Fantasia, I. F. , Hard, T. M. and Ingrao, H. C., 1970, Report No.
DOT-TSC-USCG-71-7, Clearing House Control No. P8 203 585
3. Thurston, A. D., and Knight, R. W. , 1971, Environmental Science
and Technology, Vol. 5, No. 1, pp. 64-69
4. Gruenfeld, M. , 1973, Conference on Prevention and Control of Oil
Spills, Washington, D. C., pp. 179-193
5. McKay, J. F., and Latham, D. R., 1972, Analytical Chemistry, Vol. 44,
No. 13, pp. 2132-2137
6. Parker, C. A. and Baines, W. J. , 1960, Analyst, Vol. 85, pp. 3-8
7. Levy, E. M., 1971, Water Research, Vol.5, pp. 723-733
VI - 9
-------
TABLE 1
Wavelengths of Maximum and
API
Oil Type Gravity
No. 6 Fuel Oil
Crude Oil - Bartlet
Field
Crude Oil - Lago
Marine Lube Oil
Marine Diesel
Crude Oil - Monument
Buttee
10W-30 Motor Oil
Hydraulic Oil
No. 2 Fuel Oil
No. 4 Fuel Oil
No. 5 Fuel Oil
Crude Oil - Zuitina
Jet Fuel
Crude Oil - East
Blackburn
Gasoline
Kerosene
11.4
15.0
18.0
21.8
25.2
26.2
30.0
31.6
35.0
35.0
35.0
41.0
45.0
54.2
—
42.1
Minimum Fluorescence Responses
Major in Nanometers Decreasing Miniu
379
358
360
332
333
360
332
332
331
357
358
357
326
310
290
305
364
373
373
318
347
373
344
345
314
351
373
372
336
326
322
326
406
386
335
346
318
338
312
311
340
336
346
347
290
342
326
315
431
346
427
354
355
316
293
293
350
316
404
336
303
354
336
291
337
400
316
373
374
425
370
369
366
374
432
316
349
356
352
336
VI - 10
-------
H
I
figure 1.
300 /wr
35V?
NO. 2 FUEL OIL
NQ4FUELOIL
MAR. 6AS OIL -
MAR DIESEL OL-
MAR. LUBE OIL
EXCITATION 254nM
-------
figure 2.
L^GEND
NO 2 FUEL OIL
NO 4 FUEL OIL
MAR GAS OIL
MAR.DESELOL
MAR. LUBE OIL
M
I
1*"*
EXCITATION 290MM
-------
figure 3.
1 FGFN1H
MAR. LUBE OIL -
NEUTRAL OIL -
NAPTHENIC OIL-
MAR. GAS OIL -
x—x
.0 — 0
\
-------
Iffi*
figure 4
REFNAPTHENIC01L —
HYDRAULIC FLU ID --
GEN. MACHINE OIL -*•
NEUTRAL OIL — °-
PREM. TURBINE OIL—*•
300 NM
EXCITATION 254MH
-------
H
cn
figure 5,
LEGEND
MAR. LUBE OILCweighed)
MAR LUBE OiLfesHmated)
EXCITATION 254™
-------
7K-
-=j
M
I
figure
LEGEND
NQ4 FUEL OlLfestimateel)-
EXCITATION 254NM
-------
figure!
LEGEND
MAR.-"DIESEL OlUweighed) -
MAR. DIESEL OiLjestimaTect
EXCITATION 254NH
-------
<4
M
I
figure 8.
i.EGFND.
0 hours —
48 hours
120 hours —*
168 hours —°
216 hours—A
NO. 2 FUEL OIL
EXCITATION 254-H
-------
figure 9.
LEGFNK)
0 hours —
48 hours
120 hours —x
168 hours —o-
216 hours —*•
MARINE LUBE OIL
EXCITATION 254NM
-------
3
i
figure 10.
j£_
j FGEND
UNKNOWN OIL-
SUSPECT OIL
SUSPECT OIL-'-
ACTUAL SOURCE-*
OIL SPILL ANALYSIS
WAVELENGTH IN NANOMETERS
EXCITATION 254NM
-------
July 11, 1973
THE REMOTE DETECTION AND IDENTIFICATION OF
SURFACE OIL SPILLS
By
Herbert R. Gram
Spectrogram Corporation
A need has existed for a means to continuously monitor water
surfaces for petroleum products and provide an early warning
alarm of oil spills or contamination. Of equal importance is
the requirement of such a system to type or class identify
the detected oil. Such information is meaningful in isolating
the source of a spill.
A specific situation required an early warning system to detect
a spill or leak of No. 6 fuel oil during barge transfer oper-
ations. As many pleasure craft operated in the waters im-
mediately upstream of the transfer point, and are potential
sources of general oil pollution, an oil detection system
specific to No. 6 fuel oil was desired.
This paper will present the arguments surrounding the initial
goals and the conceptual system design of the oil detecting
buoy system. The final operational buoy system provides an
alert signal when petroleum products are detected on the sur-
face of the monitored waters, and an alarm signal when the oil
detected fulfills the fluorimetric spectrochemical criteria
for No. 6 fuel oil.
We will further briefly describe the data developed during
laboratory studies, the initial breadboard design, and the de-
sign of the final buoy system including field performance.
VI - 21
-------
THE REMOTt DETECTION AND 1 DENT I r i :./.v : ON OF
S'JRfACE 0!L SP'LL'-
Ry
Herbert P.. Gram
Spectrogram Corporation
385 State Street
North Haver., Connecticut 06473
Presented at Second Conference on Environmental Quality Sensors by
the Environmental Protection Agency held at the National Environmental
Research Center, Las Vegas, Nevada, on October 10 - 11, 1973.
VI - 22
-------
Spectrogram Corporation
October 10, 1973
The Spectrogram Corporation received an inquiry from a major New
England electric power company asking if a dockside oil detection
device could be developed that would provide early warning of an oil
spill. They explained that most of the power plants have discon-
tinued burning coal and were now burning a petroleum oil classified
as Number 6 fuel oil. As the interested power plant is located on
a river, the fuel oil is brought to the power plant in large barges
of varying length, and transferred to land holding tanks. The
transfer rate is typically 600 gallons per minute and deliveries are
made both day and night.
Specifically, the power company desired a buoy-type device that
could be deployed both fore and aft of the transferring barge. They
felt the buoy approach, rather than a fixed dockside device, would
provide greater flexibility in allowing for variations in tide,
-(ind, surface currents, barge length, and point of transfer. The
buoy could be either battery powered or powered from a dockside
console, but. no ^ock hazard should be presented to operating per-
sonnel. They desired that the buoys be small, compact and light
weight. The syc» tern' mus £ operate day and'night, year round, with
minimum maintenance and service, and under the most severe of weather
conditions. The hoi OF- of detection technique was left to Spectro-
gram, h c w..: < e r , T>.. iiMlity <.".:pan;- did again stress simplicity and re
liability.
VI - 23
-------
Spectrogram Corporation
October 10, 1973
METHOD SELECTION
The Initial step for this program was to examine the literature
regarding the various analytical methods by which petroleum oils
might be detected. The principal techniques appeared-to -be: fnfra-
red absorption, infrared reflection, visible absorption and reflec-
tance, ultraviolet absorption and reflectance, ultraviolet -excited
fluorescence, and other possibilities such as aromatic vapor detec-
tion, conductivity, etc. We restricted the method selection process
by limiting the final design to no moving parts and no contact with
the water surface. This included such items as pumps, surface
sampling mechanisms, scanning monochromators, optical choppers, etc.
Although a great deal of data was available on infrared techniques
for both the detection and the identification of petroleum oils, the
approach of IR absorption was ruled out due to the lack of windows
in the absorption spectra of water, and, therefore, the need for a
mechanical device to sample the water surface. We examined the
possibilities of IR reflectance and visible reflectance and concluded
very quickly that a number of materials found as floating skim and
debris would interfere with the reliable and positive detection of
oil. We further determined that the reflectivity was a strong
function of oil weathering and solar exposure and additionally noted
that the airborne particulate when adsorbed on the surface of an
oil slick drastically affected the reflectance characteristics.
VI - 24
-------
Spectrogram Corporation
October 10, 1973
Visible and ultraviolet absorption were discarded because of the
inability to distinguish between a thick film of a heavy grade oil
and any other opaque material or debris floating on the water sur-
face. We were amazed by the amount of natural debris such as leaves,
weeds, logs, etc. found on river waters and large clumps of seaweed,
etc. found on ocean waters. Needless to say, we did not perform
detailed measurements on the above mentioned forms of debris but
gathered enough data to demonstrate to our own satisfaction that any
method of transmission would be false triggered unless the area
under surveillance could be guaranteed free of natural occurring and/or
manmade debris.
Methods such as chromatography, surface conductivity, vapor detection,
etc. were ruled out due to the complexity of the hardware and the
requirement of continuous service such as replenishment of gases,
repeated cleaning of optics, and replacement of surface probes.
The most practical and promising approach appeared to be fluorescence
or phosphorescence. This approach does not require contact with the
water surface, does not require any moving parts, has a minimum number
of components, can be made very specific to petroleum oils, and
does not require high operating power. Much is available in the
literature referencing the field identification of petroleum oils,
in particular, as related to geological exploration. Further,
optical fluorescence i f> a tried and."proven method as a laboratory
VI - 25
-------
Spectrogram Corporation
October 10, 1973
analytical tool. Typically, when used in the laboratory, the samples
are diluted in solvents such as cyclohexane by factors ranging from
10,000 to 1 to 100,000 to 1.
The sample is then placed in a sample chamber of a commercial
spectro-fluorimeter and excited by ultraviolet radiation. Typical
excitation wavelengths are in the 200 nanometer to 500 nanometer
region with particular emphasis in the region of 260 to 380. The
fluorescent radiation emitted typically at 90° from the excitation
radiation is spectrally dissected to present a display of energy as
a function of wavelength over the range of 300 nanometers to greater
than 700 nanometers. The high dilutions cited above are required
to develop the fine spectral data specific to molecular structure
under test. Much data in the form of spectral scans and tabulated
results of scans on diluted samples is available in the literature
and can be readily generated in the laboratory. As the proposed oil
detector would not sample the surface, nor would dilutions in situ
be considered, our requirements were for spectral data on undiluted
samples of all transportable petroleum products, and further, under
a wide variety of environmental conditions. We were .not able to
uncover much more than a scanty selection of data on undiluted oil
samples and were left with no choice but to develop such data in our
own laboratory.
Referring now to Figs. 1 and 2, laboratory apparatus was assembled
VI - 26
-------
Spectrogram Corporation
October 10, 1973
and set up as shown. This allowed the direct excitation of a
variety of samples placed directly on the sample stage. To study
the effects of temperature on the oil under test, a thermoelectric
heat pump was mounted on the sample stage. Prior to the development
of special excitation lamps, a standard General Electric germicidal
lamp was used to provide excitation energy in the ultraviolet region.
Low pressure mercury vapor lamps provide an intense variety of wave-
lengths in the ultraviolet region as shown in Fig. 3. Through the
use of optical filters, specific excitation wavelengths were selected
A series of oil samples of varying type and manufactured source were
placed on the sample stage, irradiated and the fluoresced radiation
scanned and plotted as signal intensity versus wavelength. These
tests were repeated on oils at different temperatures ranging from
sub-freezing to 120°F and on samples that had a variety of environ-
mental exposure including sun, rain, and flotation on pools of fresh,
brackish and salt water. In all cases, portions of the spectral
fingerprint, and, in particular, ratios of selected spectral regions
remained substantially constant. It was determined that all oils
exhibited overlapping fluorescent signals in the blue region of
the visible spectrum when excited by short wavelength ultraviolet
radiation. By carefully correlating all the data taken, a compromise
wavelength was selected for the purpose of oil detection.
Of equal importance and substantially as a bonus for our efforts,
a specific fluorescent peak was found in the red region of the
VI 27
-------
Spectrogram Corporation
October 10, 1973
spectrum that was entirely unique to No. 6 fuel oil. We elected
to include 1n future packages two optical detection channels, one
for the blue region and one for the red region; the exact operating
wavelengths to be selected by the choice of interposing interference
filters.
Based upon the encouraging results derived during the laboratory
phase, we elected to proceed with a brassboard system that would be
capable of field deployment to verify the laboratory findings. It
was further decided that an additional optical channel would be
added to monitor the ultraviolet reflectance from the water surface.
We realized that the data derived from this channel would not be
meaningful with respect to the identification or detection of oil
but would be meaningful regarding the overall operation of the buoy
and act as a water surface monitor. We recognized that the selected
detection and identification wavelengths were well within the solar
radiation region and some means must be developed to eliminate in-
terference by the sun. Up to the point of detector saturation, all
extraneous signals included within the solar radiation band, along
with other dockside illumination sources, could be eliminated through
the use of a chopped excitation source, AC amplification and phaselock
detection. In order that we abide by our initial boundary conditions
of no moving parts, a means of electronically modulating the excita-
tion lamp was developed and incorporated within the brassboard de-
sign. Prior to the incorporation of a flotation system, the electro-
VI - 28
-------
Spectrogram Corporation
October 10, 1973
optical buoy head was suspended from the bridge directly over a
small river. A section of the river was cordoned off using oil
slick booms thus providing a protected and restricted area for in
situ evaluation of the overall approach. A small quantity of a
variety of oils was spilled on the surface of the river and the
signals as generated by the different optical channels were recorded
and clearly demonstrated the feasibility of the overall approach.
In all cases, the blue channel did provide immediate and instant
indication of even mono-molecular layers of petroleum oils on the
surface of the water. In each case, the oil was removed from the
water surface by the use of commercially available oil clean-up
pads. These field tests were repeated several times under different
weather conditions both day and night, and in all cases, all oils
did trigger the blue channel and the presence of only No. 6 oil
triggered the red channel. The field data clearly supported the
data generated in the laboratory.
We next reduced the brassboard to a practical prototype package,
and deployed these packages on a 24 hour basis in actual operating
envi ronments.
PARAMETERS SURROUNDING THE PROTOTYPE DESIGN
The intended application involved only dockside conditions and
under normal conditions, the buoys would only be deployed during
transfer operations or when a barge was in position for a transfer
VI - 29
-------
Spectrogram Corporation
October 10, 1973
operation. It was, therefore, decided that cable power from a land
based console would prove adequate and, in fact, eliminate the large
weight factor of batteries. Further, batteries would require re-
charging and other maintenance, and unless expensive batteries were
employed, the extreme low winter temperature could present additional
problems.
To minimize, if not completely eliminate, the potential shock hazard
and yet maintain sufficient voltage to provide adequate power with
minimum cable current, it was decided that positive and negative
15 Volt DC power would be brought to the buoy. Inverter-type power
supplies raised the primary power voltage to the levels required by
the multiplying phototubes and the excitation lamp. The balance
of the electronics are powered by a secondary regulated supply con-
tained within the buoy.
We proceeded with a series of field tests by deploying one prototype
buoy on a river and a second unit in Long Island Sound. The time
period for this was ideal with a late summer installation continuing
until the following spring. Several engineering problems were en-
countered such as connector corrosion and packaging material fatigue
However, the units remained operational and responded to test spills
of petroleum oils for the entire test period of nine months.
VI - SO
-------
Spectrogram Corporation
October 10, 1973
A final operating console was designed to include the various logic
circuitry and contact closures necessary to provide a meaningful
output. To provide a permanent record of background level signals,
analog recorders were employed on the optical channel outputs.
A post engineering effort was devoted to update and finalize the
prototype design which has now been installed and is in operation
at the power company's oil transfer dock. Fortunately, we can not
vouch for actual oil spill detection as no accidental spill has
taken place.
VI - 31
-------
FIG.
SYNC SIGNAL
©
3
LAMP POV/ER
SUPPLY
LOW PRESSURE
MtRCUHY LAMP
^ .x-
FILTER HOLDER t=,' !^=
' 1
' \
® ® ©
PHASE LOCK
AMPLIFIER
PHOTODETECTOR
(O) "LENS
MONOCHROfvlAlOFl
| SAMPLE STAGE 1
^SAMPLE
iTHERMOELECTRIC COOLER
©
POWER SUPPLY
FIG. I..
LABORATORY TEST SET UP
FOR REFLECTION AND
FLUORESCENCE MEASUREMENT
STRIP CHART RECORDER
SPECTROGRAM CORPORATION
j
St.
No. Haven, Conn. 06H73
APPROVED OV
BREAD BOARD "A"
vi - 3:
-------
FIG, 2.
SOURCE LAMP
FILTER
EXCITATION ENERGY
FLUORESCENCE ENERGY
[COLLECTED FLUORESCENCE ENERGY
$<£&*$$}• - -
- l
-------
ENVIRONMENTAL KEYS
FOR
OIL AND HAZARDOUS MATERIALS*
by
Charles L. Rudder and Charles J. Reinheimer
McDonnell Aircraft Company
McDonnell Douglas Corporation
St. Louis, Missouri 63166
Aerial remote sensing methods can provide effective and economic means for moni-
toring real and potential spills of oil and hazardous materials. The advantages
of airborne techniques have long been recognized by the military in reconnaissance
applications. In particular, large areas can be overflown in a short period of
time, and access can be gained into industrial sites where ground truth teams or
in situ sensors may be disallowed. With properly selected airborne sensors, data
can be collected for detailed analysis of industrial operations, the location of
outfalls, the determination of run-off patterns, identification of spill materials,
and the surveillance of numerous other environmental problems that would be diffi-
cult, if not impossible, to accomplish on the ground in a timely manner. When an
episode occurs, an aircraft can be dispatched to the scene for acquisition of data
for enforcement purposes. Often times, due to time and location constraints, aerial
surveillance is the only means to assess the extent of damage. These advantages are
made possible by the aircraft, however, the viability of remote sensing is due to
the data collection systems.
The need for sensors capable of directly detecting the many varied pollutants has
caused a great deal of emphasis in developing sensor technology for data collection.
However, field use of sensors requires exploitation of sensing technology which in-
cludes data collection, data reduction and analysis, and finally information extrac-
tion. Once sensor systems are developed, tested, and deployed, the most important
part of remote sensing is in the extraction of information from that data collected.
This step places man as part of the system to make decisions. The sensor design
engineer often ignores the associated system requirements. Typically, new sensors
are tested under controlled or at least known conditions so that that information
extraction can be treated as a test result. This paper addresses the very important
requirement for imagery interpretation keys used by interpreters to extract informa-
tion from data collected with imaging sensors. Examples of aerial imagery are dis-
cussed to emphasize the value of such keys.
* This work was sponsored in part by the U.S. Environmental Protection Agency under
Contract Numbers 68-10-0140'and 68-01-0178.
VI - 34
-------
To realize the full potential of remote sensing, the complete inventory of existing
sensors and newly developed sensors may have to be exploited. Because the types of
industries and associated pollutants are so varied, the problem of data collection
and interpretation can be quite complex. This can be explained best by first look-
ing at a simple explanation of the physics of remote sensing.
Since the sensor is airborne, its only mechanism for detecting the pollutant is the
collection and recording of electromagnetic waves scattered or emitted from that
material. These EM waves carry information about the scattering material which dis-
criminates it from the surrounding material. This process is easily recognized
when the sensor is a camera and the data is displayed as a photograph. The unskilled
interpreter can look at an aerial photograph and recognize rivers and lakes and even
oil refinery storage tanks (if he knows he is looking at an oil refinery). How does
this novitiate arrive at his conclusions? He simply looks at the geometrical shapes
defined by the recorded contracts, compares these shapes with his mental storehouse
of knowledge, and makes his decision. He really doesn't have to understand any
physics or chemistry.
To elucidate further, Figure 1 is presented as an example. The factory displayed in
the aerial photograph is a titanium plant adjacent to a major river. The whole manu-
facturing facility is shown, including the several outfalls. The unskilled interpre-
ter probably would say little about the factory except that it looks complicated and
appears to be dumping something into the river. The skilled interpreter, using
direct analysis could offer an abundance of information on conditions of the facili-
ties and other apparent information, but he too could not identify the content of
the outfalls as hazardous or non-hazardous. Both interpreters have evaluated the
geometries and contrasts to arrive at conclusions. This is the essence o£ "direct
analysis".
More information could be gleaned from Figure 1 if the interpreter had an imagery
interpretation key for titanium plants designed for environmental use. Such a tool
would describe the entire plant operation with many photographic views of each type
of facility. A chemical description of the operation also would be given so that
products and wastes would have their origins and destinations identified. With a
key the interpreter may be able to determine that at location (A) the outfall con-
tains ore-gangue; cooling basin overflow river water is being dumped at (B); at
(C), (D), (E) and (G), process water with various pollutants is discharged;.and at
VI - 3J
-------
(F) the discharge contains titanium dioxide, ferrous sulphate, and sulfuric acid.
(This information was gained by ground truth collected during an EPA sponsored
study.) To arrive at such conclusions the interpreter would employ "deductive
analysis" since no direct data is available. Deductive analysis requires tracing
the outfall content back to its source.
By using other sensors, additional data can be collected to detect pollution. For
example, thermal imagery is the result of detection of radiation in the infrared
(IR) spectrum. The IR sensor provides data outside the spectral range of photo-
graphic cameras. Therefore, the sensor complements cameras by supplying added
spectral information capability. An example of thermal infrared detection of
pollution being discharged from a steel mill is shown in Figure 2. Hot drainage
is observed in the ditches leading from the plant to the adjacent river. Not only
is the thermal pollution evident, but its presence and apparent source suggests
that it may be a hazardous material. If such data were collected concurrently with
photographic data and a properly constructed imagery interpretation key were avail-
able, then the material emitting heat could possibly be identified.
By detecting discrete spectral bands in the visible and infrared spectrum, a very
crude spectrographic analysis can be accomplished. Figure 3 shows five channels
of imagery from a multispectral scanner. The scene is a portion of a river basin
downstream from industrial activity. Comparison of contrasts with the knowledge of
spectral signatures of materials reveals evidence of residual industrial pollution
in the river basin. However, without a special interpretation key providing guidance
in making a comparative analysis, such information could not be obtained. Such a key
requires corroborative ground truth in its formulation to establish confidence in
information extraction. The illustration demonstrates that the spectral character-
istics of an area can be detected for evaluation and interpretation by direct analy-
sis when the proper key is available.
The foregoing discussion points out the strong potential value of aerial remote
sensing for the detection of oil and hazardous materials. Although sensors can
collect the necessary data, the actual detection is not accomplished until that data
is analyzed and interpreted. This task is not a simple one. Types of pollution are
quite varied and can originate in many different kinds of industries ranging from
uncomplicated municipal sewage treatment plants to complex chemical plants and oil
VI - 36
-------
refineries. Furthermore, data collected with different types of sensors requires
different rules for analysis (as illustrated by Figures 1 through 3). Therefore,
an EPA interpreter is confronted with a nearly impossible task unless he has
"guidebooks for analysis" to aid in his job. These guidebooks, or keys, are actual-
ly essential to the field use of the environmental remote sensing system.
The design of an interpretation key for environmental applications must emphasize
the need to detect conditions that create pollution. An approach to such design
would be to describe the manufacturing processes for the selected industry so that
all products, by-products, and waste materials can be traced from their origins to
their final disposition. Such information would have to be presented through pro-
cessing diagrams (i.e., flow charts), photographs of facilities from several aspect
angles, and verbal descriptions of facilities and events in a language needed for
recognition by an interpreter. If special sensors are used, then interpretation
information regarding methods of analyzing the data collected by such sensors is
required. A key containing such information would permit identification of spill
sources within the confines of the industry premises. Furthermore, deductive analy-
sis of outfalls would be more reliable when direct analysis isn't possible.
McDonnell Aircraft Company, under contract to the EPA, used the approach described
above to develop an Aerial Spill Detection Key for Petroleum Refineries. The
selection of this industry allowed a reasonable demonstration of the approach since
the refinery processes are quite complex in their photographic rendition. Figure 4
shows a flow diagram of a typical petroleum refinery similar to that which may be
found in a military key. Features that are not shown are the origin, flow, and dis-
position of waste materials. Process descriptions, simplified flow diagrams, and
aerial photographs for each phase of the operation depicted in Figure 4 with the
additional environmental data were included to provide the interpreter with the
required information.
The aerial surveillance system used to collect the data for the development of the
key consisted of a cartographic camera, a multiband camera array, and low perform-
ance commercial aircraft. These are shown in Figure 5. The cartographic camera
was chosen to provide baseline imagery for standard photogrammetric (cartographic
when needed) and interpretation analysis. The multiband cameras were used as the
special sensor to collect spectral information. This was accomplished by judicious-
ly selecting photographic film and filter combinations to record only that light
which would emphasize the spectral reflectance of petroleum waste material or
VI 37
-------
spilled oil. Comparison of the simultaneously obtained photographs from the four
cameras of the array provided a crude spectral analysis. Figure 6 shows the spec-
tral sensitivity curves of the black and white films used. Color film was also
used to obtain color discrimination for cueing of suspect pollution areas. The
direct analysis techniques for the multiband imagery provided in the resulting key
are correlated with the analysis of the baseline imagery.
To illustrate features of the key that are essential to effective imagery analysis,
several sample figures have been selected for discussion. Figure 7 shows two
views of a lubricating oil refinery area. The oblique view is used to orient the
analyst (trained or untrained). The specific processing areas are identified as
deasphalting at A, solvent extraction at B, dewaxing at C, and clay treating at D.
Each area has unique features that can be readily identified. The vertical view
of the same scene, shown in Figure 7b, permits detailed analysis of the area.
Features such as storage tank conditions and revetment inadequacies can be detected
and could signal potential spill sources. Multiband imagery of the area (not shown)
could reveal spilled material.
A particular processing area of the lubricating oil refining is described in detail
by using a simplified flow diagram (Figure 8). This figure describes the solvent
extraction process and provides a functional background for recognizing facilities.
It also identifies those materials that could be spilled at this location. Such a
diagram accompanied by textual information precedes representative imagery in the
key. A view that allows detailed direct analysis is the stereo pair. Figure 9 is
an example showing the solvent extraction process where settling tanks and furnaces
are located at positions A and B respectively. This type of display presents a three
dimensional view of objects when used with any stereo optical device. Sizing and
volumetric data can be obtained in addition to terrain features that may establish
the run-off patterns.
The products of the petroleum refinery are stored in tanks of many shapes and sizes.
The storage area may occupy up to 75 percent of the refinery area. It is in this
area that the potential .for spills is most prevalent. Normally, the analyst would
be concerned with recognizing the types of storage tanks and, by inference, the
stored material; the condition of storage tanks; and the condition and adequacy of
the revetments. It is important to note that various industries have different
practices in storing material so that the interpretive inferences for petroleum
refinery storage would be inapplicable to other industries.
VI - 38
-------
The potential pollution source in a storage tank area shown in Figure 10 is typical
of data needed in an environmental key. The leaking tank in the ground view is
readily observed at position (A) in the stereo pair. Of course, the analyst would
use a stereo viewer to provide an enlarged three dimensional image to perform his
task.
The type of storage tank when correlated with location and related features often
provides a reliable clue to the identity of the tank contents. However, shortage
of space or increased demand for specific products may dictate that tanks be used
for products other than those for which the tank was designed. Ideally, low vola-
tile crude oil is stored in fixed roof cylindrical tanks. To reduce evaporation,
the more volatile products are best stored in breather tanks, floating roof,
spherical, or spheroidal tanks. The tanks designed primarily for storing gases
under pressure are the butane, spherical, and spheroidal tanks.
The oblique photograph shown in Figure 11 shows a typical mix of tanks commonly
seen in a refinery. Examples include a group of fixed roof cylindrical tanks at
(A), spherical tanks at (B), vertical butane at (C), horizontal butane at (D), and
floating roof cylindrical tanks at (E).
The multiband imagery shown in Figure 12 is the final example of imagery used in the
petroleum refinery key. These simultaneous photographs show a waste area that is
typical of refineries. Often times such areas are located near the water treatment
ponds that lead to outfalls. These parts of the refinery compound can be suspect as
pollution sources. However, the goal of the industry is to properly treat, confine,
and eventually safely remove the waste material. *The numbers appearing beneath each
picture indicate the film and filter combination (the four digit number specifying
film type and the two digit number specifying the filter). The information provided
in the key tells the interpreter how to make contrast comparisons to perform a crude
spectral analysis for materials identification.
The Aerial Spill Detection Key that was developed was used to evaluate imagery of
several refineries taken with the same aerial camera systems. With some time spent
in familiarization, the interpreters found the key to be invaluable. Clearly, the
benefits of the aerial technique were demonstrated during the aerial surveillance
program. Ground truth teams collected correlative data which confirmed the effec-
tiveness of the imagery interpretation. Also this coordinated correlative study
VI - 39
-------
emphasized the tremendous time advantage and data collection capability inherent
in the airborne remote sensing method. At the very least, such techniques com-
plement ground surveillance by locating suspicious areas so that field investiga-
tion teams operating on the ground need not spend time observing "clean" areas.
The importance of environmental keys has been demonstrated. The successful imple-
mentation of aerial remote sensing for detecting and identifying pollution sources
through field investigations necessitates effective information extraction methods
and tools.
VI - 40
-------
FIGURE 1 TITANIUM PLANT SHOWING OUTFALLS
FIGURE 2 THERMAL INFRARED IMAGE OF STEEL
MILL AND INDUSTRIAL WASTES
VI - 41
-------
Band 1
0.5 - 0.6 Mm
Band 2
0.6- 0.7 Mm
Band 3
0.7 0.8
Band 4
0.8 1.1
Band 5
10.4 12.6
FIGURE 3 MULTISPECTRAL SCANNER IMAGERY OF A RIVER BASIN
VI - 42
-------
Crude Oil Storage
Vapor Recovery
Crude Distillation
Vacuum and Atmospheric
Gases-
Low Grade Gasoline—-
Kerosene !To Treating)
Dieiel Oil (To Treating)
CAS - Oil
Gases/Gasoline
Reforming
1
A
],[
m
,
m
A
I
A
1
-^•Fuel Gas
———— Propane
(+ Propylene)
^ Butane —
(•>• Butylene)
—Pentane -
(+ Pentylene)
I — »•
Gasoline
(To Treating)
Ethane1
*—Propane
-Butane
r Gasoline—i L_J
Polymerization/Alkylation
Treating
Deaspbalting Solvent
Extraction
,F(GURE4 FLOW OF TYPICAL REFINERY
71 - 45
-------
Zeiss RMK 1523 Camera
Hasselblad
Camera Array
; m Type 2475
Film Type 2403
Film Type 2424
•
\
Aero Commander Cessna 336
FIGURES AERIAL SURVEILLANCE SYSTEM
250 350 450 550 650 760 850 950
Wavelength Nanometers
FIGURE 6 SPECTRAL SENSITIVITY CURVES FOR
PHOTOGRAPHIC FILMS
FIGURE 7a LUBRICATING OIL REFINING
OBLIQUE VIEW
FIGURE 7b LUBRICATING OIL REFINING
VERTICAL VIEW
VI - 44
-------
Deaspnalted
O.I
Solvent
Storage
Impurities
Processor
Treating Furnace Stripper
Column
Flash Flash Stripper 0|1
Column Column
FIGURE 8 SIMPLIFIED FLOW OF SOLVENT EXTRACTION
FIGURE 9 LUBRICATING OIL REFINING, SOLVENT EXTRACTION
FIGURE 10 STORAGE, LEAKING TANKS
VI - 45
-------
FIGURE 11 STORAGE TANKS
a) 2403/99
b) 2403/35
c) 2424/99 d) 2424/35
FIGURE 12 PETROLEUM WASTE AREA
VI - 46
-------
VIDEO DETECTION OF OIL SPILLS*
John P. Millard
Ames Research Center, NASA, Moffett Field, Calif. 94035
and
Gerald F. Woolever, LCDR
USCG - Headquarters, Washington, D. C.
ABSTRACT
Three airborne television systems are being developed to evaluate techniques for oil-spill
surveillance. These include a conventional TV camera, two cameras operating in a subtractive mode,
and a field-sequential camera. False-color enhancement and wavelength and polarization filtering are
also employed. The first of a series of flight tests indicates that an appropriately filtered conventional
TV camera is a relatively inexpensive method of improving contrast between oil and water. False-
color enhancement improves the contrast, but the problem of sun glint now limits the application to
overcast days. Future effort will be aimed toward a one-camera system. Solving the sun-glint
problem and developing the field-sequential camera into an operable system offers potential for color
"flaging" oil on water.
INTRODUCTION
Airborne surveillance of oil spills offers the advantages of large-area coverage, covert detection,
and quick response to spill accidents. Unfortunately, the human eye is not a particularly good
detector of thin films of oil on water1'2 for reasons having to do with the spectral response of the
eye,3 the absorptance/reflectance characteristics of oil and water,4 and the polarization
*This work was performed under Coast Guard MIPR Z-70099-2-23146. The opinions or
assertions contained here are those of the writers and are not to be construed as official or reflecting
the views of the Commandant of the Coast Guard.
VI - 47
-------
characteristics of skylight.5 Due to these factors, an unaided observer viewing from an aircraft
often sees little contrast between oil and water. Various systems are currently being evaluated for
their ability to provide a better means of detection. Among these are the video systems discussed in
this paper.
Video systems have potential for displaying a greater contrast between oil and water than the
unaided eye would observe. These systems are real-time and can operate in the near-ultraviolet
and optical-infrared portions of the spectrum. These systems are amenable to polarization techniques;
they can produce a real-time, false-color display of a low contrast scene, and they have potential for
night surveillance. Prior studies indicate improved contrast between oil and water when sensing in the
near-UV6"12 and optical-infrared6'9'10'13 portions of the spectrum, when viewing through a polarizer
oriented to transmit the horizontal polarization component,2'5 and when subtracting signals acquired
through two orthogonally oriented polarimeters.6
Ames Research Center is developing and/or flight testing basic video systems and techniques
for oil spill surveillance by the U.S. Coast Guard. Two systems have been flight tested, and a third is
being developed. This paper describes the systems, techniques, and flight results.
SYSTEMS AND TECHNIQUES
Systems
The three basic systems being developed for flight test are illustrated schematically in Fig. 1, and
their components are specified in Table I.
System 1 is composed of a conventional TV camera and a black/white monitor (Fig. 2). The
camera has a silicon-diode-array image tube, as do all the cameras in this study, and a lens with an
automatically controlled iris. As shown in Fig. 1, the camera output was also processed through
system 2 to display a false-colored image.
System 2 is composed of two conventional TV cameras (Fig. 3) viewing the same scene through
different filters, a video processor designed to subtract one image from another and false-color the
resultant image, and a color monitor (Fig. 4). The processor will also accept a single video signal and
false-color it.
VI - 48
-------
System 3 consists of a high signal/noise field-sequential camera, a special processor to enhance
the information content of the video signal, and a color monitor modified for field-sequential
operation. A field-sequential camera has a single-image tube with a spinning filter wheel in front.
The filter wheel may contain up to three filters (polarization and/or wavelength filters). In operation,
the video signal through each filter actuates a color gun in a monitor. Thus, the color displayed on the
monitor corresponds to wavelength or polarization characteristics of the scene being viewed.
The TV cameras were mounted in the nose of a Cessna 402 (Figs. 5 and 6), downward viewing at 45°.
The cylindrical tubes in these figures are air-sampling ducts for another project. All other equipment
was mounted in racks in the aircraft cabin (as shown in Fig. 4).
Techniques
The techniques being evaluated and the rationale behind them are:
Technique 1 — Wavelength and Polarization Filtering
This technique enhances the contrast between oil and water by viewing the target scene through
selected wavelength and polarization filters. The rationale for selective filtering is that it allows sur-
face features of water to be emphasized rather than subsurface features. In the near-UV and optical-
infrared portions of the spectrum, water absorbs much of the light backscattered beneath the surface,
causing the contrast between oil and water to be determined primarily by the surface reflectances of
oil and water. Oil has a higher reflectance than water and thus appears brighter. The silicon diode
array tubes used in this study are useful because of their broad spectra response. With glass optics,
measurements were made over spectral bands from 370 to 1000 nm.
For polarization filtering, Ref. 2 reported high contrast by measuring only the horizontal
component of polarization. This high contrast is attributable to the high reflectance of liquid
surfaces for this component and the difference in reflectance between oil and water. Figure 7
illustrates the reflectance characteristics of oil and water for the two principal polarization components.
Technique 2 — Subtraction of Orthogonal Polarization Components
This technique utilizes two bore-sighted cameras, each viewing through a polarizer oriented 90°
to the other. In real time, one image is subtracted from the other and the resultant image is displayed
on a monitor. This technique was previously reported in Ref. 6, where measurements were made with
VI - 49
-------
radiometers. By means of this technique, redundant information (unpolarized radiation) is
canceled and the contrast between oil and water due to polarization differences is enhanced. A
second possible improvement in contrast is based on the fact that skylight polarization varies with
position in the sky. An airborne observer viewing an oil slick sees different portions of sky reflected
by oil and water. Thus he sees the polarization characteristics of two different portions of sky
modified by the reflectance characteristics of oil and water. In the present study, this technique was
evaluated with system 2.
Technique 3 — False-Color Enhancement
The objective of this technique is to enhance the contrast between oil and water by causing
each to appear in a distinct vivid color. To accomplish this, a video signal normally corresponding to
various gray levels is automatically "sliced" into a number of amplitude ranges. A color is assigned
to each range, and the colored image is displayed on a TV monitor. All ranges need not be colored
or displayed. If oil is brighter or darker than the surrounding water, it .will appear as a different
color. This technique has been used for several years in the laboratory (e.g., Ref. 14) to enhance
subtle features in photographic imagery, but no prior real-time application is known.
Technique 4 — Field-Sequential Processing
This technique is designed to produce high contrast between oil and water again by using color
to enhance the visual display. A specially designed, field-sequential camera is used, and the video
signals from up to three filters individually drive the color guns on a monitor. When a disparity
between oil and water exists in any one of these filter regions, the corresponding color gun will be
affected and oil will be displayed in a color different from the surrounding water. To accentuate
color differences, the system uses an automatic gain control to keep the maximum video signal at a
preset level, and a threshold detector to drive any one color gun to maximum output when its video
signal exceeds preset amplitude levels.
RESULTS AND CONCLUSIONS
The first in a series of flight tests of various video systems and techniques was conducted over a
period of several weeks and under a variety of weather conditions in the vicinity of the oil platforms
in the Santa Barbara channel. The results and conclusions are as follows:
VI - 50
-------
System 1 — Conventional TV Camera / Balck/White or False-Color
The result of most immediate importance is that significantly enhanced detection of oil films,
relative to the eye, was achieved with an appropriately filtered, conventional TV camera and a black/
white monitor. The camera contained a silicon-diode-array image tube and standard glass optics. It
was filtered with a polarizer oriented with its principal axis in the horizontal direction and a Corning
7-54 filter, which blocks out the visible and transmits the near-UV and optical-infrared portions of the
spectrum. Examples of the imagery are shown in Fig. 8; the oil appears as a bright white against a
dark water background. The video presentation was often used to direct the pilot over a slick area
when it could not be seen visually. Measurements were made in various wavelength regions, with and
without polarizers. Best contrast was consistently obtained with a polarizer oriented to transmit the
horizontal component and a Corning 7-54 filter.
Unknowns presently associated with the.single-camera system are the thicknesses and types of
oil to which it will respond. There were instances (as illustrated in the last two photographs of Fig. 8)
where a portion of a slick appeared dark. This was probably associated with a thick portion of the
slick. There were instances over other geographical areas where what appeared to be "thin" slicks
did not show up well at all. These occurrences were probably associated with both thickness and
type of oil. (Evaluation of these parameters will be conducted by the authors in the near future.)
The video signals were false-colored in real time. This technique provided an easy means for
detecting anomalies on a water surface; for instance, the system was operated so that natural water
appeared as yellow and oil appeared as blue; however, sun glint was a problem on clear days. Sun
glint caused a spike in the video signal with the result that the photograph contained a series of con-
centric colored circles (Fig. 9) emanating from the specular direction. Under overcast skies, this
problem did not exist. Work is underway to investigate techniques for filtering the solar spike out
of the video signal.
System 2 — Two-Camera System /Black/White or False-Color
The two-camera system was operated with polarizers, rotated 90° to each other, in a subtractive
mode. Oil slicks were easily detected by this technique, but generally no more easily than with a one-
camera system — with one exception: the subtractive technique minimized the effects of solar spikes
in video signals when the sun glint was reflected directly into the TV cameras. Generally, however,
VI - 51
-------
the necessity of aligning two cameras appears to outweigh the advantages of this technique. The
false-color technique also worked well with this system.
System 3 - Field Sequential
The field-sequential system was not tested because its design and manufacture were not ye't
complete.
General
The results described here are those from the first of a series of tests to be conducted. It cannot
be stated which system is optimum because of the limited testing at the present time. However, the
authors believe that future efforts should be aimed toward a one-camera system. The one-camera
conventional system described here offers a relatively inexpensive method of improving contrast
over oil spills. Solving the sun-glint problem associated with false-color and developing the field-
sequential system offer potentials for further system improvement.
VI - 52
-------
REFERENCES
1. Catoe, Clarence E., "The Applicability of Remote Sensing Techniques for Oil Slick Detection,"
Offshore Technology Conf., Paper OTC 1606, Dallas, Texas, May 1-3, 1972.
2. Millard, J. P., and Arvesen, 3. C., "Polarization: A Key to Airborne Optical Detection of OH on
Water," Science, Vol. 180, June 15, 1973, pp. 1170-1171.
3. •Electro-Optics Handbook. RCA, Commercial Engineering, Harrison, New Jersey, 1968, pp. 5.3-5.7,
4. Horvath, R., Morgan, W., and Spellicy, R., "Measurement Program for Oil-Slick Characteristics,"
U. of Michigan Rept. 2766-7-F, Final Rept, U.S. Coast Guard Contract DOT-CG-92580, Feb.
1970.
5. Millard, J. P., and Arvesen, J. C., "Effects of Skylight Polarization, Cloudiness, and View Angle
on the Detection of Oil and Water," Joint Conference on Sensing of Environmental Pollutants,
Palo Alto, Calif., Nov. 8-10, 1971, AIAA Paper 71-1075.
6. Millard, J. P., and Arvesen, J. C., "Airborne Optical Detection of Oil on Water," Applied Optics,
Vol. 11, No. 1, Jan. 1972, pp. 102-107.
7. Estes, J., Singer, L., and Fortune, P., "Potential Applications of Remote Sensing Techniques for
the Study of Marine Oil Pollution," Geoforum, Sept. 1972, pp. 69-81.
8. Chandler, P., "Oil Pollution Surveillance," Joint Conference on Sensing of Environmental
Pollutants, Palo Alto, Calif., Nov. 8-10, 1971, AIAA Paper 71-1073.
9. Catoe, C., and Ketchal, R.,, "Remote Sensing Techniques for Oil Pollution Detection,
Monitoring, and Law Enforcement," Proceedings of the Society of Photo-Optical Instrumentation
Engineers, Seminar-in-Depth; Solving Problems in Security, Surveillance, and Law Enforcement
With Optical Instrumentation, Sept. 20-21, 1972, New York City; Edited by L. M. Biberrnan,
F. A. Resell, Vol. 33, 1973, pp. 79-98.
10. Horvath, R., and Stewart, S., "Analysis of Multispectral Data From the California Oil Experiment
of 1971," Remote Sensing of Southern California Oil Pollution Experiment, Project 714104,
Pollution Control Branch, Applied Technology Division, U.S. Coast Guard Headquarters,
Washington, D. C.
11. Welch', R. J., "The Use of Color Aerial Photography in Water Resource Management," in New
Horizons in Color Aerial Photography. Joint ASP-SPSE, New York, N. Y., 1969.
VI - 53
-------
12. Webber, F. J., "Imaging Techniques for Oil Pollution Survey Pruposes," Photographic
Applications in Science, Technology, and Medicine, Vol. 6, No. 4, July 1971.
13. White, P. G., "An Ocean Color Mapping System," Internal Report, TRW Inc., 1970.
14. Jensen, R. C, "Application of Multispectral Photography to Monitoring and Evaluation of
Water Pollution," Joint Conference on Sensing of Environmental Pollutants, Palo Alto, Calif.,
Nov. 8-10, 1971, AIAA Paper 71-1095.
VI - 54
-------
TABLE L- SYSTEM COMPONENTS
Component
Description
Comments
Filter #1
Filter #2
Filter #3
Filter #4
Filter #5
Filter #6
Corning 7-54
Polaroid Polarizer HN 32
Kodak 89B
Optics Technology 166
Optics Technology 787
Optics Technology IR
absorbing glass
Transmits below 420 nm and above 670 nm.
Transmits above 680 nm
Transmits from 410 to 600 nm
Transmits from 390 to 560 nm
Absorbs above 700 nm
TV Camera #1
TV Camera #2A
TV Camera #2B
TV Camera #3
Sanyo Model VCS-3000
Sierra Scientific Minicon
Sierra Scientific Minicon
Zia Associates Inc.
Field-Sequential
2/3" silicon-diode array tube; autocontrolled iris
1" silicon-diode array tube; manual or remote
controlled iris
1" silicon-diode array tube; manual or remote
controlled iris
1" silicon-diode array tube; S/N > 200:1; auto-
controlled iris; auto-shading correction
Processor #1
Processor #2
International Imaging
Systems Differential
Video Processor
Model 4490
Zia Associates Inc.
Field-Sequential
Processor
Capable of subtracting one video signal from
another, false-coloring video signals, and process-
ing the signals so they may be presented in com-
binations of false color and black/white
Capable of eliminating unwanted background
portion of video signal and amplifying informa-
tion content. Provides selection of colors in
which signal may be presented. Provides threshold
level for producing saturated colors for signals
above a selected value.
Monitor #1
Monitor #2
Monitor #3
Tektronix Model 632
Tektronix Model 654
Sony Model PVM 1200
Black/white
Color
Color
VI - 55
-------
ff\
cc
UJ
I .
T.V. CAMERA 1
1
y*
^ J~ PROCESSOR
"!____ _j
CAMERA 1
(B/W)
MONITOR 2^
. J
(COLOR^J
SYSTEM 2
M
1
en
ct:
UJ
H
T.V. CAMERA 2A
TV. CAMERA 2B
PROCESSOR I
(SUBTRACT,
FALSE COLOR, B/W)
MONITOR 2
(COLOR)
rMONITOR f]
I |B/W) j
SYSTEM 3
CO
C£
UJ
t-
T.V. CAMERA 3
PROCESSOR 2
(MODULATION
ENHANCEMENT)
MONITOR 3
(COLOR)'
Figure 1. Schematic of baste systems.
-------
Figure 2. Conventional TV camera, with zoom lens, and monitor.
-------
LT1
CC
Figure 3. Two bore-sighted TV cameras,
-------
— VI - 59
Figure 4. Processor, monitor, and auxiliary equipment used for system 2.
-------
<
H
Figure 5. TV cameras mounted in nose of Cessna 402, shroud removed.
-------
Figure 6. Nose-shroud of Cessna 402 modified to accommodate TV cameras.
-------
o>
TO
10 r
HORIZONTAL
POLARIZATION
COMPONENTS
UJ
o
Z^-
< c
h- 0)
09
UJ
VERTICAL
POLARIZATION
COMPONENTS
o>
Q_
UJ
(T
0
BREWSTER
ANGLE FOR //
WATER //
y
30 40 50 60
ANGLE, deg
Figure 7. Reflectance of polarization components as a function of the
angle of incident light for oil and water; the index of refraction for
water is 1.34, and the index for oil is 1.57.
-------
Figure 8. Examples of imagery acquired over natural slicks in the Santa
Barbara channel; imagery acquired with a conventional TV camera con-
taining a silicon-diode-array image tube, filtered with a Corning 7-54
filter and a polarizer oriented to transmit the horizontal polarization
component.
-------
<
H
Figure 9. Sun-glint effects on video imagery.
-------
Measurements of the Distribution
and Volume of Sea-Surface Oil Spills
Using Multifrequency Microwave Radiometry
J. P. HOLLINGER AND R. A. MENNELLA
E. O, Hulburt Center for Space Research
Space Sciences Division
ABSTRACT
Multifrequency passive microwave measurements from aircraft
have been made of eight controlled marine oil spills. It was found
that over 90 percent of the oil was generally confined in a compact
region with thicknesses in excess of 1 mm and comprising less than
10 percent of the area of the visible slick. It is shown that micro-
wave radiometry offers a means to measure the distribution of oil
in sea-surface slicks and to locate the thick regions and measure
their volume on an all-weather, day-or-night, and real-time basis.
PROBLEM STATUS
An interim report on a continuing NRL problem.
AUTHORIZATION
NRL Problem G01-08
Project DOT/USCG 2-21881
Manuscript submitted April 6, 1973.
VI - 65
-------
MEASUREMENTS OF THE DISTRIBUTION AND VOLUME OF SEA-SURFACE
OIL SPILLS USING MULTIFREQUENCY MICROWAVE RADIOMETRY
There is mounting concern by the public and governmental agencies over the ever-
increasing number of marine oil spills and the more serious resulting pollution. Before
appropriate corrective action can be taken, a knowledge of the nature, thickness, areal
extent, direction, and rate of drift of the oil spill must be promptly established. This
requires a detection and measurement system capable of rapidly responding to a require-
ment of surveying large and often remote and inaccessible waters on a nearly all-weather
and day-or-night basis.
Reliable determination of oil-film thickness is of major importance. It is the film
thickness along with areal 'extent which allows the volume of the slick to be estimated.
A knowledge of the volume of oil is essential for litigation and damage claims resulting
from major oil spills, as well as for assessing the impact of the spill on marine life and
environment. A knowledge of the oil distribution and the location of those regions con-
taining the heaviest concentration of oil would enable the most effective confinement,
control, and clean-up of the oil and perhaps is most important.
Sea-surface oil spills do not spread uniformly nor without limit (1, 2). Thick regions
which contain the majority of oil are formed and are surrounded by very much thinner
and larger slicks. For example, in controlled oil spills of 200 to 630 gallons (760 to 2380
liters), which .will be described in detail later, the oil typically formed a region with a
thickness of 1 mm or more containing more than 90 percent of the oil but comprising
less than 10 percent of the area of the visible slick. The remaining oil formed a large
slick, hundreds of times thinner, surrounding the thick region.
Microwave radiometry offers a unique potential for determining oil-slick thicknesses
greater than about 0.05 mm. The apparent microwave brightness temperature is greater
in the region of an oil slick than in the adjacent unpolluted sea by an amount depending
on the slick thickness. In effect the oil film acts as a matching layer between free space
and the sea enhancing the brightness temperature of the sea. The calculated increase in
microwave brightness temperature due to an oil slick above that due to the unpolluted
sea as a function of slick thickness is shown in Fig. 1 for the three microwave frequencies
at which measurements were made. As the thickness of the oil film is increased, the ap-
parent microwave brightness temperature at first increases and then .passes through al-
ternating maxima and minima, due to the standing-wave pattern set up by the sea surface.
The maxima and minima occur at successive integral multiples of a quarter wavelength in
the oil. By using two or more frequencies, thickness ambiguities introduced by the oscil-
lations can be removed and the film thickness determined for a wide range of thicknesses.
* The reflection coefficients for a smooth dielectric material covered by a uniform dielectric film of
finite thickness, necessary to calculate the brightness temperature, are given in Ref. 3.
VI - 66
-------
HOLLINGER AND MENNELLA
SLICK THICKNESS (MM)
Fig. 1 — Calculated increase in microwave brightness temperature due to
an oil slick above that due to the unpolluted sea as a function of slick
thickness at 19.3, 31.0, and 69.8 GHz. The calculations are at 0° in-
cidence angle for a smooth sea at a temperature of 20° C and salinity
of 35 ppt covered by a uniform oil film with a complex dielectric con-
stant of 2.1 — 0.01J.
To verify the calculated behavior with film thickness and to measure the dielectric
properties of the three oil types used in the controlled ocean oil spill tests, microwave
radiometric measurements were made using a small test tank. The tank was filled with
fresh water, and then a known volume of oil was added to the surface. The incremental
increase in the oil-film thickness resulting from the addition of oil was calculated assum-
ing uniform spreading of the oil over the surface area of the tank. Measurements were
made for No. 2 fuel oil and No. 4 and No. 6 crude oils. Number 2 fuel oil spread uniformly
over the tank even for thicknesses as small as 0.1 mm. However No. 4 and No. 6 crude
oils tended to form lenses or blobs until the entire tank surface was covered. This re-
quired thicknesses in excess of about 4 mm, after which the oils apparently did spread
uniformly for small incremental increases. Therefore the results for these oils are less
accurate than those obtained for No. 2 fuel oil. The complex dielectric constant of the
oil was determined by adjusting it to obtain the best fit of the calculations to the measure-
ments. The measurements for No. 2 fuel oil and the best-fitting calculated curves are
shown in Fig. 2. The complex dielectric constant 6j — jez, determined from the measure-
ments for No. 2, No. 4, and No. 6 oils, is given in Table 1.
The measured antenna temperature, rather than brightness temperature, is given in
Fig. 2 to present the magnitudes that were actually observed during the observations.
The antenna temperature is the average of the brightness temperature over all directions,
weighted by the antenna response pattern (4). That is,
(1)
(6,
VI - 67
-------
NRL REPORT 7512
234
SLICK THICKNESS (mm!
Fig. 2 — The data are measurements at 19.3 and 69.8
GHz of the increase in antenna temperture due to No.
2 fuel oil spread over a smooth water surface in a test
tank as a function of film thickness. The curves rep-
resent the calculation which best fit the measurements.
where f (&,(?) is the normalized antenna response pattern and Tg (Q,p) is the total bright-
ness temperture in the direction Q,v and is composed of not only the radiation emitted
by the sea surface but also the downwelling sky radiation reflected by the surface as well
as the emission and attenuation of the atmosphere between the surface and the radiometer.
Atmospheric effects are usually less than 10 percent for observational frequencies away
from the water-vapor absorption line at 22.235 GHz and below about 40 GHz, and in
general approximate corrections can be applied to partially remove them.
A series of eight controlled oil spills was conducted during the period August 1971
through August 1972 in cooperation with the NASA-Wallops Island Station, the Virginia
Institute of Marine Science, and the U. S. Coast Guard to investigate the possiblity of
determining the thickness of an oil slick using passive microwave radiometry. The spills,
of from 200 to 630 gallons (760 to 2380 liters) of either No. 2 fuel oil or No. 4 or No.
6 crude oil, were performed in accordance with the guidelines established by the Environ-
mental Protection Agency for the discharge of oil for research purposes (5). All of the
spills were conducted in relatively calm sea conditions of less than 2-m swell and 10-m/s
surface winds. The oil was transported in 50-gallon (190-liter) drums to the ocean test
site, about 10 mi (16 km) east of Chesapeake Light Tower off the east coast of Virginia.
The drums were off-loaded, herded together, and emptied from small rubber boats in a
manner so as to obtain as nearly an undistrubed point release as possible.
The documentation of "ground truth" gathered included the type and volume of oil
spilled, in situ measurements of oil-slick thickness, and airborne natural and color IR
photography and thermal IR imagery, as well as the environmental parameters of sea
temperature, air temperature, relative humidity, wind speed and direction, sea state, and
general weather and cloud conditions. The oil in two spills was dyed with an oil-soluable
VI - 68
-------
HOLLINGER AND MENNELLA
Table I
Measured Complex Dielectric Constant of Oil
Oil Type
No. 2 Fuel
No. 4 Crude
No. 6 Crude
Temperature
CO
19
26
26
Complex Dielectric Constant
f = 19.3 GHz
el: 2.10 ± 0.05
+ 0.02
^:0-01-o.oi
el: 2.4 ± 0.1
e2: 0.06 ± 0.04
el : 2.6 ± 0.2
e2: 0.05 ±0.05
f = 69.8 GHz
el: 2.10 ± 0.05
+ 0.02
e2: 0.01
2 - 0.01
ej: 2.2 ± 0.1
e2: 0.07 ± 0.04
el: 2.6 ± 0.2
e2: 0.05 ± 0.05
red dye to aid in establishing the distribution of oil over the sea surface. The dye allowed
the thick regions of oil to be easily identified visibly. Figure 3 is a series of drawings
traced from color photography of the July 11, 1972, oil spill. This spill consisted of 630
gallons (2380 liters) of No. 2 fuel oil dyed red. The sea conditions were calm, with about
1-m swell and winds of 2 - 4 m/s. The outer line in each drawing represents the extreme
edge of the visible slick, the next inner line is the region of color fringing when visible in
the photograph, and the crosshatched area is the region of thick oil. The oil formed a
well-defined thick region surrounded by a very much larger and thinner region. In situ
thickness measurements showed the oil to be 2.4 ± 0.3 mm thick at spots in the cross-
hatched region and typically 2 to 4 Mm thick outside this region. The thick inner region
spread at a much slower rate than the total slick. This is shown in Fig. 4 where the area
of the inner region and the total area of the visible slick are displayed as a function of time
on a log-log plot. If the dashed lines are taken to represent the measurements, the total
area grew at a rate proportional to the time to the 0.6 power; the thick region grew at a
rate proportional to time to the 0.2 power. The spreading rate of the total area most
nearly matches the gravity-viscous spreading phase described theoretically by Fay (6),
which grows at a rate proportional to the square root of the time. It is somewhat slower
than spreading rates reported by Guinard (7) or by Munday et al., (8). However the
spreading rate is dependent on many variables—such as initial volume, age, density and
viscosity of the oil, the surface-active materials present, interfacial surface tension, surface
wind, sea state, and surface current present—and will vary widely. Most significant is the
dichotomous behavior of the oil, dividing clearly into a thick, relatively compact region
surrounded by a second much larger and thinner region. All of the spills conducted of
each oil type exhibited this behavior. It may well be due to small quantities of surface-
active materials in the oil which spread more rapidly than the bulk of the oil, surround-
ing it, inhibiting its growth, and thus containing and controlling the oil.*
The microwave observations were taken using the NASA-Wallops Island DC-4 aircraft.
Measurements were made at 19.4 and 69.8 GHz for the initial spills and at 19.4 and 31.0
* Private communication with W. D. Garrett.
VI - 69
-------
TIME: 09:32
09:36
10:09
10:46
-3
O
"Z
JO
r*
»
13
-J
01
12:02
13:27
I METERS
100
200
300
4OO
500
'Fig. 3 — Tracings of color photography of the oil slick resulting from a controlled oil spill of 630 gallons of No. 2 fuel oil. The oil had been dyed
red to allow the thick regions of oil to be identified visibly. The outer line in each drawing represents the extreme edge of the visible slick, the
next inner line is the region of color fringing when visible in the photograph, and the crosshatched area is the region of thickest oil.
-------
HOLLINGER AND MENNELLA
10
Itl
a:
10'
iO*
SLOPE 0.6
, _- SLOPE 0.2
I02 I03
_1 1 I I I, ! I I
10"
TIME (SECONDS)
Fig. 4 — Area of the inner thick region (dots) anil the
total area of the visible slick (crosses) measured from the
photography of the oil spill represented in Fig. 3 is plot-
ted versus the time the picture was taken. The dashed
lines are possible representations of the measurements.
GHz for the last three spills. The latter combination proved more effective for the oil
thicknesses of up to several millimeters which were encountered. The half-power antenna
beamwidth at all three frequencies was 7.2 degrees, which gave a beam spot on the sur-
face of about 50 ft (15 m) for the aircraft altitude used of about 400 ft (122 m). A
two-dimensional antenna-temperature map of the oil slick was built up by making repeated
aircraft passes over the extent of the slick. Approximately 15 to 30 minutes before
and after the nominal time of the map were required to acquire sufficient scans for the
map.
Contour maps of the increase in antenna temperature above that for the unpolluted
sea at 19.4 and 31.0 GHz are shown at the right in Pig. 5 superimposed on the outline
of the visible slick for the spill of July 11, 1972. These antenna-temperature distributions
were used to derive the thickness contours shown at the bottom left of the figure. The
antenna temperatures and derived thicknesses are weighted averages over the antenna
beam, as given by Eq. (1). The half-power beam spot on the surface is represented by
the small circle. The microwave signals coincide closely with the region of thick oil and
show that average thicknesses over the antenna beam of up to 1.5 mm are present in
good agreement with in situ spot measurements in this area of 2.4 ± 0.3 mm. Integration
of the thickness contours derived from the microwave data gives a volume of 650 ± 65
gallons (2460 ± 246 liters), which taken with the volume of oil spilled of 630 gallons
(2380 liters) indicates that nearly all of the oil is in the thick region. This is consistent
with in situ measurements of film thicknesses of 2 — 4 urn outside the thick region,
since only 15 to 30 gallons (57 to 114 liters) of oil would be needed to cover the entire
VI - 71
-------
VISIBLE SLICK
31.0 GH*. ATa(°K)
3db BEAM SPOT
THICKNESS (mm )
19.3 GHz. ATa(°K)
-=:
H
i
ro
ya
%
o
93
H
100 200 300 0 100 200 3
DISTANCE (Meters)
Fig. 5 — The upper-left-hand drawing is a tracing of a color photograph of the oil slick resulting from a
controlled spill of 630 gallons of No. 2 fuel oil. The oil had been dyed red to allow the thick regions of oil
to be identified visibly. The outer line is the extreme edge of the visible slick, the next inner line is the re-
gion of color fringing, and the crosshatched area is the region of thick oil. The antenna temperature mea-
sured at 19.3 and 31.0 GHz are shown at the right superimposed on the outline of the visible slick. The
thickness contours derived from the microwave data are shown at the bottom left.
-------
HOLLINGER AND MENNELLA
area of the visible slick of 33 X 103 m2 with a uniform film to thicknesses of 2 — 4
The ratio of slick thickness in the two regions of nearly 1000 also shows that nearly all
of the oil is located in a small region of the slick.
The microwave measurements of all of the oil spills of each oil type showed results
very similar to those just described for the spill of July 11, 1972. The slicks always
formed an identifiable region with film thicknesses of a millimeter or more and contain-
ing the majority of oil, which was surrounded by a very much larger and thinner slick
which contained very little of the oil. In general the thick region contained more than
90 percent of the oil in less than 10 percent of the area of the visible slick. It was al-
ways possible to locate and delineate the thick region solely form the microwave obser-
vations, and the total volume of oil present derived from the microwave measurements
was within about 25 percent of the volume of oil spilled.
In summary, multifrequency passive microwave radiometry offers the potential to
measure the distribution of oil in sea-surface oil slicks and to locate the thick regions
and measure their thickness and volume on an all-weather, day-or-night, and real-time
basis. As such it should prove a useful tool in the confinement, control, and cleanup
of marine oil spills.
ACKNOWLEDGMENTS
The authors thank S. R. MacLeod and J. A. Modolo for their fine work in the
design, construction, calibration and installation of the radiometers and for their help in
taking the measurements. We express our appreciation to the staff at NASA-Wallops
Island Station for their invaluable assistance and aircraft support and to the Virginia
Institute of Marine Science for their aid in conducting the oil spills and obtaining ground
truth. This work was sponsored by the U. S. Coast Guard under contract number
Z-70099-2-21881.
REFERENCES
1. D.V. Stroop, "Report on Oil Pollution Experiments-Behavior of Fuel Oil on the
Surface of the Sea," presented to the Committee on Rivers and Harbors, House of
Representatives, Seventy-First Congress, Second Session, Hearings on H.R. 10625,
U. S. Government Printing Office, 1930, p. 41.
2. W.D. Garrett, "Impact of Petroleum Spills on the Chemical and Physical Properties
of the Air/Sea Interface," NRL Report 7372, Feb. 16, 1972.
3. L.M. Brekhovskikh, Waves in Layered Media, Academic Press, New York, 1960,
p. 45.
4. R.N. Bracewell, "Radio Astronomy Techniques," Handbuch der Physik 54, 42
(1962).
5. "Discharges of Oil for Research, Development, and Demonstration Purposes,"
Federal Register Document 71-5369, Federal Register 36, No. 75, 7326 (1971).
VI - 73
-------
NRL REPORT 7512
6. J.A. Fay, Oil on the Sea, Plenum Press, New York, 1969, p. 53.
7. N.W. Guinard, "Remote Sensing of Oil Slicks," in Proceedings of the Seventh
International Symposium on Remote Sensing of Environment, Ann Arbor, Michigan,
May 1971, p. 1005.
8. J.C. Munday, Jr., W.G. Maclntyre, M.E. Penney, and J.D. Oberholtzer, "Oil Slicks
Studies Using Photographic and Multiple Scanner Data," in Proceeding of the Seventh
International Symposium on Remote Sensing of Environment, Ann Arbor, Michigan,
May 1971, p. 1027.
VI - 74
-------
THE NATIONAL ENVIRONMENTAL
MONITORING SYSTEM
by
Theodore Major
Environmental Systems Manager
The Magnavox Company
Fort Wayne, Indiana
ABSTRACT
The first NASA developed GOES (Geostationary Operational Environ-
mental Satellite) is scheduled for deployment early In 197**. After
engineering checkout its operational control will be assumed by NOAA
(National Oceanic and Atmospheric Administration). NOAA has planned
to make the communications relay (transponder) capability of the
satellite available to all agencies planning or operating Environmental
Monitoring Systems comprised of automatic unattended Data Collection
Platforms (DCP). The objective of this paper is to provide a review of the
capabilities of the satellite communications system, its developmental status
as well as similar descriptions for some of the DCP planned for incorpora-
tion in the overall system which has been titled 'The National Environ-
mental Monitoring System." The applicability of the system to water
pollution control systems is discussed and some recommendations for
specific pollutant sensor developments are made*
SYSTEM 'DESCR1PT ION
The "National Environmental Monitoring System11 (Figure 1) is
based on a plan by the United States to deploy two Geostationary
Operational Environmental Satellite (GOES) over the western hemisphere.
The first of these is scheduled for launch early in 197^. The European
Space Research Organization, the USSR, and the Japanese each plan to
launch a satellite in their segment of the belt. The five satellites,
all commun i cat ions- compat i ble, will form a worldwide belt of Environ-
mental Satel1i tes.
VI - 75
-------
Each GOES satellite (Figure 2) will have the capability to relay
data from ^0,000 Data Collection Platforms every six hours. The Data
Collection Platforms, consisting of both land-and water-based stations,
will be equipped with environmental sensors and radio sets which will
transmit the digital sensor signals over one of 150 possible UHF channels
at 100 bits per second.
One of the principal applications of the Data Collection Platforms
will be that of hydrological monitoring of fresh water resources.
Fresh water control is necessary not only to agriculture, for flood
control and irrigation, but to all citizens as a vital resource for
health and quality of life. The degree of water pollution i'n our streams,
rivers, and lakes is directly proportional to the volume of water available
for dilution and natural processing of waste materials. Accordingly,
until the zero pollution discharge objectives of the Environmental
Protection Agency (EPA) are achieved, water resource conservation and
control will continue to play a vital role in water pollution control.
The hydrological stations will be installed in grids across major
water sheds to measure precipitation, water levels, water flow, and in
some instances water quality. The data from these grids of DCP's wi1!
be used to manage the retention or release of water by the existing
networks of flood control dams as well as to predict flash floods.
The National Environmental Research Center (NERC) in Cincinnati
have been developing unattended water quality monitoring stations.
(Figure 3) They have successfully demonstrated the feasibility of trans-
mitting water quality da*"a via satellite from the Great Miami River
Pollution Monitoring Station. (Figure k) Plans have been made to equip
this station with the capability of monitoring specific pollutants
(heavy metals) and to transmit the data via the ERTS satellite. The
specific pollution monitoring equipment will be Hes gned :n ru^gedized
form for om^at i bil i ty with shipboard environmental parameters so that
the modules can be assembled to provbe a pollution monitoring capability
for relatively small patrol vessels, as shown in Figure 5. Bo^h EPA and
the USCG currently operate a number of such vessels which perform
visual and sample acquisition monitoring functions.
Manual shipboard monitoring on a large scale can be prohibitively
expensive. Water Quality Monitoring Buoys (Figure 6) equipped with low
power, reliable sensors (currently available or in development) provide
a cost effective alternative. Typical parameters that can be monitored
from a buoy include DO, pH, temperature, conductivity, chlorophyl and
turbidity. An outline drawing of a developmental ch 1 or^phy 1 - ^urb idometer
package is depicted in Figure 7- The buoy Data Collection Platform
becomes particularly attractive if the monitoring data suite includes
meteorological and oceanographic parameters such as currents, tempera-
ture profiles, etc. When synoptic reporting is required, the buoy
DCP become economically mandatory.
VI - 76
-------
The National Weather Service Automation of Field Operation and
Services (AFOS) Program and related programs provides for automation
of bo«~h manned and unmanned weather stations. An unmanned automated
weather station is depicted in Figure 8. This station will be battery
powered and modular for packaging into remote areas. The highly
directional antenna can be manually trained to the calculated GOES
satellite elevation and azimuth. The directivity of the antenna reduces
DCP transmitter power and thus permits a smaller battery pack.
The National Oceanic and Atmospheric Administration plans to deploy
a number of moored meteorological buoys (similar in configuration to
the experimental buoys shown in Figure 9) off the North American Pacific
coast. Data from these buoys will be used locally to provide maximum
safety to the Trans-Alaskan Pipeline tankers as well as to improve weather
forecasting over the North American continent. The tankers and similar
platforms of opportunity, as shown in Figure 10, may be equipped with
the sensors and electronics installed in the buoys to provide mobile
automatic meteorological platforms. Another mobile meteorological
platform is the Drifting Meteorological Buoy developed by the NOAA
National Data Buoy Office. (Figure 11) This drifting buoy is equipped
with a position location system utilizing NNSS or Omega data to repo~t
buoy position along with the environmental data.
REGIONAL AND LOCAL MONITORING SYSTEMS
The National Environmental Monitoring System can be used to supply
data from inaccessable remote DCP's to a regional or local monitoring
system, and to report summary data from the regional system into the
national system. A Great Lakes Environmental Monitoring System is
depicted in Figure 12 as an example of a regional system. In this
system, data from the buoys inaccessable land stations will be transmitted
via the GOES satellite to the CDA station, and then back to the Regional
Data Acquisition and Processing Center via land line. To fulfill all
data requirements, the Great Lakes Environmental Monitoring System will,
on a selective parameter module basis, be capable of providing meteorological
and 1imnologicat, as well as water quality data.
The need and requirements for automatic unattended Water Quality
Monitoring Systems have been fully established and documented (see
References 1,2, and 3). In summary, the systems must provide the follow-
i nq .
1. Rapid intelligence and alarms
2. Continuous surveillance
VI - 77
-------
3. Effective enforcement
k. Data for water quality models
5. Basis for issuance of permits
6. Effects of natural phenomena
7. Large numbers of measurements
8. Computer analysis of data
9. Cost effectiveness
10. Measurement of specific pollutants
11. Monitoring of water quality trends
It is anticipated that the Great Lakes Water Quality Monitoring
Platforms will include fixed land-based stations equipped with modular
sensors and electronics as required to perform the measurements to be
made at the selected site. In addition, mobile platforms such as tra-Mers
or vans and vessels will be used, particularly to investigate critical
areas.
An example of a local area monitoring system is the USCG Oil Spill
Surveillance System shown in Figure 13. The objective of this system
will be to protect a harbor from accidental oil spills through the
provision of oil slick sensors mounted on buoys, bridges, piers, build-
ings etc. Data from these sensors will be telemetered to a central
processing control station via radio and hard-wire circuits. After
processing, the oil spill detection data could be transmitted via the
GOES satellite for action by the USCG and the responsible governmental or
civilian organizations involved.
THE GOES COMMUNICATIONS SYSTEM
Figure 1^ shows that the major components of :he GOES Data Collection
System (DCS) are the Command and Data Acquisition (CDA) Station, the
GOES spacecraft (SC) and the remotely located Data Collection Platforms
(DCP). While the spacecraft carries a family of remote sensors and
associated data distribution systems, it is the cross-trapped transponder
that is of primary interest to the technical community concerned with
earth-based Data Collection Platforms (DCP). Transmission between the
CDA and spacecraft is at S-band, while transmission between the field
of DCP's and the spacecraft is at UHF. A block diagram of the GOES
Data Collection System is shown in Figure VS.
The NOAA National Environmental Satellite Service (NESS) has developed
Data Collection Platform Radio Sets (DCPRS) for both mobile (Ship, buoy,
and fixed (land-based) Data Collection Platforms (Figure 16)). The fixed-
location DCPRS are further provided in self-rimed and interrogated con-
figurations. The self-timed sets automatically transmit available data at
preset intervals while the interrogated set transmissions are commanded
from the COA station. The specifications for these sets are tabulated
in F igure 17.
?! - 78
-------
For the fixed location DCPRS the helical antenna shown in Figure 18
is used. The gain of this antenna is considered to be +10db for link
calculation purposes, but careful pointing in the field will yield
nearly 13db of gain. For the mobile platforms, an omnidirectional antenna
is required; a conical log-spiral antenna with right-hand circular
polarization is used. The gain for link calculation purposes is considered
to be kdb. To preserve the link margin on the mobile DCP, the transmitter
power is increased to **0 watts.
The interrogated DCPRS uses the Interrogation Command and Reply
Message Formats shown in Figure 19. The detailed frequency plan is
shown in Figure 20. For the DCP to SC reply messages, 150 channels are
available Studies have shown that the average transmission from DCP
to SC will be 80 seconds in duration; accordingly up to ^0,000 DCP
transmissions can be processed over a six-hour period. For those messages
that require more than the two-minutes maximum DCP reply, the
platform may be re interrogated until all of the data is accumulated.
Wallops Island will perform CDA station functions for the GOES
satellite. It will be equipped with the Data Collection Subsystem Control
and Demodulator shown in Figure 21. This equipment has the capabilities
listed in Figure 22. For the DCP interrogation functions, the DCP addresses
will be received over land lines from the computer-at the NOAA Suit land
Maryland facility, and will be stored in the Data Collection Subsystem.
After data formatting, the Data Collection Subsystem transmits interroga-
tion commands to the Data Collection Platforms via the spacecraft and
receives data responses from the DCP. The data from the DCP is multi-
plexed in the Data Collection Subsystem.for transmission over unconditioned
telephone lines to the computer at Suitland.
SENSOR DEVELOPMENTS
The success of the National Environmental Monitoring System is
largely based on the availability of sensors capable of providing accurate
measurements for months of unattended operation under adverse operating
conditions. NOAA has recognized the problem in the National Data Buoy
Program and has supported a meteorological and oceanographic sensor
development program which has begun to yield significant results.
This agency has recently undertaken the development/evaluation of reliable,
low-power water quality sensors compatible with the long-life, unattended
requirements of the National Data Buoys. The results of this program will
be available to the environmental community late in 197**.
- 79
-------
Similar developmental programs should be initiated in analytical
instrumentation for specific water pollutants. Figure 23 indicates applicable
laboratory analytical methods for the detection/quantification of the
pollutant materials listed in the Oanadian/US Agreement for Improving
Water Quality in the Great Lakes. In general the laboratory equipment is
designed for attended operation with some of the functions such as sample
pretreatment, transfer and disposal performed by the operator. Further
the operator is often required to read meters, log results etc'. The
ultimate objectvies of analytical sensors development programs must
i nclude:
1. Automatic sample acquisition
2. Automatic sample pretreatment
3. Selection of techniques compatible with automation
k. Selection of techniques providing modular assembly of
an unattended Data Collection Platform
5. Design for installation in all possible types of platforms
6. Outputs compatible with available communications systems and
computer processing
Basic technology and specific related experience are already
available to perform unattended analytical sensor development for
specific water pollutants. The NERC Laboratory in Cincinnati has ex-
tensive experience in Sample Acquisition Systems and Water Quality
Sensors. In the area of severe, corrosive, marine fouling environmental
the NOAA National Data Buoy Office and the Navy ASW offices and associated
industrial contractors have a great reservoir of talent and technical
experience. In the area of analytical instrumentation, the Instrument
Industry stands ready to make its contributions in design for specific
applications. A mu 11 i-d isc ipl i ne team of systems-oriented engmeers
and scientists working under the leadership of the responsible govern-
mental agency and drawing on the extensive technology available can,
with minimum risk, assume the challenge to extend unattended monitoring
to specific water pollutant control.
The primary objective of this paper is to acquaint the reader with
the developmental status of the National Monitoring System. Since the
water pollution instrumentation area is the pacing item relative to the
full implementation of this system it may be useful to review the develop-
ment . elements of one of the more promising analytical methods. Energy
Dispersive X-Ray (EDX) Spectroscopy was chosen since it is a relatively
new technique with great potential for rapid and accurate analysis of the
heavy metal family of materials. Since pumps, reagents and considerable
power consuming equipment- are involved with fhis t-echnique, it is not
applicable to buoy platforms but is very compatible with vessels and fixed
land platforms.
VI - 80
-------
The first function to be performed is Sample Acquisition (Figure 2k),
Reliable pumps and inert materials must be selected in combination with
filter mechanisms and flushing systems to prevent fouling by marine
organism and contamination of the sample. Provisions must be made to
capture and hold a sample in response to commands issued by the logic
built into the analyzer, or in response to those generated by the data
processing computer or operating personnel. The sample from the Sample
Tank may be applied to a number of analyzers and their associated pre-
treatment equipment.
Figure 25 depicts the Sample Pretreatment envisioned for the EDX
analyzer for dissolved heavy metals. The auto-analyzer metering pump
draws a 0.1 liter sample which is reagent-pretreated to form the metal
precipitates. The sample is passed through a tape filter to collect
the precipitates. The filter is then indexed to a position below the
x-ray generator and the concentric diode detector for the derivation
of the element energy spectrum shown in Figure 26. The detector is
shown in greater detail in Figure 27• The output of the 1 ithium drifted
silicon diode detector is applied to a mini-computer which functions
both as a mu1ti-channel analyzer and a data manipulation device; thus the
spectrum shown in Figure 26 does not need to be generated as a display,
but can be printed out on one of the many available computer output
pr i nters .
The principle of detector operation is based on the generation
of secondary x-rays in the sample by illuminating it with low level
primary x-rays such as those radiating from a radio-isotope. The
secondary x-rays, of unique energy for each element, are columnated and
directed through a beryllium window to strike the liquid nitrogen cooled
silicon diode. Both the diode and its associated FET pre-amplifier
are operated in the liquid nitrogen environment to minimize thermal noise
The secondary x-rays that strike the detector ionize the silicon atoms
in the diode, creating electron hole pairs and free electrons proportional
to the energy of the incoming x-rays. This signal is amplified and stored
in a memory bin corresponding to its energy level. The electronic system
can handle up to 20,000 signals per second of various energy levels.
The detection process continues over a preset time base in the order of
100 seconds. At the end of the detection process, the memory bins may
be read out in the form of an energy spectrum with the energy of each
signal identifying the metal and the peak amplitude the concentration.
This read-out process can be done automatically in the form of a set of
numbers .
The EDX laboratory equipment (shown in Figure 28) has demonstrated
sensitivities in the order of 60 parts per billion. The EDX system
performs elemental analysis on all material above sodium (Atomic No.11).
VI - 31
-------
References
1. Meriting A. F.. March 1965. "Instrumentation for Water Quality
Determination" presented at the ASCE Wster Resources Engineering
Conference.
2, Bal linger, 0. G., May 1969, "Analytical Instruments in Water Pollution
Control" presented at the ISA Symposium (15th) on Analysis
Ins<• rumentat ion.
3. Maylath, R. E., April 1971, "Automatic Surveillance of New Yorks
Waters'1 presented a*" the ISA Symposium (17"h) on Analysis
Ins trumentation.
VI - 22
-------
NATIONAL ENVIRONMENTAL
MONITORING SYSTEM
T Seismic
Stations
c^
Ships
°f 4
Opportunity JijL
AY
Cv A
-4
4.
4
A
JjfT
Command
and Data
Aquisition
Weather
Stations
REGIONAL
STATIONS
A
Moored and Drifting ^_
Environmental Buoys
Hydrological
Stations
,£.
VI - 53
-------
HYDROLOGICAL STATION
a
i
FWD149-11
-------
I
CD
reat Miami River GMR Pollution Monitoring Station
Well Station Water Sample
GMR Monitoring Station
FWD375-37
-------
Pollution Monitoring Vessel
<
n
I
SAMPLE
DISCHARGE
AUTOMATIC
SPECIFIC
POLLUTANT
MONITORING
CONSOLE
FWD149-10
SAMPLE
INTAKE
-------
WATER QUALITY MONITORING MOORED BUOY
73-8-3538
VI - 87
-------
Chlorophyll Detection and Turbidimeter Package
CHLOROPHYLL DET.
SECTION x FLASHLAMP
ELLIPSOIDAL MIRROR SCATTERED RADIATION
DETECTOR
/ CONDENSING LENSES
FLASHLAMP
COLLIMATING
OPTICS
DETECTOR
WINDOW
COLLIMATED
BEAM STOP
) BACK6RQUND
/N RADIATION
\ SHIELD
TBTO IMPREGNATED
GUARD RING
DIRECT BEAM DETECTOR
TURBIDIMETER
SECTION SPECTRUM-SHAPING
- _ FILTERS
SECTION A-A
ELECTRONICS SECTION
(AMP, LD, & INV)
WATERPROOF SEALED CABLE TO
ELECTRONICS IN OIL DETECTION
PACKAGE
TURBIDIMETER
ENERGY STORAGE
SANK
CHLOROPHYLL
DETECTOR
ENERGY STORAGE
BANK
CALIBRATION DYE
RESERVOIR
BOTTOM VIEW
FWD375-22
VI - 88
-------
METEOROLOGICAL STATION
1
oo
0897TM
-------
MOORED
ENGINEERING EXPERIMENTAL BUOY
(40' DIAMETER)
H
o
-------
-s
M
I
-^., .-
*"_, _^^*"*-
-
-------
VI - 92
-------
Great Lakes Environmental Monitoring System
M
I
<0
THUNDER
BAY Q
TAWAS CITY
o
EPA, REGION V
DATA ACQUISITION
AND PROCESSING
^•ROCHESTER
=S
SAGINAW
ST. CLAIR
RIVER *
DETROIT
*
MONROE
TOLEDO.
*
MAUMEE
RIVER
ERIE
O
ASHTABULA
* Ol
CLEVELAND
FWD319-13
-------
USCG - Oil Spill Surveillance System
POTENTIAL OIL POLLUTION
SENSOR LOCATIONS
0897(TM)
-------
Data Collection System Employing Interrogated DCPRS
••*!»
B>:0
°'
HlliOC
GOES SATELLITE
S-BAND
2034.9 MHz
INTERROGATE
1694.45 REPLY
UHF
468.825 MHz
INTERROGATE
401. 7-402.0 MHz
REPLY
CDA STATION
FWD37-1
VI - 95
-------
VJV^U,«J LSAld, \^,\JLL
SATELLITE 401.7—402 MHz
T 7 -4 ,,. ••_,
c^nuii oyaitm
\/ x7 401.7— 402 MHz -> ""
^ tf V \^468.82b MHz
/^[^ 401.84 MHz \
T 1 \ PT = 10 W/40 W
\ \ g =. 8 dB
2094.9MHz \ \ 1694.5 MHz
g = 13.7 dB PT - 20 W
g = 19.1 dB 468.825 MHz v^v 401.85 MI
^^N. ij]
Y'5
DUPLEXER ^
2034.9 MHz \ \ 1694.5 MHz
\ u
x \ \ 1 y TRANSMIT i '
\^llxf PT = 1U W
7~^\ o - 48 dB RECEIVER
AFC CON.
T •
J^t^TO 74.9 MHz INTERR.
MOD = ±70 PSK TOR
MANCHESTER CODED
1 , .
74.45 5 150
MHz MHz> DEMOD
)Ub-ll
468.825 MHz
xir MOD -
DATA RATE
lz RECEIVE MC
MANCHESTER
IANSMIT
, = 5 W/40 W
PILOT 5 MH-
i TRANS- ^ •
MITTER
^ *- _.,...,.,™™
ADDRESS ADDRESS
r FACE
DATA
FIELD OF
DCPRS
±70° PSK
100 EPS
)D = ±70° PS1
!. CODED
CDA
5 MHz
STATION
STANDARD
COMPUTER
-------
NOAA/NESS OCR RADIOS
FIXED
i/icn o
-------
8
DCP Radio Set Specifications Fixed And Mobile
VG8505-14
PARAMETER
• FREQUENCY
RECEIVE
TRANSMIT
• NO OF CHANNELS
RECEIVE
TRANSMIT
• MODULATION
TECHNIQUE
• TRANSMITTED POWER
• RECEIVER SENSITIVITY
• ANTENNA
• STANDBY POWER
• SUPPLY VOLTAGE
• TEMPERATURE
• SIZE (LESS CASE)
• WEIGHT
-------
-------
INTERROGATION DCPRS COMMAND FORMAT
4
15 BITS
FIME
CODE
31 BITS
CMnnF ADDRESS
4
TIME
CODE
REPLY MESSAGE FORMAT
RF CARRIER ONLY
* SEC 1
» . ...- - _. - - -,.„_, — — —
ALT.
W
I T ]
2-1/2
«- CCP«
"^bttl
7
15 BITS
31 BITS
DATA
MLS ADDRFSS
CODE AUUKt^
Ml MI ITT MAY
^•i
UNLESS REIN1ERROGATED
FOR ANOTHER 2 MINUTE
PERIOD.
VG8505-8
VI - 10Q
-------
DCPRS/SC/CDA Detailed Frequency Plan
10
Ol
01
<*
o
un
oo
1/3 Ol O
O IO O
Ol IT) O
1^ Ol O t—
01 cn o o
Ol CTl l~
1^
^
o
•t
CO
CDA
DATA DEMODULATOR
BAND
10
Ol
01
O
ro
in CTI o
O ID O
o> m o
00
01
CDA
REPLY IF
FREQUENCIES
O!
CTl
O
O
to a\ o
o ID o
cn in o
•* ^ »4
r^ cn O
^-«3- uv
oo co oo i—
« • -o
I—* t-H »—I I
O O Ov-
• o.
00
cn
ov
UHF REPLY LINKDCPRS/SC
SELF TIMED
50 CHANNELS
SPACED 3.0 KHZ
INTERROGATED
100 CHANNELS
SPACED 1.5 KHZ
S-BAND REPLY LINK SC/CDA
0>
vo
LO
to
co
10
co
10
«*•
IT)
CM
CO
CO
>£>
o
o
01
o>
co
UHF INTERROGATION
LINKSC/DCPRS
10 KHZ
S-BAND INTERROGATION
LINK CDA/SC
NOTE:
ALL FREQUENCIES IN MHZ
r~
1
10
KHZ
T
1 1 1
i
o
in
s
CM
If)
CM
CO
O
00
O
o
01
CO
O
CM
LO
l-~
Ol
oo
o
CM
o
uo
o
in
oo
o
CM
o219-14
VI - 101
-------
Data Collection System
CDA Control And Demodulator
VI - 102
-------
CDA GROUND STATION DATA COLLECTION SUBSYSTEM
•3
f
FEATURE
• Simultaneous reception of 20 reply channels
• Automatic acquisition and phase lock tracking of reply channels
Matched filter data detection
Automatic acquisition and phase tracking of pilot carrier (referenced to CDA 5 MHz standard)
Pilot carrier synthesis referenced to COA 5 MHz standard
AFC Control of Interrogate Channel frequency
Multiplex of reply data at 2400 bps for transmission on unconditioned lines
Automatic generation of address request
Self-contained built-in test equipment
Ease of maintenance through plug-in modules
DESIGN FLEXIBILITY
• Expandable to 150 channels
• Expandable to provide Bit Error Rate testing
• Adaptable to on-site computer control
ES64451-1
-------
MATERIALS
vs
METHODS
MATERIAL
i
X
oc
O
5
<
a:
a.
•<
cr
o
O
ce
cr a.
H- o
oo o
5 1
o o
< S
PHENOL
T-DDT
DIELDRIN
HEPTACHLOR
METHOXYCHLOR
LINDANE
ALDRIN
ENDRJN
IRON
MERCURY
ARSENIC
CADMIUM
ZINC
PHOSPHORUS
CHROMIUM
SELENIUM
COPPER
LEAD
NICKEL
JARIUM
MICROBIOLOGY
DISSOLVED OXYGEN
DISSOLVED SOLIDS
pH
RADIOACTIVITY
TEMPERATURE
SUSPENDED SOLIDS
OIL
AMMONIA
CHLORIDE
CYANIDE
FLUORIDE
SULFATE
_SPIEC_IFIC_
GENERAL
FWD319-1
VI - 104
-------
SAMPLE ACQUISITION
<;
M
*
o
ut
\
WATER
D. I. COLUMN
SAMPLE
PUMP
MICRO
FILTER
1
D. I. WATER
FLUSH TANK
THROTTLE i-
VALVE
TANK
HOLDING
TANK
SAMPLE
TANK
TO ANALYZER
FWD319-9
-------
SAMPLE PRETREATMENT
8
r
1 L
DiAn
u-i
I
1
1
1
rZER
1
SONIC
OUAUQED
SAMPLE RUMP "'"]"•"""
V V /
1 1 -. _
EDX
1 1 1 1 1
1 -*•_—».-
__ __ //• \
! - , ,-L-L, (O ^
i T ! CHEMICAL
1 1 REAGENTS
1 i
i i i
r- 1 r-**-| r~ 1 r 1
AA j j FE j j ASP 1 COLORIMETRY f ""
L__i L_J i 1 i J— ~
7 \ ° Cl
FILTER
PAPER
TO DRAIN
FWD460-5
-------
ENERGY-DISPERSIVE X-RAY
a
i
I
ED X-RAY
PRETREATMEI^
/ / / /
III!
II I i
REAGENTS
-------
AUTOMATED ENERGY DISPERSIVE X-RAY SYSTEM
PRE-
TREATMENT
00000
IS
F™o
000
(&
I
u
EDX
I
REEL
DRAIN
/ 'L a
FILTER ' •=
TAPE
r
ing)
REEL
1
LIQUID N2
77°K FET
Si (Li
)
^
DIODE
„ Be WINDOW
yl
FILTER TAPE . ^^*^^v"i
>. \ f* I
X \Xvi
V
1V
A
,—
^
i'^—
i
- SAMPLE F
'
'PT
FWD170-3
VI - 108
-------
1
I—'
o
ENERGY DISPERSIVE X-RAY
(Laboratory Configuration)
!2-1349
-------
SESSION VII
SATELLITE ENVIRONMENTAL MONITORING APPLICATIONS
CHAIRMAN
MR. JOHN D. KDUTSANDREAS
OFFICE OF MONITORING SYSTEMS, OR&D
-------
SKYLAB APPLICATION TO ENVIRONMENTAL MONITORING
by:
Victor S. Whitehead
NASA Johnson Space Center
Abstract
The Skylab Earth Resources Experiment Package (EREP)
provides a view of the'Earth with relatively high spa-
cial and spectral resolution through the visible and
parts of the near and thermal infrared regions of the
spectrum. Although the design of Skylab restricts its
use as an operational platform for environmental moni-
toring, observations taken with the Earth Resources
Experiment Package are being used in some research
studies directly related to environmental monitoring
and in a larger number that have an indirect bearing
on environmental quality in general. There are 148
experiments in progress using the EREP facility.
These include such varied areas of investigation as:
vegetative characteristics such as crop vigor, land
use and regional planning, sea surface characteristics
and circulation, atmospheric contamination, and water
pollution, all these are factors to be considered in
environmental quality. These experiments will not only
provide increased knowledge of the environment but will
also provide guidance on design requirements for future
observation systems. EREP observations are taken at
the times and places required to support these 148 ex-
periments. The data collected, however, is placed
immediately in the public domain, available to any in-
vestigator.
The systems within the EREP that are being used in
environmental monitoring are:
Multispectral Photographic Facility S190A
Earth Terrain Camera S190B
Skylab Spectrometer SI91
Multispectral Scanner SI92
The capabilities and restrictions of these systems,
along with a description of the quality of data they
are providing, is presented in the text.
¥11 - 1
-------
SKYLAB APPLICATION TO ENVIRONMENTAL MONITORING
by:
Victor S. Whitehead
NASA, Johnson Space Center
The Skylab was launched and inserted into a 234 n. mi. circular
orbit at 50° inclination on May 14, 1973. Eleven days later, the space-
craft was occupied by a three man crew to begin the first of three planned
extended periods of scientific observations from space. The scientific
objectives of Skylab are:
0 To study man - Determine physiology conditioning and perform-
ance capability in real time in zero-gravity environment for long-duration
space flight.
0 To study the sun - Synoptic survey and study of special phenom-
ena on the solar disk in x-ray, ultraviolet, and visible spectral wave-
lengths.
0 To study space technology - Evaluate coating degradation,
spacecraft contamination, manufacturing and repair techniques, and manned
maneuvering units.
0 To study the earth - Synoptic survey of selected areas on the
earth in visible, infrared, and microwave spectral wavelengths.
This last objective will be the only one discussed here. It should
be noted that the facility, crew-time, and some consumables are shared
among objectives. Scheduling of experiment performance then is a criti-
cal effort in achieving optimum use of the Skylab and, at times, can be-
come quite complex when individual experiment requirements conflict with
each other.
To observe the earth, the crew has as its tool, the Earth Resources
Experiment Package (EREP)(Fig. 1), a cluster of five instruments systems
designed to aid the scientific community in determining the spectral and
spatial resolution required for earth resources applications, the utility
of microwave in earth resources surveys, and the effects of atmosphere in
optical and electronic data analysis.
VII - 2
-------
Of the five EREP systems three
1) The multispectral photographic facility with earth terrain
camera,
2) The infrared spectrometers system, and
3) The multispectral scanner
produce data which are directly applicable to some areas of environ-
mental monitoring.
The other two systems,
4} Microwave radiometer-scatterometer and altimeter, and
5) The L-band radiometer
are oriented more to test of.the feasibility of acquiring microwave
(
data from space than to direct application of the resulting data. The
discussion here is directed to the utility of the data collected by
the first three systems. Only a cursory description of the individual
systems and integrated Skylab system is given here. Those interested
in a more detailed description are referred to the EREP Users Handbook,
"NASA Document S-72-831-V and the Skylab EREP Investigators Data Book."
Multispectral Photographic Facility (S190A) with Earth Terrain Camera
(S190B)
The S190A provides radiometrically and metrically accurate images
of ground radiance for a wide range of user-oriented studies. It con-
sists of a band of six calibrated boresighted 6" focal length cameras
capable of handling a variety of film-filter combinations. Output for-
mat is 70 m square. The field of view of 21.2° provides an 88 n. mi.
by 88 n. mi. view from 234 n.ymi. altitude. Forward overlap of up to
90% is possible. All housekeeping data is recorded. The standard load
for S190A and its characteristics are:
Wavelength Resolution
Film Filter Micrometers Ft.
PAN-X B&W (S0022) AA 0.5-0.6 99
PAN-X B&W (S0022) BB 0.6-0.7 91
IR B&W (EK2424) CC 0.7-0.8 223
IR B&W (EK2424) DD 0.8 - 0.9 223
IR COLOR (EK2443) EE 0.5-0.88 223
AERIAL COLOR (S0356) FF 0.4-0.7 78
VII - 3
-------
The S190B provides high resolution photography in support of other
EREP sensors and user-oriented studies. It is compensated for spacecraft
motion and can provide up to 85% forward overlap. The focal length is
18-inches providing ground coverage of 59 n. ml. square. The film for-
mat is 4.5 inches square. Film and filters available for use in S190B,
and their characteristics, are:
Wavelength Resolution
Film Filter Micrometers Ft.
AERIAL COLOR (S0242) None 0.4-0.7 60
AERIAL B&W (EK3414) W-12 0.5 - 0.7 40-60
AEROCHROME IR, COLOR W-12 0.5-0.88 150
(EK3443)
The Infrared Spectrometer System (SI 91)
The Infrared Spectrometer System with Viewfinder tracker, more
appropriately called the Skylab Spectrometer, provides quantitative
determination of the effects of atmospheric attenuation upon radiation
from surface features over a broad spectral range thereby providing in-
puts into correction for these effects. This instrument consists of a
filter wheel spectrometer that spectrally scans the radiation entering
its aperture from 0.4 urn to 2.5 urn and from 6.6 urn to 16.0 urn, once
each second and a viewfinder tracker system, with zoom magnification
from 2.25 to 22.5, that looks in same direction as the spectrometer and
allows the astronaut to find, track and photograph for record, the site
the spectrometer views. The system has internal calibration sources
that can be inserted into the field of view on command. The spot size
on the surface is 1/4 mi. diameter. The spectral resolution of the sys-
tem is of the order of 2 or 3% of the wavelength at any wavelength.
The tracker and spectrometer is gimbaled to view 45° to the front
10° to the rear and 20° to either side of track.
Multispectral Scanner (S192)
The SI92 scanner provides radiance values simultaneous in 13 bands
in the visible, near IR, and thermal IR portions of the Spectrum. Each
VII -
-------
channel is calibrated 100 times each second. It has a conical scan
with a radius of 22.6 n. mi. The instantaneous field of view is 260
feet at the ground (.182 mi Hi radians). The swath width -on the ground
is 37 n. mi. The bands with asterisks listed below are sampled 1240
times per scan line the remainder at 2480 times per scan line. There
are 94.79 scan lines per second. Design absolute accuracy of the system
in the visible and near IR was 5%, in the thermal IR I.^OK.
The spectral bands covered are:
Band Coverage, Mi crometers
1* 0.41 to 0.46
2* .46 to .51
3 .52 to .56
4 .56 to .61
5 .62 to .67
6 .68 to .76
7 .78 to ,88
8* .98 to 1.08
9* 1.09 to 1.19
10* 1.20 to 1.30
11 1.55 to 1.75
12 2.10 to 2.35
13* 10.20 to 12.5
There are restrictions due to the design and operational require-
ments of Skylab that limit its EREP data gathering ability. The orbital
inclination restricts its nadir viewing capability to +_ 50° latitude.
The.orbit was also planned to provide repetitive coverage,
at five day intervals of defined ground tracks; this, coupled with the
high resolution - narrow field of. view of the sensors leaves, on a nomi-
nal mission, significant gaps between tracks that cannot be covered.
The Skylab is designed to operate in a solar inertial attitude to provide
maximum power from the solar cells and to support astronomical observa-
tions; to operate EREP in an earth viewing mode over a ground track it
must leave this attitude, thereby expending consumables and decreasing
the power supply. This restricts Skylab to one or two EREP passes per
VII - 5
-------
day. Except for system tests, data taken by EREP is not telemetered back
to the surface, instead it is stored and returned with the crew. This
weight and volume of tape and film also place an upper limit on data tak-
ing ability. In spite of these restriction Skylab will provide from EREP
upon its completion: sixteen rolls of 5 mm film (450 frames per roll),
78 rolls of 70 mm film (400 frames per roll), nine rolls of 16 mm film
for the viewfinder tracker (5,600 frames per roll), and 25 reels of 28-
track magnetic tape (7,200 feet per reel).
Given the capabilities and restrictions of Skylab how can it best
be applied to the problem of environmental monitoring? The system has
the spatial and spectral resolution required to detect features of day-
to-day concern, such as oil spills, but the restrictions of EREP would
make such observations almost worthless for operational use. Such obser-
vations could, however, be used in research to determine the aerial ex-
tent of such spills under the existing conditions. The most promising
application of Skylab to the problem of environment monitoring, however,
does not appear to be the detection or tracking of specific contaminates,
but rather that of providing a better understanding of the environment.
These synoptic multispectral and sometimes repetitive views, used in
conjunction with other data sources, and with models, are providing a
new perspective to environmental observers. Examples of application are:
Determination of the circulation patterns in an estuary, determination
of area! extent of effect of air pollution on vegetation"about a city,
Determination of land'use patterns and desirable land use patterns,
modeling of atmospheric diffusion.
There are 148 approved Skylab EREP experiments in which the crew
obtains the data requested over the site described and under the conditions
desired (to^the extent possible) by the investigators. In many cases,
the investigator is acquiring complementary data at the surface at the time
of the overflight. The investigators obtain no proprietary rights of
the Skylab data. It goes immediately into public domain upon retrieval.
The 148 investigations are broken down into nine broad discipline areas
which are described here. Subdivision of these areas which are most
likely to be of interest to environmental monitors are underlined,
VII - 6
-------
however, almost every investigation has as its prime or secondary
objective a better definition of the environment or to test the feasi-
bility of using high spatial and spectral resolution information acquired
from spacecraft to better define the environment.
AGRICULTURE/RANGE/FORESTRY - 20 Tasks
Studies of crops, forests and rangelands, including management, re-
source eyaluation, mapping and inventory information, crop identification,
acreage estimation, irrigation needs, land productivity and crop yicfor,
soil surveys, soil conditions, soil distribution of insect infestations
and freeze-thaw line/snow cover supportive studies.
GEOLOGY - 37 Tasks
Geomorphology, soil erosion, volcanic activity, resource and envir-
onment , geothermal anomalies, lithology, surface water loss, fault tec-
tonics, earthquake hazards, geologic mapping, mineral exploration, high-
way engineering.
CONTINENTAL WATER RESOURCES - 17 Tasks
Mapping snow field distribution and water equivalent, investigating
soil moisture distribution in plains area, measuring ice parameters,
charting and cataloging estuary effluents, measuring changes in migratory
bird habitats, delineation of good quality ground water and areas of high
saline, and monitoring of flood control.
OCEAN INVESTIGATIONS - 16 Tasks
Obtain data on ocean currents, potential fish abundance, geoidal
undulation, sea surface conditions, sea and lake ice, water depth, estuar-
ine and coastal processes, water color and circulations and plankton pop-
ulation in upwelling areas.
ATMOSPHERIC INVESTIGATIONS - 13 Tasks
Various meteorological investigations, including mesoscale cloud,
features, orographical effect on the formation of mesoscale disturbance,
VII - 7
-------
solar and terrestrial radiation measurements, aerosol concentrations,
cloud statistics and characteristics, day/night detection of cirrus
clouds and other particulates, severe storm environments, and'Stratos-
pheric aerosol concentrations as_ a function of altitude.
COASTAL ZONES, SHOALS AND BAYS - 12 Tasks
Happing hydfobi 61 ogical conrtunities, s6nSinc[ bay'and coastal envi r-
onments, coastal water circulation, river plumes, sedimentology analysis,
water depth analysis, water pollution, developing monitoring techniques.
REMOTE SENSING TECHNIQUES DEVELOPMENT - 11 Tasks
Relate signatures of EREP imagery to ground spectra, adaptation of
discrimination techniques, radar altimeter terrain characteristics
identification, microwave pulse response of rough surfaces, and sensors
performance evaluati ons.
REGIONAL PLANNING AND DEVELOPMENT - 30 Tasks
Land use mapping, crop identification, acreage mensuration, urban
studies, land classification, effects of strip mining^ e f f 1uent water
patterns, recreation site analysis, water resource development and
management, transportation planning, assess^ f 1 re damage and erosion
sources.
CARTOGRAPHY - 8 Tasks
Photo mapping, map revision, map accuracy, thematic mapping, surveys,
mapping techniques.
- 8
-------
On the basis of samples of data taken from the first manned mission
and telemetered data from the second mission, it appears that the EREP
systems S-190A and B, and 5-191 are performing as well as, or better than,
expected. These preliminary data also indicate that S-192 is not quite
performing as expected in all channels, however, considering the use of
ground computer preprocessing techniques these data will be provided to a
quality comparable to ERTS.
VII - 9
-------
NASA-S-73-2842
MULTICHANNEL
SCANNER
MICROWAVE
RADIOMETER
CAMERAS
CAMERAS
L BAND
RADIOMETER
INFRARED
SPECTROMETER
-------
ENVIRONMENTAL APPLICATIONS OF THE
EARTH RESOURCES TECHNOLOGY SATELLITE
Dorothy T. Schultz
General Electric Company
Space Division
Beltsville, Maryland 20771
and
Charles C. Schnetzler
Planetology Branch
Goddard Space Flight Center
Greenbelt, Maryland 20771
On July 23, 1972 NASA launched the first earth-resources dedicated re-
search satellite, called ERTS-1 (for Earth Resources Technology Satellite),
into a near-circular, near-polar orbit at an altitude of 920 km (570 miles)
( Figure 1 ) . The satellite reoccupies the same orbital path every eighteen
days, and thus it is possible, depending on cloud cover, to obtain up to
20 views of the same area in a year. The satellite orbits the earth about
every 103 minutes. It carries a multispectral scanner (MSS) imaging system
that produces images about 185 km (100 nautical miles or 115 statute miles)
on a side. Thus an area of approximately 34,000 sq. km (13,000 sq. miles)
is presented in a single image. Another imaging system, the return beam
vidicon (RBV) is also on the satellite but a malfunction occurred in this
system soon after launch, and the system was deactivated.
The MSS is a line-scanning device which records a scene simultaneously in
four bands, each band image being composed of approximately 7.5 x 10 picture
elements (pixels). The four solar-reflected spectral bands are: 0.5 - 0.6
um (green), 0.6 - 0.7 urn (red), 0.7 - 0.8 um (near infrared), and 0.8 - 1.1
urn (near infrared). Figure 2 shows the Salt Lake, Utah area imaged in the
four bands. Note that the different bands bring out different features - for
example, vegetation and water turbidity are best shown in the shorter wave-
lengths, while geologic structure and the water/land interface are best shown
in the infrared images.
The video MSS scanner signal is converted to digital data and telemetered
from the satellite to a receiving station on earth. The final data products
which are produced at the GSFC Data Processing Center are digital tapes,
black and white images of individual bands made from the digital data, or
false-color composites of several bands. The resolution of the ERTS
imagery is about 70 - 80 meters, approximately the size of a pixel, but
varies depending upon the shape of the object and the contrast between the
object and its surroundings.
There are approximately 330 NASA-supported investigations which use data
from ERTS-1 in such disciplines as agriculture and forestry, mineral
VII - 11
-------
resources, water resources, land use, meteorology, and environment sur-
veys. There are presently about twenty-seven investigations in the
environment discipline, and about twice this number in other disciplines
which are secondarily concerned with environmental problems. The fol-
lowing are just a few examples of environmental applications from ERTS
gleaned from progress reports to NASA or papers presented at a symposium
on ERTS results held in March of 1973.
To date, ERTS data have been found useful for a variety of environmental
applications. Studies of air pollution, of fresh water and ocean water
pollution, of disrupted land surfaces, of regional and local environmental
management planning - all have used ERTS data successfully.
The New Jersey Department of Environmental Protection has found ERTS a
unique tool for environmental planning \_L~] . Figure 3 is an ecozone
map, prepared for the entire state from three ERTS images. Ecozones are
defined as "regional areas characterized by homogenous interrelationships
of soils, landforms, vegetation, geology, drainage and land use". Because
of their regional areal size and uniform characteristics, the state con-
siders that ecozones should logically be recognized as integral regional
planning units. The ecozones listed below contain critical environmental
resources, and the state Department of Environmental Protection has expressed
a need for special protection and regulation of these zones: the coastal
zone (coastal bays and wetlands); the pine barrens (unique forest associ-
ations and extensive aquifer zone); the agricultural belt (prime agricul-
tural land); the highlands and the Kittatinny Mountains (relatively undis-
tured forest areas). A small scale, synoptic view is required for the
recognition and delineation of regionally similar land areas. The state of
New Jersey found ERTS data uniquely suited to this purpose because each
image covers approximately 34,000 sq. km (13,000 square miles).
Imagery from the ERTS satellite is also capable of contributing to environ-
mental management on the local scale. The coastal wetlands environment is
one which requires careful management because of its importance to wildlife
and its value to land development. The Nanticoke River Marsh is part of the
Chesapeake Bay wetlands network which provides food and habitat for a
variety of marine life and which serves as one of the most important win-
tering areas on the Atlantic Flyway. ERTS imagery has been used for the
development of a vegetation group classification map of the Nanticoke
Marsh [Y] .
The amount of flooding in a wetland area is a major determining factor in
vegetation species mixture. "Low marsh" areas, or those which ate most
often flooded, cover the greatest area of the Nanticoke River Marsh. Large
stands of Juneus roemarianus, Scirpus sp., Distichlis spicata, Spartina
alterniflora and Salicornia sp. are found in these areas. Low marsh appears
darker than high marsh on ERTS imagery because of the low reflectivity of
the deep water in which the plants stand. The classification "high marsh"
describes drier areas. These often occur at the wetland/water boundary
where the land surface has been built up by deposits from streams. This
category is largely made up of Spartina cynosuroides. Spartina patens/
Distichlis spicata association, Iva futescens and Baccharis halimifolla.
The 10 October 1972 ERTS image was used to produce the vegetation map
shown in Figure 4. High marsh and low marsh areas are mapped, together
with a third classification described as low marsh/water. The latter has
VII - 12
-------
been established due to the difficulty in defining some of the interior
boundaries. Low growth or sparse plant cover exhibit a very low reflectance
because of the water background, and so are not easily distinguished from
pools of surface water. The vegetation map also delineates major high
ground islands which are covered with trees.
ERTS imagery is a useful tool for determining wetlands extent and boun-
daries, plant community composition, the position of mud flats and berms,
the extent of ditching, filling and other man-related activities and
successional trends in vegetation. Such information is required by federal
agencies involved with wetland mapping and coastal zone management as well
as by state agencies concerned with wetland and coastal problems. For use
by these agencies, the specific, up-to-date information supplied by ERTS
is uniquely applicable to studies of the constantly changing wetlands
environment.
Several ERTS investigations of air pollution are presently in progress.
Studies include those involving the location and monitoring of individual
pollution sources, the altered reflected radiance from the surface as a
result of air pollution, the effects of air pollution on vegetation, and
the geometry and movement of air pollution plumes. One of these investi-
gations has resulted in proof of the theory that pollution of the air in
our cities can cause local weather modification £3] . Figure 5 shows
an ERTS image of Lake Michigan taken on November 24, 1972. Particulate
plumes are visible from at least seven steel mills and power plants located
in the Chicago - Gary, Indiana area. The plumes feed directly into the
shallow convective cumulus clouds over the lake. It is apparent that the
smoke causes premature cloud formation, as those clouds along the plume
line form noticeably closer to the shoreline and become more dense and
well-developed over the lake. Snow was reported falling from these cloud
"fingers".
Several investigations have demonstrated the usefulness of ERTS imagery
for delineating strip-mined areas and for evaluating reclamation success.
Because vegetation is highly reflective in the 0.8 to 1.1 um ERTS band and
bare soil absorbs near-infrared radiation, the satellite images show dis-
tinct contrast among stripped land, partially vegetated areas and fully
vegetated areas. In Figure 6 is shown an enlargement of an ERTS image of
Indiana acquired in August of 1972. The two counties shown were last map-
ped for strip-mined'acreage in 1968. The accompanying figure shows a map
of stripped land made from this ERTS image \Jt^j . The resulting map was
in close agreement with inventory information obtained from individual
coal companies. A comparison of the image and map reveals that certain
areas which were mined before 1968 appear to remain partially or wholly
unvegetated at the time of this ERTS pass. Acreage and reclamation status
of stripped land can be monitored from the satellite on a yearly or even
seasonal basis. Mapping these areas by ERTS imagery both saves time and
reduces cost in comparison to the conventional ground survey methods.
Another application of ERTS in the area of disturbed land surfaces is the
mapping of fire burns. The Division of Forestry in California is required
VIT - 13
-------
to submit a fire map and fire report to the District Office within ten
days after a wildfire has been extinguished. The present procedure for
drawing up such maps requires that individuals walk or drive around the
fire perimeter, drawing boundaries on a topographic map. For very large
fires, low-flying aircraft are employed while an individual draws the
fire boundaries freehand on a map sheet.
Figure 7A is a map of the Fiske Creek (left) and Pocket Gulch (right) fire
burns, as it was drawn up by the California Division of Forestry. The CDF
spent approximately 4 to 8 hours (including flying the fire to draw the
map, plus time to use a dot grid for area estimation), or about $500, to
map the Pocket Gulch burn. The map in Figure 7B was drawn in about 25
minutes (after the image was in hand) from an ERTS image of the burned
area (Figure 7C) (jjj . The cost of drawing up the' ERTS map was about
$50. Low altitude oblique aerial photography has indicated that the fire
perimeter map in Figure 7B is more accurate than the one drawn up by the
CDF.
A variety of studies involving water resources and water pollution have
found ERTS to be a significant tool in the study of pollutional processes
in lakes, in rivers and in coastal and estuarine waters. The satellite
data have registered a variety of pollutant types including sediment,
organic pollution such as sewage sludge, acid industrial wastes, oil slicks,
and surface algal blooms. An example of the latter appears in Utah Lake
in the Figure 2 images. The green algae, bearing reflective properties
in the near-infrared spectrum similar to land vegetation, show white in
the 0.8 to 1.1 urn ERTS band against the black of the highly absorptive
water. The bloom is the result of an overabundance of nutrients such as
phosphorous and/or nitrogen in the lake. Not only is the bloom an indica-
tion of the trophic state of the lake, but also demonstrates the pattern
of wind circulation against the Wasatch Mountains east of Provo, Utah.
The October 11, 1972 image of the Washington, D.C. area (Figure 8) shows
the Potomac River heavily laden with sediment which had washed down from
the west in a heavy storm which occurred on the 10th of October.
Suspended sediment is more light-reflective than is clear water, and so
it appears lighter in the image. The Potomac and the Rappahannock Rivers,
whose watersheds drain from the west, are obviously high in sediment con-
tent in this image. However, the Patuxent River, which drains mostly from
the north, and the Choptank and Nanticoke which drain the eastern shore of
the Chesapeake Bay, are much less turbid, and therefore darker in color.
It is apparent that on this day the sediment content in the Occoquan River
(entering the Potomac from the west, south of the D.C. metropolitan area)
was considerably lower than that in the .Potomac. The outfall from this
tributary, dark on the image, moves southward along the west bank of the
Potomac, while the more turbid waters from the upper Potomac cling to the
right. Gradual mixing occurs, and most of the heavily concentrated sedi-
ment has been dispersed before the entrance of the Aquia River and the
Potomac Creek at the first major eastward bend.
VII - 14
-------
While many of the investigations being carried out with ERTS-1 data have
relied on photographic reproduction of the imagery, we find that digital
interpretation and manipulation of the data products leads to a capability
for resolving features which are indistinguishable in the photographic pro-
ducts and, in addition, allows identifications and/or correlations which
are amenable to quantitative and statistical treatment. For example, the
October 11 image, shown in Figure 8, has been studied in greater detail by
computer analysis of the digital data [6^] . Sedimentation in the Potomac
and Anacostia Rivers on this date and on the preceding ERTS pass were com-
pared using pattern recognition and multi-channel density slicing techniques,
The result of the analysis is shown in Figure 9. Most of the sediment in
the Potomac estuary in the 23 September scene was found either in the
Anacostia River channel or downstream below Occoquan Creek. Low sediment
values in the region below Roosevelt Island are associated with local sewage
outfalls. Using this digital technique, most major sewage effluent sources
in the estuary can be located. The very high sedimentation values found
in the Anacostia are from two major sources: a sand and gravel operation,
and runoff from open surfaces associated with construction of roads and
buildings. When compared with the earlier image, it is clear that the
heavy storm runoff on October 11 has resulted in high sedimentation levels
throughout the estuary, obscuring effluent outfa.lls.
Several types of water pollution were identified in an ERTS image of the
New York Bight area. Figure 10, which is an enlargement of the red ERTS
band image of that scene provides an excellent example of the detectability
of industrial wastes, sewage sludge, pollution plumes and water mass bound-
aries in estuarine waters by ERTS imagery (_7J • ^e serpentine curve in
the ocean south of the center of the image represents a dumping of acid-iron
wastes, containing about 8.5% H2S04, 10% FeSO^, and small quantities of
various metallic elements. Immediately to the north of this spill is a
dark outline indicating a sewage sludge dump, and directly to the west of
the dump can be seen the polluted waters of New York Harbor, Note the
almost physical boundary between the harbor waters and the ocean. Ground
truth confirmation of these features was obtained by aircraft underflight
and the waste disposal authorities of New York City for the day of the pass,
August 16, 1972. Continued monitoring of such features provides useful
information about circulation patterns, thus aiding in the optimization of
ocean dumping for maximum dispersal and minimum environmental impact.
In summary, although monitoring of pollution and environmental degradation
by satellite remote sensing is inherently less specific than in-situ
measurements, it has the advantage of global and repetitive coverage, thus
providing the basis for an economic, operational capability. Unfortunately,
the temporal or spatial resolution of the ERTS system is not satisfactory
for many environmental problems. Ideally, a satellite system dedicated to
environmental monitoring should have a spatial resolution of at least 10
meters, be independent of cloud-cover or sunlight restrictions, and should
be able to essentially continuously monitor any particular area. The study
of images received from the first Earth Resources Technology Satellite,
however, has demonstrated the capability of remote sensing by satellite to
detect and monitor naturally occurring environmental changes as well as
man-induced stresses to the environment.
VII - 15
-------
References
[Y) Yunghans, R. S. and Mairs, R. L.: Application of ERTS-1 data to
the protection and management of New Jersey's coastal environment: in-
vestigation review summary, October 29, 1973.
[2] Anderson, R.: Semiannual progress report to NASA, May 1973.
(~3j Lyons, W. A. and Pease, S. R. : ERTS-1 views the Great Lakes,
paper W17 from Symposium on significant results obtained from the Earth
Resources Technology Satellite - 1, vol. I: p.847 - p. 854, March 5-9,
1973.
(4] Wier, C. E.: Fracture mapping and strip mine inventory in the
midwest by using ERTS-1 imagery, paper El from Symposium on significant
results obtained from the Earth Resources Technology Satellite - 1, vol.
I: p. 553 - p. 560, March 5-9, 1973.
[Sj Colwell, R. N.: Semiannual progress report to NASA, January 1973.
[&] Schubert, J. S. and MacLeod, N. H.: Digital analysis of Potomac
River Basin ERTS imagery: sedimentation levels at the Potomac - Anacostia
confluence and strip mining in Allegheny County, Maryland, paper E13
from Symposium on significant results obtained from the Earth Resources
Technology Satellite - 1, vol. I, p. 659 - p. 664, March 5-9, 1973.
[TJ Wezernak, C. T. and Roller, N.: Monitoring ocean dumping with
ERTS-1 data, paper E10 from Symposium on significant results from the
Earth Resources Technology Satellite - 1. vol. I, P. 635 - p. 642, March
5-9, 1973.
VII - 16
-------
RETURN BEAM
VIDICON CAMERAS
DATA COLLECTION
SYSTEM ANTENNA
COMMAND
ANTENNA
SOLAR
PADDLE
ATTITUDE
CONTROL
SUBSYSTEM
.WIDEBAND
ANTENNAS
UNIFIED
S-BAND
ANTENNA
MULTISPECTRAL
SCANNER
Figure 1. The Earth Resources Technology Satellite.
VII - 17
-------
•
*
V
r
Figure 2. ERTS imagery of the Salt Lake City, Utah area: 0.5-0.6 pm band
upper left, 0.6-0.7 urn band upper right, 0.7-0.8 um band lower left,
0.8-1.1 um band lower right.
VII -
-------
REGIONAL ECOLOGICAL MAP
NEW JERSEY
LEGEND
A COASTAL ZONE: coastal lands, wetlands and
water directly affected by coastal processes
B PINE BARRENS: contiguous forest cover with
low Intensity land use
C LAKEWOOD: forested area with mixed residen-
tial and commerical land use
D VINELAND: mixed agriculture and forest
E AGRICULTURAL BELT: extensive farmland with
small woodlots and some urban development
F URBAN AND INDUSTRIAL ZONE: areas of Inten-
sive land use
G PIEDMONT PLAIN: mixed cropland and urban
land with scattered forested traprock ridges
H HUNTERDON PLATEAU: curvilinear forested
ridges and cleared valleys
I UPPER DELAWARE RIDGE AND TERRACE: rolling
terrain with forest and agricultural use
J KITTATINNY MOUNTAIN: steep series of forest-
ed ridges with low intensity land use
K KITTATINNY VALLEY: rolling topography with
forested ridges, cleared valleys (agricultural
use), and numerous small lakes
L HIGHLANDS: rugged, partially forested area
with numerous lakes
M WASHINGTON: level valley (rural land use)
enclosed by Highlands Ecoione
N PASSAIC BASIN WACHUNG MOUNTAINS:
forest cover and urban land use in a level river
basin ringed by forested, traprock ridges
O RIDGEWOOD: urban land use and forest cover
Scale in Miles
^SfEEE
10 20 30
This pkotomap produced from a NASA ERTS-1 mosaic of MSS band 5 taken on October 10, 1972
Figure 3. New Jersey ecozone map made from ERTS imagery.
VII - 19
-------
1:250,000 enlargement of Nanticoke R. salt marsh
Scale 1:250,000
Nanticoke River Marsh
1:250,000 map of Nanticoke River marsh.
Figure 4.
-------
••'.///
MICH
Figure 5. ERTS mosaic of Lake Michigan showing inadvertent
weather modification by pollution plumes.
VII - 21
-------
I
ro
EXPLANATION
Areas strip mined
prior to 1968
Areas mined since
1968. Mapped from
ERTS-1 imagery
Figure 6. Indiana strip mine map and enlarged ERTS image,
-------
ro
0)
A. (Scale « 1:125,000)
B. (Scale • 1:110,000)
C. (S^ale = 1:780,000)
Figure 7. (A.) Ground survey map, (B.) ERTS data map and (C.) ERTS image of fire scars in California
-------
Figure 8. ERTS red band image of the Washington, D. C. area,
VII - 24
-------
POLLUTION DETECTION IN POTOMAC RIVER
NORMAL POLLUTION LEVELS
SEPTEMBER 23,1972
•
AFTER A STORM
OCTOBER 11, 1972
VERY HIGH SEDIMENT LEVEL
IX,8,9, BLANK)
IB HIGH SEDIMENT LEVEL
(H)
IB SEDIMENT
(S)
I 1 NORMAL BACKGROUND
IN)
ORGANIC POLLUTION
(P,L,0)
HUNTING
CSEEK
ERTS DIGITIZED DATA
AMERICAN UNIVERSITY ERTS PROJECT UN443
Figure 9. Digital enhancement of ERTS data for two dates
at the confluence of the Potomac and Anacostia Rivers.
"II - "5
-------
H
'
Figure 10. Industrial waste and sewage sludge dumps in the New York Bight area.
-------
ERTS-1 DETECTION OF ACID
MINE DRAINAGE SOURCES
by
Elliott D. DeGraff and
Edward Berard
Ambionics. Incorporated
400 Woodward Building
Washington, D.C.
638-6469
VII - 27
-------
ERTS-1 DETECTION OF ACID
MINE DRAINAGE SOURCES
Sponsored by EPA's Office of Water Programs and Office of
Monitoring, through their interagency agreement with NASA LaRC
we are using data from ERTS-1 in a program to locate and monitor
sources of acid drainage pollution from the coal fields in the
Upper Potomac River Basin.
ERTS-1 was launched by NASA in July, 1972. Three of its
unique features have been brought together for the first time
in resource inventory and monitoring. These are spatial
coverage of 34,000 square kilometers on each image scene;
repetitive coverage of any area on earth every 18 days,* and
registered simultaneous viewing in four separate, spectral
bandwidths ranging from green to near infrared.
This new technology has potential application to locating
sources and identifying causes of pollution on earth. As you
have seen, ERTS images show contrasts in areas of known pollu-
tion. These data require further interpretation and confirming
VII - 28
-------
measurements to determine the nature and quantity of the
polluting source. The development of a data acquisition pro-
gram appears feasible which will utilize these data and could
be used to monitor large land areas for pollution resulting
from surface and shaft mining operations as well as other
causes. Continuous monitoring of these areas may well indi-
cate the status of both existing and new pollution sources.
Such rapid detection capability will allow corrective action
to be promptly implemented.
Description of ERTS-1 data appear elsewhere. Suffice
it to say that analysis of the imagery shows that the per-
formance of the Multi-spectral scanner (MSS) is excellent.
There is clear separation between the bands and ground resolu-
tion is approximately 150 feet. Of interest to EPA, indica-
tions of stream flow, algae blooms and sedimentation and
dispersion patterns have been noted. Urban areas show clear
differences from current land use maps (in many cases based
on data 5-10 or even 40 years old) and watershed topographic
features (including surface-mined terrain) stand out in clear
detail.
VII - 29
-------
With these facts in mind EPA and NASA have contracted
with Atnbionics to perform a study and test of the applica-
bility of using ERTS data, combined with aircraft data and
traditional methods to inventory the sources of mine drainage
pollution in the Upper Potomac Basin.
It is our intent in this program to investigate the
best means of integrating this new source of data into a
total EPA monitoring system. By demonstrating the usefulness
of advanced technology and remote sensing to EPA's needs in
one small area we hope to provide a model for larger scale
future applications.
The Potomac River is entirely within the state of
Maryland, but is fed by tributaries from Pennsylvania and
West Virginia as well as Maryland. The Maryland water
Resources Administration Laboratory at Cumberland, Maryland
and its corresponding member in West Virginia have extended
their cooperation to not only take periodic field samples
but to provide their analyses. These include the following:
PH
Total Solids
CaCO3
Ferrous Iron
VII - 30
-------
Ferric iron
Aluminum
Sulfates
Turbidity
We are now identifying and acquiring existing data about
the region. These include existing maps, photographs, and
reports compiled by agencies such as USDA, SCS, BuMines,
State Game and Fisheries Commission, etc. These data are
in the form of aerials, reports, land use maps, etc. We have
had excellent cooperation from the local authorities and have
a flood of data coming in. In addition, we are monitoring
and acquiring normal and special data collected during the
period of the study. (i.e. surface data taken contemporaneous
with NASA or state aircraft sensing, and/or ERTS passage over
the test site.) This will extend throughout the length of
the contract.
We have identified and ordered from USGS EROS the applicable
ERTS images. Now that the EROS computer is back on line, we
should have this data shortly.
Analysis will proceed in this approximate sequence.
Location of pollution sources and types on the ERTS imagery
from the previously obtained data. As a control, tributaries
VII - 31
-------
within the test site known not to be subject to mine
pollution, such as Patterson Creek in West Virginia, will
be studied to determine the appearance of regional tributaries
not subject to degradation from active or abandoned mines.
Topographic changes resulting from surface mining
activities shall be located on aircraft data and marked on
ERTS frames.
With the locations determined and checked through
utilization of maps and identifiable landmarks the following
will be checked for appearance and compared with the previously
obtained surface data:
Texture (degree of coarseness)
Shapes
Associations of shapes
Response to varied color bandwidths
All of the above will be analyzed by seasons, location,
and relationships to developments (including mines) and
general direction of slope. From these factors patterns
will be sought relating vegetative cover types to topography
and drainage basins.
While other investigators report the identification of
acid pollution in water, our primary approach will be to use
VII - 32
-------
ERTS in conjunction with traditional methods to identify land
use and pollution sources as well as secondary effects of
pollution e.g. ecological degradation along stream banks.
Our next step will be analysis of the ERTS transparencies
2
using I S Addxtive Color Equipment for enhancement of image
features and varying color bandwidths in combinations and
different degrees of intensity. Results will be recorded
for each image.
A land use map will be prepared based on current U.S.D.A.
thematic maps presently within each county. However it may
be necessary to redefine or modify existing classifications
to types more similar to systems presently in use in the
Census Cities program. This latter is readily digitized
and will save time and money when used with the new G.E.
Image 100 computer system using ERTS images, coded informa-
tion and CCT's.
Utilizing the data from surface investigations and
aircraft to establish accurate controls, the G.E, Image 100
system will be tested and adjusted for accuracy in its
interpretation of the ERTS data entered into the system. Once
perfected, this program may be employed on consecutive mine
drainage search and control programs.
VII - 35
-------
The land use maps based on ERTS will be studied for the
appearance of areas known to be subjected to mine pollution
and similarities noted. Areas of similar appearance, but
heretofore unchecked or with little previous record of
despoliation will be subject to surface and aircraft investi-
gation.
This analysis technique should improve upon the
efficiency of small staffs, traveling over large areas of
difficult terrain taking small samples in the field often at
long intervals, and too often sampling downstream samples of
pollution from unknown sources, whether surface or subsurface,
or broken seals.
We have had great cooperation from the West Virginia
Department of Natural Resources and the Maryland Department
of Water Resources in not only obtaining stream samples and
analyses on specific sites but also maps on farmed areas,
forests and cutover lands, county roads, trails, paths, and
wet weather springs.
The area office of the U.S.D.A. Soil Conservation
Service has an aerial photointerpretation section which has
recently completed land use survey maps based on aircraft
VIII -
-------
photography. At this writing we have supplied and ordered
larger scale ERTS-1 imagery and the SCS is transcribing its
data to the ERTS imagery.
Mine locations within the Test Site and their relation-
ships to the steep choppy topography characteristic to this
area is, in our opinion, a key element to this program.
Direction of slope, rate of fall per mile of principal
streams and their tributaries, and stream density (Total
length of stream in a basin (miles) divided by area of basin
(sq. miles) = stream density (sq. miles)] will contribute
to the distribution, strength, and rapidity of spread of acid
mine pollution. The Federal Bureau of Mines Eastern Field
Operation Center mapping office is locating all abandoned
and existing known mine sites on topographic maps. These
data are computerized by topographic sheet title and are
readily available. All such data for the Upper Potomac Basin
test site are now being transposed onto tographic sheets
obtained by Ambionics from U.S.D.I. and from there will be
plotted on ERTS-1 imagery.
Similarities will be sought on the ERTS imagery common
to the affected areas. The I 2S Additive Color process and
VII - 35
-------
the G.E. Image 100 will be employed particularly where
topography affects vegetative patterns and possibly emissions
and reflectance. It is here that the land use maps may be
most useful. Other features noted will be changes and
differences in colors, shadings, textures, shapes and com-
binations of the above which can be indicative of degraded
areas in the test site. A close watch will be kept for
seasonal and annual characteristics which may provide
dynamic indicators of acid mine pollution that can be
applied elsewhere.
Our ultimate goal is to formulate a methodical approach
in locating and monitoring mine polluted waters that will
be more efficient and economical than techniques in existence
today. These is promise of greater success in dealing with
this earthbound problem through space technology than has
previously been possible.
VII - 56
-------
WATER QUALITY ANALYSIS USING ERTS-A DA.TA
H. Kritikos
L- Yorinks
The Moore School of Electrical Engineering
University of Pennsylvania
Philadelphia, Pennsylvania
H. Smith
Environmental protection Agency
Region 3, Philadelphia
Abstract
The magnetic digital tapes of the imagery obtained by ERTS-A on September 23,
1972, have been analyzed for selected areas of the Chesapeake Bay and the
Potomac River. A statistical analysis of all four bands has been carried out.
The results show that band III is useful in determining the vater to land interface
Data on bands I and II suggest the existence of three distinct types of water
those having low, medium and high reflectivity. Available information from
published literature shows that suspended matter (silt) produces higher than
normal reflectivity. It is reasonable therefore to suggest that the recorded
area of high reflectivity contain high concentrations of suspended matter.
A computer processing technique has been developed which identifies the
above areas and produces thematic maps showing their geographical distribution.
VII - 37
-------
1 . INTRODUCTION
Remote sensing is now rapidly becoming an important tool in the detection
and surveillance of water pollution. The basic physical concepts which led to
the present developments were developed by a number of investigators some of
(k)
v '
which are F. Hulburt' ', j. Williams^ ', and S. Duntley^ . Application of
remote sensing techniques to oil slick detection have been made by N. Guinardv
J. Aukland^ ', j. Monday^1 '. Water quality measurements have been carried out by
M. Querry^ ' and J. Scherz^ ' , M. Golberg^ ' has experimented with Raman scattering
techniques, p. White ^ ' has carried out a study of the spectral characteristics
of sewage outfalls.
A great opportunity for the further development of the art of remote sensing
appeared in MSA's Earth Resources satellite program. A large number of data
for extended geographical areas become available and it appears that some
significant results have been obtained. Some relevant work has been carried out
by the following investigators: v. Klemas^ ' t B. Bowker^ 'and Rugles^ 3^ have
studied the dispersion of high sedimentation plumes from photographic images .
C. Wezernak^1 ' , H. larger ^', R. Mairs'1 ' and J. Schubert^17^ also have investigated
the surveillance of areas with high sedimentation.
!Qie present paper is an additional contribution to the effort of assessing
water quality from digitized ERTS-A imagery.
VII - 38
-------
2 . THE REFLECTANCE OP WATER
The satellites observations were made by using the Multispectral Scanner
which has four spectral bands. The bands are defined in Table I.
TABLE I
MSS SEECTRAL BAUDS
Band Spectral Width in Resolution Gray
Micrometers \± bits level
I .5 - .6 7 12?
II .6 - .7 7 127
III .7 - .8 7 127
IV .8 - 1.1 6 &
The information is transmitted digitally and then it is converted to a photographic
product . Each digital count represents the radiance of a cell of 50 by 80 meters .
It is important to notice that while the information is recorded at the tapes
with a 7 bit resolution (127 gray levels) the photographic products are seriously
degraded. Only a fraction of the gray level resolution remains at the photographic
products (16 gray levels) . it is very important therefore to recognize that
accurate radiometric work can only be carried out with computer compatible tapes
(CCT) .
The total reading which is recorded at the satellite is the ••product of a
number of complex physical phenomena several of which will be discussed below.
The radiance L is given by the well known expression
L=|HT + LA
where R is the reflectivity of the target
H Irradiance illuminating target
T Transmissivity of Atmosphere
L. Path Luminance
- 39
-------
It Is important in the course of our investigation to establish the order
of magnitude of these quantities..
(rn
H, Rogers% ' by measuring separately the ground reflectance has published
some typical values for Barton Pond, which is a small lake. They are:
TABLE II
RADIANCE READINGS FOR BARTON BOND
Sun, Zenith angle ^9°, Date: 9-28-72 (From H. Rogers)
1AND 2 L
mw/cm /sr
1
2
3
if
0
0
0
0
.U?6
.2^2
.lUl
.23k
bits
2k
15
10
3
•3
•3
.3
.50
mw/cm /sr bits
0
0
0
0
.271*
.118
.082
.1062
Ik
7-5
6
1.5
T
.810
.865
.909
.913
R
9
5
2
0
•3
.5
.8
•9
H ,
mw/cm
8. in
S.llf
7.38
5.02
Note in the above table that the path luminance is approximately 50% of
the total reading. This is potentially a source of large errors because of
the uncertainty in its determination. Also it is significant to observe that
the reflectance of the water in band TTT and TV is small. Ibis takes place because
the incident light is mostly absorbed. A number of other investigators have also
studied the optical properties of the water and it has been established that an
optical window exists which is approximately ,2ji wide in the visible range.
She peak of the window lies somewhere between ,k and .5u depending on the
(2)
physical composition of the water. For example for distilled water J. Williams*• '
shows that the peak is at 3\j. and the penetration length is approximately 100 m.
Ihe same data show approximately that average penetration length for band I and
II of MSS is approximately 10 and 1 meter respectively. For natural water the
peak of the window is shifted towards higher wavelength and the penetration length
naturally decreases.
VII - 40
-------
A large number of experiments have been made to determine the spectral
response of the water as a function of its constituents. The results are
difficult to relate quantitatively with ERTS-1 data because of the variable
atmospheric conditions. In general however the following qualitative conclusions
can be drawn. Algae is expected to give low readings in band I, medium in II
and high in m and IV. Sedimentation is expected to give high readings in all
bands except IV.
In this study no independent readings were taken at the ground consequently
it was impossible to make absolute radiometric measurements. It was possible
however from the relative radiance of the scene to derive a substantial amount
of information. It was found that the best procedure was to use band III to
identify water and then band II to assess the water quality. This is discussed
in greater detail in section IV.
VII - 41
-------
3. WATER QUALITY CONDITIONS IN THE
UPPER POTOMAC RIVER TIDA.L SYSTEM
The Potomac River is one of the most environmentally stressed areas in the
(19)
Middle Atlantic states. Effluents from a number of major wastewater treatment
plants serving a population of about 2,500,000 people are discharged into the
system. Figures I and II show the geography of the region and the major wastewater
sources. Some of the important physical phenomena which characterize the Potomac
are discussed below.
a. Tides
The tidal system extends from the Chesapeake Bay all the way to Little Falls
in Washington, B.C. The time at which the satellite was taken was 9:20.2 A.M.
E.S.T. which is 1 hour and 40 minutes after the high tide. This means that
approximately one third of the tidal movement has taken place and that plumes
should point downstream.
There are two areas which have presented tidally induced problems. These
are the Anocostia River and the Piscataway Creek. In the Anocostia River the
exchange rate is very small and consequently the high loads of silt which are
discharged into it remain entraped for long periods of time.
A similar phenomenon takes place in the Piscataway Bay. In this region it
has been observed that the wastewaters of the Piscataway Sewage Treatment Plant
(10 MGD) have a very small dispersion rate and occasionally reach Broad Creek
which lies upstream.
b. Water quality measurements
Unfortunately at the time of the satellite overpass no ground data were
taken. There are however data taken at an earlier date, 9-6-72, close to
Low Tide Slack (13-32) which provide some insight into the state of water quality
of the Potomac.
VII - 4E
-------
The data is shown in Table III and the location of the stations is shown
in Figures I and II. Prom the results the following observations can be drawn.
1. In the Key Bridge region a high chlorophyl reading (5^.8 Hg/l) and- l°w
Secchi Disk reading (26.0") probably indicate a high concentration of algae.
2. in the l^th St. Bridge region a low chlorophyl reading (21.0 Mg/l) and
a high Secchi Disk reading indicate a patch of clear water with low concentration
of algae and sediments.
3- The most environmentally stressed area is the W. Wilson Bridge station
where the dissolved oxygen (2.2? mg/1) has its lowest value, the nutrient loading
its highest (1.35 K>L, 1.90 NH~ in mg/1) and the chlorophyl count is medium
(28.5 Hg/l)• This takes place because this region is downstream of the two
most important sewage treatment plants in Arlington (20 M3D) and Blue Plains
(300 M3D).
k. In the Mathias Point region an interesting phenomenon is observed.
Going from stations 15 to 15A and 16 the salinity nearly goes up by a factor of
four, the chlorophyl count decreases by factor of two and Secchi Disk Depth
increase by a factor of two. These data indicate that at some point in between
lies the salt to fresh water interface.
VII - 43
-------
River Miles from Chain Bridge = 0
Fort Belvoir
STP Ik MCD
Pentagon
STP 1 MGD
HTP k3 MGD
Arlington
20 MGD
PEPCO
HTP 315 MGD
Alexandria
STP 20 MQD
Westgate
STP 12 MGD
Little Hunting Ck.
STP 2.5 MGD
scataway c
STP 10-12 MGD
ZONE I
District of Columbia STP 300 M3D
River Miles from Chain,Bridge = 15
Andrews A .F.B.
ZONE II
Gunston Cove
Figure I
Wastewater Discharge Zones
in Upper Potomac Estuary
VII - 44
-------
Chain Bridge
Mathias Point , Testing
/ Stations
Potomac River Bridge
* ?* '
Chesapeake
Figure II
Potomac River Tidal System
VII - 45
-------
TABLE III
WATER QW.LITY MEASIRENENTS
DATE: 9-6-72
Stations
River Time Secchi Disk Water Total p NH^-N DO Total c Cblorophyl Salinity
Miles Inches Temperature POj^ mg/1 mg/1 mg/l pg/1
1.
2.
2A.
3.
U.
5-
6.
7-
15-
ISA
16.
Key Bridge
Memorial Bridge
ll*th St. Bridge
Hains Point
Bellvue "N8"
W. Wilson Bridge
Broad "N86"
Piscataway "77"
Nonjeraoy F113
. MatMas Pt.
301 Bridge "CN"
98.3
97
96
9>*-5
93
91
88
85
52
1*7
1*3
11.20
11.35
11.1*5
11.55
12.05
12.15
12.30
13130
11.15
10.55
16.17
26.0
30.0
31* .0
2l*.0
22.0
ll*.0
2l*.0
21* .0
2U.O
33-0
1*2.0
°C
23.0
23-5
2l*.0
21* .5
2l*.5
21* .5
2U .0
23.20
21* .80
25.90
15-20
mg/1
.233
.21*3
.2M*
.399
1.181
1.35
1.058
.816
.377
.351
..269
.130
.070
.100
.260
1.680
1.90
2.20
1.80
.150
.210
.115
8.6l
8.02
7.39
5.27
3-17
2.27
2.59
3-31
5.81
1*.75
7.85
29.72
30.71
30.95
27.85
29.22
29.15
29.1*5
26.86
18.1*7
18.59
17.1*7
51*. 8
U3.5
21.0
22.5
21.0
28.5
18.0
1*2.0
13-5
7.5
6.5
O/ 00
1.1*0
3.00
6.60
Readings were taken at midstream
Low Tide Slack Time 13.32
-------
U . IDENTIFICATION OF WATER
In order to determine the areas covered by water it was necessary to first
examine some areas where the ground truth was known. Table IV shows the results
of the survey.
TABLE IV
AVERAGE RADIANCE FOR DIFFERENT TARGETS IN GRAY LEVELS
BAND
I
II
III
IV
WATER
25.32
16.96
9.^6
1.33
URBAN AREAS
28.68
22.37
27.8
lk.3h
VEGETATION
30.61
25.1*5
38.5
22.15
M/UC. READING
12?
12?
12?
Notice that the radiance readings for water was very close to those which
Rogers observed for Barton Bond which indicate that probably the same water
quality and atmospheric conditions were present . Band III presented the largest
contrast between water and its environment and for this reason it was decided
to use it to generate a mask. In order to justify this choice the statistical
distribution of the radiance of the water and its environment were determined and
are shown in Figure III . It can be easily seen that there is no overlap and
consequently there was no ambiguity in the determination of the water .
The water can also be identified by using an adaptive technique as follows .
First choose an ad hoc threshold slightly above the mean level. Find the water
and determine the boundary points . Relax the condition at the boundary points
by increasing the threshold by a predetermined amount (e.g. 10$). *This has an
effect of edge sharpening. This technique was originally tried out and proved
to be successful.
VII - 47
-------
Fraction of Samples
o +
1 +
2+
3 +
4 +
5+**
;,.!.*««.>*»*.>
7+*«v.»»«». **.»***•>*
|l + 4>??9*9*9}*.ttV»V<>M >V* + **«»->*«* ^v
12***
15+
1A»
17+
18 +
19+
ZO+
Z1 +
2A + *
Z5+*
Z'i + *
Z7 +
26 + **
29 + **
30+***
33+** *»***#*+*<
33s::::::::,,,***.,
3f.t****»*««»+ ^ - Environment \ Land
37 + #»*p*vx'«'>4iJt*'W /
3a+***«*****+« (JLZrban Areas
3V + »»**!"»w
<,0t to <•»•»••*«»•> + '*»
<,5+**«
51+*
52.+
53 +
S'.*
55 +
56 +
57 +
5fl +
59 +
60+
61 +
62 +
63 +
64 +
65 +
6<- +
67 +
6F»
69 +
Figure III
Histogram of Brightness of Water
and its Environment. Band III.
VII - 48
-------
5- DATA REDUCTION
Once the water has been identified and the appropriate masks have been
generated the variations of its brightness are examined. From the statistical
distribution of the brightness of all four bands it can be seen that band II
shows the maximum spread. The histogram of band IE also suggests that three
distinct brightness clusters exist. (See figures k and 5.) This observation
serves as the basis for identifying three classes of water. Those which have
high brightness (shown by 0) medium brightness (shown by .) and low brightness
(shown by 1). The choice of the thresholds is not obvious. The thresholds
were arbitrarily chosen to conform with our best judgment. The results show
that with very few exceptions regions of high brightness in bandll had high
or medium brightness in bands I,H, IE. Since bandll offered the maximum
discrimination it was decided to work with that band alone rather than considering
the spectral signature in all four bands
VII - 49
-------
6. INTERERETATION AND SIGNIFICANCE OF RESULTS
The CCT (Computer Compatible Tapes) offer the maximum possible gray level
resolution that can possibly be obtained from the ERTS-1 data. For example in
band II the water brightness covers a region of 9 gray levels (out of a max.
reading of 128) with approximately an error of 1 to 2%. (Approximately one
gray level.) In a photographic product this resolution is greatly reduced
the maximum range has been compressed from 128 to l6 gray levels.
In the computer generated thematic maps (figures k and 5) areas of high
sedimentation content or algae are easily determined because of their high
reflectivity and consequently brightness. These areas are the Anacostla river
and generally most of the western bank of the Potomac. A number of large Sewage
Treatment plants are known to lie in the western bank. These are the Arlington,
Alexandria, and Westgate. In the same vicinity lies the largest one, the Blue
Plains STP which discharges its effluents underwater midstream. The high
sediment areas as predicted from the thematic maps seem to correlate well with
the turbidity of the water as indicated by low Secchi Disc measurements and or
high chlorophyl concentrations. Although the ground truth measurements were
taken 20 days before the satellite overpass it is believed that the conditions
were similar (a few hours after high tide with essentially a constant-slope
declining river stage) to suggest that the correlation is more substantial than a
coincidence. This can be seen in figure 6 which shows the flow conditions of
the Potomac River.
From previous published investigations it can be inferred that areas of
low reflectivity correspond to clearer water (less sedimentation and algae) .
VII - 50
-------
In the thematic maps these areas are shown to be in the vicinity of the lU St.
Bridge and generally the lower eastern bank of the Potomac river. Here again there
is qualitative agreement between the ground truth measurements and satellite data.
The ground measurements have shown large Secchi Disc depts and low chlorophyl
counts in the same areas. An interesting coincidence can be pointed out here;
the water of low reflectivity appears to lie in the vicinity of thermal plants.
This is true in both the lU St. Bridge area and Goose Island one where it is
known that the Pentagon thermal plant (U3 M3D) and the PEPCO generating station
(315 M5D) discharge their thermal effluents respectively.
The usefulness of the results lies in the formation of qualitative over
all picture of the geographic distribution of pollution in the Potomac river.
The most significant observation is that after high tide the pollutants seem to
accumulate in the west bank of the river. This phenomenon has been reported before
and is attributed to the Coriolis forces.
There is evet-y reason to believe that from the present data quantitative
information on water quality can be obtained for water quality surveillance.
This, however, can be accomplished with simultaneous spot measurements of ground
truth which will calibrate the observational system.
ACKNOWLEDGMENT
We would like to thank Mr. R. Berstein of IBM for providing us with the
CCT tapes. We would also like to acknowledge the valuable assistance of
N. Melvin of Ef& (Region 3) in gathering the ground truth and in the interpretation
of results.
VII - 51
-------
i., '. ..... . Pentagon Plant
">: i: ..... STP 1
\HXE3 MSD .
••:;;::::::::;.- •;:•„ i'.. Anacostia
Waterfowl^:.. ":::::.:;:;::,. .,.VRiver
Sanctuary , .:: :i::;::;:i;;:i;;-1 :;;i.
,,1.1 -;---•• .
«.,„,','.«' .Ju.o
(ver *.«<«<•« i
Arlington
STP 20 HCD^iv,,,:
, 9-0BO»o'e:i if*eos
> **
OJ "»
-« 1*
>> 2*
to f*
co •; .
§10.
4J U*
L*.tMri»«i,.>,;i
.... ..
PEPCO ' ifIH::::i;i:!ir-iSJ'ue Plains lit'
Generating 4!:!:':';:::' STP 300 NED
Station - |;f^w;-«
HTP 315 MJD ."igrl^'llii;:- Oxon Creek
Sun-Angle kk
Time 9.20 EST
Date 9-23-72
i
Histogram
Fraction of Samples
I .t-ul /.I.- M 3.1 - .1
Ssiiii'iJJSSiiiir Goose Island
'IvSiU-iiJitf.,
•il't*!*.*•*•••* '.'I"* '•
r. Wilson Brd.
a-e •-.,...,..-, ,,,
Alexandria
STP 20
---- ...
n . i tm, »*
•
Westgate-
STP 12
Mean Brightness
Standard Deviation
legend
^
lh^^'T'7
J.D S • S, JL f
_ n
TW ^ r\
-i-W ;» W
-17.0
1.68
Broad
» Creek
Figure IV
Band II. Upper Potomac
VII - 52
-------
c-
01
Port Tobacco
River
.d
-:-ti!):,J
:-'•'-,'.'.'.','•-
.:
i;:l
•il^I! '. ','.','.".','.'.'.. ..'.'.'.. i
;~-'::!:;"^;-!!:;;:;!
...:, .::.
M;
R
:-;;.
3 /
it
ji
«»* * "\
11 •*•«•-
,,..i. . .
u IM-i"
^-S
hia
nt
.4
-'.
1,
3
.. M
-' '
"! "
'i
..:::..,
; »-n-
:::;
't',',1,
'j
'-;; :"-iti;:::;:f;:"i:i.
^;::::::::::::::::::::::::::::::::::.:;'
-i- ; •*» 1 • j^ » ;- ii ; 1 1 ) I ; ; 1 1 ; I ; ; I ; 1 ; ; ; 1 1 1 " 1 ; _\
"—;:>:';'93""' - i
i* -«o:>fc'-;a':-6p!i!a-;ii;!i!^;'";j;;
^i ;--^!! '-^'.'.'.'.'.ii'.i*'.'.'.','..',-:','.'.'.'.'.]'.'.
'.,> |"..:.i ii5t.-!!..-.«!!!ll!!!)I)I!!)||)
;%:::::;: :i:iiii:'iii::;:::::l!::;::::::
"l i:::::;:;:;:;:;;:::::::;::::::::-::-::
Legend
s 13
Sun Angle U4
Time 9.20 EST
Date 9-23-72
. ^Histogram
< n Fraction of Samples
t.t-Jl 2.1-01 3.t->.l
03 •;'
a '1
c >*
3 "•
7»
03 ft
03 o.
s ».
•P 1 !»*•
"S [,!!*."***
V
......
21****
17
18 ^ 0
Mean Brightness
Standard Deviation
16.1*
2.56
.Figure V
Band II. Mathias Point
-------
o
c
•H
100,000
50,000
20,000
10,000
5,000
2,000
r^
.,000
M 500
i
Cn
200
100
HURRICANE AGNES
PEAK
6/23 189,000 tngd.
6/2U 230,000 mgd.
6/25 1^1,000 mgd,
GRObirD VSA31BSKENTS
SATELLITE OVERPASS
1910 LCD
9/30/72
JAN
FEB MAR
MAY
JIH
JULY
AU3
SEPT
Figure VI
Potomac River Stream Discharge
Near Washington, D. C.
-------
References
1. E. Hulburt
2. J. Williams
3. S. Duntley
k, N. Guinard, et. al.
5. j. Aukland
6. J. Monday
7. M. R. Querry
8, J. Scherz
9. M. Golberg'
10. P. White
"Optics of Distilled and natural water", J.O.S.A., Vol.
35, No. 11, November 19^6.
"Optical properties of the sea", United States Naval
Institute, Annapolis, 1970.
"The visibility of submerged objects", Cambridge, 1960.
'Remote Sensing of Oil Slicks", University of Michigan,
Proceeding of 7th International Symposium, May 1971-
"Multi-sensor oil spill detection", University of Michigan,
Proceeding of 7th international Symposium, May 1971.
"Oil slick studies using photographic and multispectral
scanner data", University of Michigan, Proceedings of
7th international Symposium, May 1971.
"Specular reflectance of queous solutions,
University of Michigan, Proceedings of 7th International
Symposium, May 1971•
'Demote sensing considerations for water quality monitoring,
University of Michigan, Proceedings of 7th International
Symposium, May 1971.
"Applications of spectroscopy to remote determinations
of water quality", Ifth Annual Earth Resources Program
Review, Vol. Ill, MSC-05937, Houston, Jan. 1972.
"Remote Sensing of Water Pollution", international
Workshop on Earth Resources Survey Systems, Vol. II,
May 1971.
VII - 55
-------
11. V. Kleuas
12. F. Ruggles
13. D. Bowker
-. C. Wezernak
15. A. Liud
16. H. larger
I?. J. Schenbert
18. R. Rogers
19. J. Aalto
"Applicability of ERTS-1 imagery to the study of
suspended sediment and aquatic fronts", Symposium on
Significant Results obtained from ERTS-1 NASA, March 1973-
"Plume development in Long Island sound observed by
remote sensing", Symposium on Significant Results obtained
from ERTS-1 NASA, March 1973.
"Correlation of ERTS multispectral imagery with suspended
matter and chlorophyl in lower Chesapeake Bay, Symposium
on Significant Results obtained from ERTS-1 NASA, March 1973.
"Monitoring Ocean Dumping with ERTS-1 Data", Symposium
on Significant Results obtained from ERTS-1 NASA, March 1973-
"Environmental study of ERTS-1 Imagery", Symposium on
Significant Results obtained from ERTS-1 NASA, March 1973.
"Water Turbidity Detection using ERTS-1 Imagery", Symposium
on Significant results obtained from ERTS-1 NASA, March 1973-
"Digital Analysis of Potomac River Basin ERTS-1 Imagery,
Symposium on Significant Results obtained from ERTS-1
NASA, March 1973-
"A Technique for correcting ERTS Data for Solar and
Atmospheric Effects", Symposium on Significant Results obtained
from ERTS-1 NASA, March 1973-
"Current Water Quality Conditions and Investigations in
the Upper Potomac River Tidal System", Technical Report
Wo. ifl, U.S. Dept. of Interior, Federal Water Pollution
Control Administration, Middle Atlantic Region.
VII - 56
-------
Satellite Studies of Turbidity, Waste Disposal Plumes
and Pollution-Concentrating Water Boundaries
V. Klemas
College of Marine Studies, University of Delaware, Newark, Delaware 19711
VIT - 57
-------
Abstract
Satellite imagery from four successful ERTS-1 passes over Delaware
Bay'during different portions of the tidal cycle are interpreted with
special emphasis on visibility of turbidity patterns, acid disposal
plumes and convergent water boundaries along which high concentrations
of pollutants have been detected. The MSS red band (band 5) appears to
give the best contrast,, although the sediment patterns are represented by
only a few neighboring shades of grey. Color density slicing improves
the differentiation of turbidity levels. However, color additive enhance-
ments are of limited value since most of the information is in a single
color band. The ability of ERTS-1 to present a synoptic view of the
turbidity and circulation patterns over the entire bay is shown to be a
valuable and unique contribution of ERTS-1 to both coastal ecology and
coastal oceanography.
VII - 53
-------
Introduction
The Earth Resources Technology Satellite (ERTS-1) has proven its
ability to observe large and inaccessible regions of the earth instan-
taneously in four spectral bands. However, 'ERTS-1 must still show it
can provide information which is unique or less costly than that attain-
able by other means, such as aircraft. In the discussion which follows,
imagery from four successful ERTS-1 passes over Delaware Bay are inter-
preted with special emphasis on turbidity, off-shore acid disposal
plumes, and boundaries between masses of water having different physical
and chemical properties. Attempts to enhance these features, with color-
additive techniques and color density slicing are described. The ability
of ERTS-1 to present a synoptic view of the turbidity and water boundaries
over the entire Delaware Bay region at various stages of the tidal cycle
is shown to be a valuable and unique contribution that ERTS-1 is making
to coastal ecology and oceanography.
VII - 59
-------
Physical Characteristics of Delaware Bay
The Delaware Bay Estuary is a relatively prominent coastal feature
which bounds the Delmarva Peninsula on its northern side. The geography
of this region, including the locations of several convenient reference
points, is shown in Figure 1. Trenton, New Jersey is generally taken to
define the upper limit of the estuary so that its total length is over
130 miles.
Fresh water input to the system is derived mainly from the Delaware
River at an average rate of 11,300 cfs which, in terms of volume flow,
ranks this as one of the major tributaries on the eastern coastal plain.
Together with this large volume of fresh water, the river also discharges
a heavy load of suspended and dissolved material, since its effective
watershed encompasses an area typified by intensive land use, both
agricultural and industrial (Oostdam, 1971). Seaward of the Smyrna
River, the bay undergoes a conspicuous exponential increase in both width
and cross-sectional area so that the strength of the river flow is rapidly
diminished beyond this point. Ketchum (1952) has computed the flushing
time of the bay (defined in this case as the time required to replace
the total fresh water volume of the bay) to be roughly 100 days.
Seasonal variations in river flow cause this figure to fluctuate within
a range of from 60 to 120 days. Consequently, river flow is not a
significant factor in determining the current pattern in the bay except
in the consideration of time-averaged flow. In terms of short period
studies, it is mainly important as a source of suspended sediment and
contaminants.
VII - 60
-------
The seaward boundary of the bay extends from Cape May southeast
to Cape Henlopen, a distance of eleven miles. Tidal flow across this
boundary profoundly affects the dynamic and hydrographic features of the
entire estuary. The effect is especially pronounced toward the mouth
where conditions are generally well mixed. The dynamic behavior of the
tides is closely approximated by the cooscillating model described by
Harleman (1966). In this model, the upper end of the estuary is assumed
to act as an efficient reflecting boundary. Consequently, the actual
tidal elevation at any given point is a result of the interaction of both
a landward directed wave entering from the ocean and a reflected wave
traveling back down the estuary. The tidal range is a maximum at the
reflecting boundary and decreases toward the mouth in a manner dependent
upon the relative phase of the two components. Observations of the tide
at Trenton show a 7-foot range compared to a 4-foot range at the mouth.
A relative maximum appears at roughly the location of Egg Island Point,
where the phase relationship is temporarily optimal. An important
consequence of this behavior is the occurrence of .strong reversing tidal
currents in the Delaware River. The tidal wavelength as computed by
Harleman is 205 miles, so that both peak currents and slack water may
occur simultaneously at either end of the bay.
The tidal flow is modified by the bay's rather complex bathymetry
(Figure 1). Most prominent are several deep finger-like -channels which
extend from the mouth into the bay for varying distances -(Kraft, 1971).
Depths of up to 30 meters are present, making this one of the deepest
natural embayments on the east coast. The channels alternate with narrow
shoals in a pattern which is shifted noticeably toward the southern shore.
VII - 61
-------
on the northern side, a broad, shallow mud-flat extends from Cape May
to Egg Island Point. Considerable transverse tidal shears result from
these radical variations in bottom contours. As a consequence, a marked
gradient structure may form normal to the main axis of the bay as a
function of tidal phase. This intrinsic two-dimensional character, together
with the complex, super-imposed time variations, represents an almost
insurmountable task when it is confronted with the tools of conventional
hydrographic surveys. The problem is ready-made, however, for the techniques
of high-altitude photography.
VII - 62
-------
ERTS-1 Imagery
ERTS-1 passes every eighteen days over Delaware Bay at an altitude
of about 500 miles, imaging the area with a Multispectral Scanner
(MSS). The MSS has four spectral bands as follows: band 4 (0.5 -
0.6 microns), band 5 (0.6 - 0.7 microns), band 6 (0.7 '- 0.8 microns),
and band 7 (0.8 - 1.1.microns). Since its launch on July 23, 1972,
the satellite has made fifteen passes over the-Delaware Bay test site,
four of which occurred on days with exceptionally good visibility, and
produced imagery which is nearly free of clouds and ground haze. The
resulting pictures are shown in Figures 2,3, 4, and 5, consisting of
MSS band 5 images.taken on October 10 and December 3, in 1972, and on
January 26 and February 13 in 1973. (NASA-ERTS-1 I.D. Nos. 1079-15133,
1133-15141, 1187-15140, and 1205-15141 respectively.) Each time, the
images were taken at about 15:14 G.M.T. or 10:14 a.m. local time,
resulting in low solar altitudes from 23° to 39°, a condition favoring
visibility of water features by avoiding sunlight directly reflected
off the water surface. Bands 4 and 6 of the October 10th frame are
shown for comparison in Figure 2
VII - 63
-------
Optical Properties of Suspended Sediments
Extensive investigations of suspended sediments in Delaware Bay and
laser transmission tests in a test tank facility have been conducted
respectively by (Oostdam, 1970) and Hickman, 1972). The results can be
summarized as follows:
0 suspended sediments in Delaware Bay averaged 30 ppm. During
July-August the average sediment level was 18 ppm.
0 turbidity increased with depth in the water column, except
during periods of bloom, when surface turbidities at times
exceeded those at greater depths.
0 suspended sediment concentration gradients were greater during
ebb than during flood because of greater turbulence and better
mixing during flood stage.
0 the turbidity decreased from winter to summer.
° marked increases in turbidity which were observed during May
and September were caused mainly by plankton blooms.
0 suspended sediments were silt-clay sized particles with mean
diameters around 1.5 microns.
0 the predominant clay minerals are chlorite, illite and kaolinite.
0 reflectivities for the Delaware Bay sediments were measured to be
about 10%.
At the time of the ERTS-1 overpasses, Secci depth readings ranged
from about 0.2 meters near the shore up to about 2 meters in the deep
channel. Preliminary "equivalent Secci depth" measurements with green
and red boards indicated that neither "color" exceeded the readings
obtained with the white Secci disc. Therefore it is quite unlikely
VII - 64
-------
that the bottom will be visible in any of the ERTS-1 channels and, at
least in Delaware Bay, most of the visible features will be caused by
light reflected off the surface or backscattered from suspended
matter.
"Red" filters, such as the Kodak Wratten No. 25A, have frequently
been used in aerial photography to enhance suspended sediment patterns
(Klemas et a_l. , 1973, Bowker et^ _al. , 1973). In Delaware Bay red filters
have been particularly effective for discriminating light-brown sediment-
laden water in shallow areas from the less turbid dark-green water in
the deep channel. For comparison, photographs taken by a U-2 plane in
the green, red and near-infrared bands during the December 3, 1972,
ERTS-1 overpass are shown in Figure 6. Note that even from 65,000 feet
altitude, suspended sediment patterns photographed in the green band
result in poorer contrast than the imagery obtained with a red filter.
To the ERTS-1 >tultispectral Scanner suspended sediment
patterns also appear most distinct in band 5, the "red"
channel, in agreement with aircraft results. Band 4 displays a more
complex pattern of suspended matter, which is further aggravated by a
masking effect resembling scattering by the atmosphere or haze. Due to
its limited water penetration, the band 6 picture shows
only weak patterns of suspended sediment near the surface.
Figure 7 contains microdensitometer scans between Cape Henlopen and
Cape May at the mouth of the bay. ERTS-1 images taken in bands 4,5, and
6 were scanned, and grey scales equalized, to enable comparison on the
same set of coordinate axes. Although suspended sediment is most clearly
visible in MSS band 5, the sediment patterns are caused by only three
VII - 65
-------
to four neighboring shades of grey. In agreement with other investi-
gators (Ruggles, 1973 and Bowker e_t_ _al. , 1973), we found that when the
sought features were in the first few steps of the grey scale, it
was best to use the negative transparencies. When the analysis
was performed above the first few steps of the grey scale, positive
transparencies proved somewhat more useful.
Turbidity, salinity, temperature and sediment concentration have
been measured from boats along the same transsect between the capes
as the microdensitometer scans. Preliminary comparisons indicate
that for the upper one meter of the water column, band 5 (red band)
correlates more closely with turbidity and sediment load measured
from boats,' than either band 4 or band 6. (Figure 8). Consequently, ih
our discussion of circulation patterns, we will emphasize imagery
obtained in band 5.
VII - 65
-------
Suspended Sediment and Circulation Patterns
Since suspended sediment acts as a natural tracer, it is possible
to study gross circulation in the surface layers of the bay by employing
ERTS-1 imagery and predicted tide and flow conditions. In addition to
ERTS-1 pictures, Figures 2, 5, 4, and 5 contain tidal current maps for
Delaware Bay. Each ERTS-1 picture is matched to the nearest predicted
tidal current map within + 30 minutes. A closer match was not attempted
at this point, since quantitative comparison would require compre-
hensive current measurements over the entire bay at the time of each
satellite overpass — a feat not attainable with our limited resources.
The current charts indicate the hourly directions by arrows, and the
velocities of the tidal currents in knots. The Coast and Geodetic
Survey made observations of the current from the surface to a maximum
depth of 20 feet in compiling these charts.
The satellite picture in Figure 2 was taken two hours after maximum
flood at the entrance of Delaware Bay on October 10, 1972. The sediment
pattern seems to follow fairly well the predicted current directions.
A strong sediment concentration is visible above the shoals near Cape
May and in the shallow nearshore waters of the bay. Peak flood velocity
is occurring in the upper portion of the bay, delineating sharp shear
boundaries along the edges of the deep channel. At the time of this
ERTS-1 picture, the wind velocity was 7 to 12 miles per hour from the
north.
Figure 3 represents tidal conditions two hours before maximum
VII - 67
-------
flood at the mouth of the bay observed by ERTS-1 on January 26, 1973.
High water slack is occurring in the upper portion of the bay, resulting
in less pronounced boundaries there as compared to Figure 2. The shelf
tidal water is not rushing along the deep channel upstream anymore as in
Figure 2, but is caught between incipient ebb flow coming down the upper
portion of the river and the last phase of flood currents still entering
the bay. The sediment plume directions in Figure 4 seem to show flood
water overflowing the deep channel arid spreading across the shallow
areas towards the shore. On the morning of December 3, 1972, there was
a steady wind blowing over the bay at 7 to 9 miles per hour from the
west.
As shown in Figure 4, on December 3, 1972, ERTS-1 passed over
Delaware Bay one hour before maximum ebb at the mouth of the bay. In
addition to locally suspended sediment over shallow areas and shoals
observed in Figures 2 and 3 plumes of finer particles are seen in
Figure 4''parallel to the'river flow and exiting from streams and inlets
of New Jersey's and Delaware's coastlines. Shear boundaries along the
deep channel are still visible in the upper portion of the bay; however,
they are beginning to disappear as slack sets in, resembling conditions
in Figure 3. U-2 photographs taken from 65,000 feet 41 minutes after
the ERTS-1 overpass are shown in Figure 6. The wind was variable in
speed and direction. At the time of the overpass, it was from the south
at about 4 miles per hour.
The satellite overpass on February 13th, 1973 occurred about one
hour after maximum ebb at the capes. The corresponding ERTS-1 image and
VII - 68
-------
predicted tidal currents are shown in Figure 5. Strong sediment trans-
port out of the bay in the upper portion of the water column is clearly
visible. Sediment boundaries such as the one in Figure 5 are frequently
observed near Cape Henlopen. Changes in Secci depth from about 0.5
meters to 1.1 meters were observed from boats crossing the boundaries
toward the less turbid side. The wind velocity at the time of the
satellite overpass was about 7 to 9 miles per hour from the north-northwest.
VTI - 69
-------
Water Boundaries and Fronts
Boundaries or fronts (regions of high horizontal density gradient
with associated horizontal convergence) are a major hydrographic feature
in Delaware Bay and in other estuaries. Fronts in Delaware Bay have
been investigated using STD sections, dye drops and aerial photography.
Horizontal salinity gradients of 4% in one meter and convergence velocities
of the order of O.lm/sec. have been observed. Several varieties of fronts
have been seen. Those near the mouth of the bay are associated with
the tidal intrusion of shelf water (Figure 9). The formation
of fronts.in the interior of the bay appears to be associated
with velocity shears induced by differences in bottom topography
with the horizontal density difference across the front influenced by ver-
tical density difference in the deep water portion of the estuary (Kupfer-
man, IClemas, Polis, Szekielda, 1973). Surface slicks and foam collected
at frontal covergence zones near boundaries contained concentrations of
Cr, Cu, Fe, Hg, Pb, and Zn higher by two to four orders of magnitude than
concentrations in mean ocean water. (Szekielda, Kupferman, Klemas, Polis,
1972) Figure 10, (Band 5, I.D. Nos. 1024-15073) obtained by ERTS-1
on August 16, 1972 contains several distinct boundaries. The southern-
most boundary, as shown in Figure 11 is of particular interest. Since
it has frequently been observed from aircraft (Figure 12). At the
time of the ERTS-1 overpass, divers operating down to depths of 6 meters
noted increases in visibility from about half a meter to two meters
as the boundary moved past their position .
VII - 70
-------
Waste Disposal plumes
Careful examination of the ERTS-1 Image of January 25, 1975,
shown in Figure 15, disclosed a plume 36 miles east of Cape Henlopen
caused apparently by a barge disposing acid wastes. The plume shows
up more strongly in the green band than in the red band. Since some
acids have a strong green component during dumping and turn slowly-
more brownish-reddish upon contact with seawater, the ratio of
reflectance signatures between the green and red bands may give an
indication of how long before the satellite overpass the acid
was dumped.
VII - 71
-------
Image Enhancements
Color density slicing and optical additive color
viewing techniques were employed to enhance the suspended
sediment patterns. Grey tone variations were "sliced" into
increments and different colors assigned to each increment
by using the Spatial Data Datacolor 703 system at NASA's
Goddard Space Flight Center. MSS band 5 clearly contains
more density steps than bands 4 and 6. Density slicing
emphasizes the difference between tidal conditions on
October 10th, 1972, and January 26th, 1973.
Additive color composites of bands 4 and 5 of the
October 10th overpass using a photographic process of silver-
dye bleaching. This process bleaches out spectral separations
of each MSS band to produce the color composite. The additive
color rendition is then reproduced on Cibachrome CCT color
transparencies. Comparison of the composite with the
equivalent band 5 image indicated that the composite does
not contain more suspended sediment detail than the individual
band 5 image. For similar reasons, composites prepared with
the International Imaging Systems Mini-Addco"Additive Color Viewer,
Model 6030 did not improve contrast beyond what was attainable
in band 5 directly.
{The color composites could not be included in this
report. Interested parties should contact the author).
Acknowledgements
This project is partly funded through ONR Geography
Programs, Contract No. N00014-69-A0407; and NASA-ERTS-1,
Contract No. NAS5-21837.
VII - 72
-------
Conclusions
a) ERTS is a suitable platform for observing suspended sediment
patterns and water masses synoptically over large areas.
b) Suspended sediment acts as a natural tracer allowing photo-
interpreters to deduce gross current circulation patterns from
ERTS-1 imagery.
c) Under atmospheric conditions encountered along the East Coast
of the United States MSS band 5 seems to give the best representation
of sediment load in the upper one meter of the water column. Band 4
is masked by haze-like noise, while band 6 does not penetrate suf-
ficiently into the water column.
d) In the ERTS-1 imagery the sediment patterns are delineated by
only three to four neighboring shades of grey.
e) Negative transparencies of the ERTS-1 images give better contrast
whenever the suspended sediment tones fall within the first few steps
of the grey scale. Considerable improvement in contrast can be ob-
tained by more careful development of film and prints.
f) Color density slicing helps delineate the suspended sediment pat-
terns more clearly and differentiate turbidity levels.
g) Sediment pattern enhancements obtained by additive color viewing
of the four ERTS-1 MSS bands did not noticeably improve the contrast
above that seen in the best band, i_.e_., MSS band 5.
h) Water!)boundaries containing high concentrations of toxic substances
identified in several of the ERTS-1 frames.
i) ERTS-1 bands 4 and 5 of Oct. 10, 1972, contain a clearl/ visible acid
disposal plume 36 miles east of Cape Henlopen.
VII - 73
-------
References
Bowker, D. E., P. Fleischer, T. A. Gosink, W. J. Hanna and J. Ludwich,
Correlation of ERTS Multispectral Imagery with Suspended Matter
and Chlorophyll in Lower Chesapeake Bay, paper presented at
Symposium on Significant Results Obtained from ERTS-1, NASA
Goddard S.F.C., Greenbelt, Maryland, March 5-9', 1973.
Harleman, D. R. F., Tidal Dynamics in Estuaries, Estuary and Coastline
Hydrodynamics, ed., A. T. Ippen, McGraw-Hill, Inc., New York,
1966.
Ketchum, B. H., The Distribution of Salinity in the Estuary of the
Delaware River, Woods Hole Oceanographic Institute, Ref. No.
52-103, 38 pages, 1952.
Klemas, V. , W. Treasure and R. Srna, Applicability of ERTS-1 Imagery
to the Study of Suspended Sediment and Aquatic Fronts, paper
presented at Symposium on Significant Results Obtained from
ERTS-1, NASA-Goddard S.F.C., Greenbelt, Maryland, March 5-9, 1973.
Klemas, V., R. Srna and W. Treasure, Investigation of Coastal
Processes Using ERTS-1 Satellite Imagery, paper presented at
American Geophysical, Union Annual Fall Meeting, San Francisco,
California, Dec. 4-7, 1972.
Kraft, J. C., A Guide to the Geology of Delaware's Coastal Environment,
College of Marine Studies Publication, University of Delaware,
1971.
Kupferman, S. L., V. Klemas, D. Polis and K. Szekielda, Dynamics of
VII - 74
-------
Aquatic Frontal Systems in Delaware Bay, paper presented at A.G.U.
Annual Meeting, San Francisco, California, April 16-20, 1973.
Nat. Aer. Space Admin., Data Users Handbook, Earth Resources Technology
Satellite, GSFC Document 71504249, 15 September, 1971.
Oostdam, B. L., Suspended Sediment Transport in Delaware Bay, Ph.D.
Dissertation, University of Delaware, Newark, Del., May, 1971.
Ruggles, F. H., Plume Development in Long Island Sound Observed by
Remote Sensing, paper presented at Symposium on Significant Results
Obtained from ERTS-1, NASA Goddard S.F.C., Greenbelt, Maryland,
March 5-9, 1973.
Sherman, J., Comment made at NASA Marine Resources Working Group
Meeting, GSFC, Greenbelt, Maryland, March 9, 1973.
Szekielda, K. H., S. L. Kupferman, V. Klemas and D. F. Polis, Element
Enrichment in Organic Films and Foam Associated with Aquatic
Frontal Systems, Journal of Geophysical Research, Volume 77,
No. 27, September 20, 1972.
U. S. Department of Commerce, Tidal Current Charts - Delaware B,ay and
River, Environmental Science Services Administration, Coast and
Geodetic Survey, Second Edition, 1960.
U. S. Department of Contaerce, Tidal Current Tables, Atlantic Coast of
North America, National Oceanic and Atmospheric Administration,
National Ocean Survey, 1972 and 1973.
Hickman, G. D., J. E. Hogg and A. H. Ghovanlou, Pulsed Neon Laser
Bathymetric Studies Using Simulated Delaware Bay Waters.
Sparcom, Inc., Technical Report #1, Sept. 1972, pp. 10-13.
VII - 75
-------
FIGURE
SUBMERGED CONTOURS OF DELAWARE BAY
TONGUES OF DEEPER WATER RADIATE
FROM THE BAY ENTRANCE INTO THE BAY.
Miami
• ••ch
Clark Point
Ston*
Delaware B
-------
/""**! / ^- Urf' "*'" '"*' "*"'" '"• '"•
C. . I.' J25 rjr-l 0 C I 1.01.1 «!...!,< Co..1 .. No.,h
TWO HOURS AFTfR MAXIMUM HOOD AT DELAWARE BAY ENTRANCE
Figure 2 - ERTS-1 image of Delaware Bay obtained in MSS band 5 on October 10, 1972,
and tidal current map. (I.D. No. 1079-15133.)
VII - 77
-------
TIDAL
-------
ONI HOUR BffOKC MAXIMUM i 86 AT OCLAWARf BAr tNTRANC€
Figure 4
VII - 79
-------
IIDAL CURRENT CHART
DKL-AWARK BAY AND HIVBB
•lA^""^t' '< vV
£^to^
-
- ^
X. " V 1 "•
* /»'., \ ?-eJ. <•
ON[ HOUR AltIR MAXIMUM [ BB AT DIIAWAHt HAY INIUANi'l
Figure 5 - ERTS-1 image of Delaware Bay obtained in MSS band 5 on February 13, 1973,
and tidal current maps. (I.D. Nos. 1205-15141).
VII - 80
-------
a) 0.475—0.575 microns.
W 0.580—0.680 microns.
FIG. 6. Photographs taken at 65.000 feet altitude by a
NASA U-2 aircraft. 41 minutes after the December 3, 1972
ERTS 1 overpass. The mouth of Delaware Bay is shown in
three spectral bands.
c) 0.690—0.760 microns
VII - 81
-------
JlTV
"
i^^
VV\
A
/
\
vs
WU. f
- l-~^rJ 'wv
J
f
n
,/
/
y
^^
BAND t
BAND ft
BAfO 4
'
-CAPE HENLDPEN
CAPE MAY
it
20
co
LJ
O
10
CAPE HENLOPEN -CAPE MAY
- DISTANCE
DISTANCE
Positive trans-
parencies.
Negative trans-
parencies.
FIGURE 7
Microdensitometer traces of October 10, 1972, ERTS-1 imagery from Cape
Henlopen, Delaware to Cape May, New Jersey, using MSS bands 4, 5, and 6,
VII - 82
-------
3456
8 9
10
12
13
14
STATION
Figures Correlation between ERTS-1 Image Radiance,
Suspended Sediment Concentrations, and Secchi Depth.
SEDlMENT(lm) /^SEDIMENT
(surface)
/ M1CRODENSITOMETER
SECCH! DEPTH
(cm)
oo
e
H
Q
fcj
M
X
o
o
tu
•1.66
•1.64
1.62
•1.60
1.58
1.56
••154
•-I.52
50
-40
--30
20
10-
0
-•10
-15
--20
--25
--30
•-35
--40
•45
-1-50
-------
Figure 9 Frontal system caused by higher salinity shelf
water intruding into Delaware Bay.
Figure 12 Aerial photograph from 9000 feet altitude of foam-
line at houndary between two different water masses off Delaware's
coast.
VII - 84
-------
Figure 10 - ERTS-1 image of the mouth of Delaware Bay showing
several water mass boundaries and high concentrations
of suspended sediment in shallow waters. (Band 5,
August 16, 1972, I.D. Nos. 1024-15073).
VII - 85
-------
DELAWARE BAY
CAPE HENLOPEN
ATLANTIC OCEAN
Figure 11 Aquatic boundaries and suspended sediment plumes identified
in the ERTS-1 image of August 16, 1972, shown in Figure 14.
VII - 86
-------
Figure 13 - Acid waste dumped from barge about 40 miles
east of Indian River Inlet appears clearly as a fishhook
shaped plume in MSS band 4 image of January 25, 1973.
VII - 87
-------
DATA COLLECTION PLATFORMS FOR
ENVIRONMENTAL MONITORING
J. Earle Painter
NASA/Goddard Space Flight Center
ABSTRACT
The Earth Resources Technology Satellite (ERTS) Data Collection
Platform (DCP) Network is in operation with installations
extending from Iceland, to Hawaii and from Alaska to Honduras.
Thirty investigators have deployed 125 DCP's in twenty-five
states and five foreign countries. These installations are
used for in-situ monitoring in the areas of meteorology,
hydrology, agriculture, vulcanology and water quality.
Water quality installations in prototype operational networks
are supplying data to regulatory agencies on a daily or more
frequent basis. The failure of some installations to perform
adequately points out a definite need for further sensor
development and improved implementation techniques. The ERTS
experiment has demonstrated the feasibility and reliability
of space relay systems for automatic collection of data from
in-situ environmental sensors and indicates a promise of cost
effective operation of such systems.
VII - 88
-------
INTRODUCTION
The Data Collection System (DCS) flown on the Earth Resources
Technology Satellite (ERTS) is the first attempt to establish
a large scale telemetry network using a satellite to relay
data from in-situ environmental sensors to a central collection
point. After more than one year of operation, the attempt has
proven successful.
SYSTEM DESIGN GOALS
Design goals for the DCS (Table 1) were established by-consulting
with appropriate personnel in those federal agencies which have
environmental monitoring responsibilities. They include
reliability, versatility, and low cost as primary requirements.
These goals were all met or exceeded"by the system during the
past 14 months of operation (Table 2).
VII - 89
-------
TABLE 1
ERTS DCS DESIGN GOALS
SYSTEM PARAMETERS
ONE GOOD TRANSMISSION EVERY TWELVE HOURS
LOW ERROR RATE
RAPID DATA DELIVERY
VERSATILE TRANSMISSION FORMAT
GOOD RADIOINTERFERENCE IMMUNITY
PLATFORM (TRANSMITTER) PARAMETERS
LOW COST
LOW POWER REQUIREMENTS
RELIABLE OPERATION
VERSATILE SENSOR INTERFACE
VII -90
-------
TABLE 2
ERTS DCS PERFORMANCE
SYSTEM PARAMETERS
GOOD TRANSMISSIONS EACH TWELVE HOURS
ERROR RATE (TO USER)
DATA DELIVERY (TELETYPE)
TRANSMISSION FORMAT
RADIOFREQUENCY IMMUNITY
6 to 8
< 10"5
30 MINUTES
SATISFIES USERS
EXCELLENT
PLATFORM (TRANSMITTER) PARAMETERS
COST (200 UNIT PROCUREMENT)
POWER DISSIPATION
FAILURE RATE
SENSOR INTERFACE VERSATILITY
$2,500
50 MILLIWATTS
2% PER MONTH
SATISFIES USERS
VII - 91
-------
SYSTEM DESCRIPTION
The system (Fig. 1) consists of a data formatting and trans-
mitting unit, called the Data Collection Platform (DCP), a
receiver and a retransmitter aboard ERTS-1; and receiving,
demodulating and decoding equipment located at the Goldstone,
California and Goddard data acquisition stations. Data is
transmitted from the data acquisition stations to the ERTS
Control Center at Goddard, then to the NASA (ERTS) Data
Processing Facility (NDPF) where it is processed and
distributed to Users.
Every three minutes a sample of data from each of as many
as eight sensors is accepted by the DCP in either analog or
digital form. The DCP converts the analog data to digital
form, adds a DCP identification number, encodes the data for
error control purposes, and transmits it skyward. If the
spacecraft is in range of both the DCP and one of the receive
sites, the data is received and retransmitted by the space-
craft and received by the ground station where it is
demodulated, decoded and forwarded to the ERTS Operations
Control Center (OCC) at Goddard.
In the Control Center, the DCS data is screened for quality
and retransmitted to Users by teletype.
VII - 9-2
-------
ERTS-I
Goldstone
H
H
U)
DCP Installation
Pig. 1 ERTS DATA COLLECTION SYSTEM
-------
THE OPERATING NETWORK
A 125 DCP network is now in operation with installations
extending from Alaska to Honduras and from Iceland to
Hawaii (Fig. 2). DCP's are transmitting from locations in
22 states and five foreign countries (Table 3).
A total of 29 Users are currently involved in the program
(Table 4) representing six federal agencies, one state, one
foreign country, four universities, and one industrial firm.
These investigators are using the system for eight major
application categories (Table 5). The most active
organization is the U. S. Geological Survey, accounting for
one-third the Users and one-half the DCP's. Hydrology is
the primary application. Volcanology is second closely
followed by water quality. Table 6 is a partial list of
parameters monitored by the DCS installations.
An individual DCP in the network is contacted during two to
seven orbits per day, depending upon location (latitude and
proximity to a data acquisition station). One to four
messages are received during each contact period averaging
17 good messages per day.
VII - 94
-------
H
H
I
Ul
80
60 -
40 _
20 _
0 _
160 140
120
100
20
Fig. 2 ERTS-1 J)CP NETWORK
-------
TABLE 3
GENERAL LOCATION OP ERTS-1 DCP'S
22 STATES
H
H
ALABAMA
ALASKA
ARIZONA
CALIFORNIA
CONNECTICUT
DELAWARE
FLORIDA
HAWAII
KANSAS
LOUISIANA
MARYLAND
MASSACHUSETTS
MI GUI CAN
MISSISSIPPI
NEW HAMPSHIRE
OHIO
OREGON
PENNSYLVANIA
TENNESSEE
VERMONT
VIRGINIA
WASHINGTON
5 FOREIGN COUNTRIES
CANADA
EL SALVADOR
GUATEMALA
ICELAND
NICARAGUA
-------
TABLE 4
ORGANIZATIONS USING ERTS-1 DCS
H
H
vo
-j
ORGANIZATION
POREIGN (CANAPA)
U.S. GEOLOGICAL SURVEY
BUREAU OP LAND MGMT.
FORESTRY SERVICE
CORPS OP ENGINEERS
NAVOCEANO
UNIVERSITIES
STATES
NASA
INDUSTRY"
TOTALS
NO. INVESTIGATORS
6
10
1
1
1
1
4
1
3
1
29
ASSIGNED DCP'S
H
106
8
3
30
3
18
1
29
2
214
ACTIVE DCP»S
12
58
3
3
21
0
3
1
4
1
106
-------
TABLE 5
USE OP ERTS-1 DCS BY APPLICATION
S
H
00
APPLICATION
METEOROLOGY
HYDROLOGY
WATER QUALITY
OCEANOGRAPHY
FORESTRY
AGRICULTURE
VOLCANOLOGY
ARCTIC ENVIRONMENTS
NO. OP USERS
5
20
4
J>
1
1
2
1
DCP'S ASSIGNED
10
75
26
9
3
3
33
2
-------
TABLE 6
PARAMETERS MONITOREB BY ERTS-1 DCS
H
H
RESERVOIR LEVEL
STREAM PLOW
GROUND WATER
TIDAL VARIATION
ICE CONDITIONS
SALINITY
DISSOLVED OXYGEN
TURBIDITY
ACIDITY-ALKALINITY
BIOLOGICAL CONTENT
WATER TEMPERATURE
AIR TEMPERATURE
WIND DIRECTION
WINDSPEED
HUMIDITY
PRECIPITATION
SOLAR RADIATION
SNOW DEPTH
SNOW WATER CONTENT
EVAPORATION
SOIL MOISTURE
EARTH TILT
TREMOR EVENTS
EARTH TEMPERATURES
-------
WATER QUALITY INSTALLATIONS
The ERTS DCS installations of most acute interest to
regulatory agencies are those which monitor water quality
parameters. Several organizations are using such installations
with mixed results.
VII - 100
-------
DELAWARE BASIN WATER QUALITY NETWORK
The U. S. Geological Survey, Water Resources Division,
Pennsylvania District operates a network of twelve water
quality installations in the Delaware River Basin using
ERTS to relay the data (Pig. 3). These data are relayed
through ERT.S-1 to Goddard Space Center, then forwarded via
teletype to the USGS offices in Harrisburg, Pennsylvania^
where they are processed and distributed daily to the
Delaware River Basin Commission and other regulatory
agencies.
All the Delaware Basin Installations are single line
pumping configurations with a submerged pump located in a
stream-side well and an instrumentation assembly in a
nearby shed. Equipment from several manufacturers is used
and includes four sensors in each installation to monitor
temperature, pH, conductivity, and dissolved oxygen.
Thermocouples are used for temperature and give little
trouble. The glass pH probe with associated referende probe
are also relatively trouble free once installed properly.
Most sensor problems in the array are a result of coating
of the one mil teflon membrane in the D.O. monitor arid the
presence of conducting material on and around' the four probe
conductivity sensor.
VII - 101
-------
Pig. 3 ERTS DATA RELAY
DELAWARE RIVER BASIN NETWORK
Del. River near
E. Stroudsburg,
PA.
Del. River at
Easton. PA
42
3. Lehigh River at
Easton, PA
4. Del. River at
Trenton, N. J.
5. Del. River at
Bristol, PA
6. Del. River at
Torresdale,
Phila. PA
7. Del. River at
Ben Franklin
Bridge
8. Schuylkill River
at"Philadelphia,
PA
9. Del. River at
Chester, PA
10. Del. River at
Del. Memorial
Bridge 4C
11. Del. River at
Reedy Island
Jetty, Delaware
12. Del. River at
Ship John Shoal
Lighthouse
VII - 102
-------
In addition to the degradation and failure of sensors, the
amplifiers and other electronics associated with the sensors
are failure prone. Some of these failures are caused by
handling and lightning strikes, of course. The submersible
pumps are relatively trouble free, lasting six months to a
year.
With periodic maintenance performed at one to three week
intervals, about half the stations produce data at any one
time. This low level of operational efficiency is unfortunate,
but is tolerable for the time being since it provides data not
otherwise available.
VII - 103
-------
CORPS OF ENGINEERS NSW ENGLAND EXPERIMENT
The Array Corps of Engineers, New England Division deployed
three water quality DCS installations on the North Nashua
River at Pitchburg, Mass., the Chicopee River at Chicopee,
Mass., and the Westfield River at West Springfield, Mass.,
(Pig. 4). These installations employ basically the same
systems as the USGS Delaware Basin sites.
The Chicopee and Westfield sites have performed reliably,
though no definitive check on data quality was available at
this writing. These sites have operated for several months
with maintenance performed every two to three weeks. The
apparent success of these two sites is attributed to the
relative cleanliness of the stream sites involved.
Unfortunately, these sites are strictly experimental.
The third site at Fitchburg, Mass, (not illustrated) was
intended to provide baseline data prior to proposed construc-
tion of low flow augmentation (pollution dilution) reservoirs,
Parkinson's law holds and viscuous material from upstream
paper mills keep clogging pump screens and burning out pumps.
The site will be abandoned.
VII - 104
-------
Pig. 4
CORPS OF ENGINEERS EXPERIMENT
42*
QuaDDin
Reservoir
Northampton
CMcopee
River
Westfield
River
Westfield
West
Springfield
Springfield
Massachusetts
Connecticut
Connecticut
River
Hartford
VII -
-------
NASA- WALLOPS STATION EXPERIMENT
Within the last few months, the NASA-Wallops Station in
Virginia has deployed two DCS water quality systems in
support oi% experiments conducted "by the Corps of Engineers,
Vicksburg, Mississippi, office. The sites chosen are
near-shore locations in the Chesapeake Bay; one at the
mouth of the Rappahannock and the other near Oxford,
Maryland (Pig. 5).
Instrumentation of the NASA sites consists of submersible
instrument clusters mounted on pilings with the electronics
and DCS antenna mounted at the top of the piling out of
•range of normal wave action. Sensors are included for
conductivity, temperature, dissolved oxygen, pH, and
turbidity.
The turbidity sensor, which is a light source and photocell
device, degrades at a rapid rate with the adhesion of foreign
matter on instrument surfaces in the light path. Data from
this instrument is of little value after a week or so of
service.
The other sensors accumulate surface coatings but produce
good quality data after 18 days of use when tested in a
VII - 106
-------
Fig. 5 NASA - Wallops Experiments
,o
Annapolis
Washington
Choptank
River «
Rappahannock
River
VII - 107
-------
calibration tank pri©r to cleaning. A thick algae coating
on the D. 0. membrane had little effect.
The Rappahannock site was primarily affected by algae
aceemulation; the Oxford site by silt deposits. In these
environments, the turbidity sensor was clearly inadequate.
Additional experimentation is required to determine the
required service interval for the other sensors.
vii - 108
-------
CONCLUSIONS
The deployment- of large, automated networks to acquire
environmental data from surface monitoring stations requires
field equipment capable of many weeks of reliable unattended
operation. The ERTS-I Data Collection System experiment
illustrates a pronounced failure of pollution sensing systems
to meet this requirement. Indeed, they are the weakest
element in the experiments.
Clearly, a concerted, national effort is required to develop
pollution sensors and to approach installation and network
designs from an overall system point of view.
VII - 109
-------
ACKNOWLEDGEMENTS
The author, who is not an expert in pollution monitoring and
water systems regulation, gratefully achknowledges contributions
from the following people who are:
Mr. Saul Cooper and Mr. Joe Horowitz
U. S. Army Corps of Engineers, New England Division
Mr. Richard Paulson and Kr. Charles Merck
U. S. Geological Survey, Water Resources Division,
Pennsylvania District.
Mr. Duane Preble
U. S. Geological Survey, Gulf Coast Hydroscience Centea
Mr. Roger Smith
National Aeronautics and Space Administration
Wallops Station
VII - 110
-------
SESSION VIII
ENVIRONMENTAL MONITORING APPLICATIONS
CHAIRMAN
MR. ROBERT F. HOLMES
OFFICE OF MONITORING SYSTEMS, OR&D
-------
AIRCRAFT and SATELLITE MONITORING
of LAKE SUPERIOR POLLUTION SOURCES
By
James P. Scherz
Associate Professor,
Civil and Environmental Engineering Department
University of Wisconsin, Madison
and
John F. Van Domelen
Research Assistant,
Institute for Environmental Studies, and Ph.D. Candidate
University of Wisconsin, Madison
A paper presented at the U.S. Environmental Protection Agency's Second
Environmental Quality Sensor's Conference.
October 10-11, 1973
Las Vegas, Nevada
VIII - 1
-------
BIOGRAPHICAL SKETCHES
JAMES P. SCHERZ, born 1937, obtained B.S. and M.S. degrees in Civil
Engineering from the University of Wisconsin in 1959 and 1961. He has worked
with the Highway Commission and on the state airborne magnetic survey of
Wisconsin, and 4 years with the U.S. Army Corps of Engineers as a Regular
Army Officer. He left the service in 1967 and began research at the Univer-
sity of Wisconsin on photographing water pollution and obtained his PhD in
1967. Since then, he has taught in the photogrammetry and remote sensing area
and has been active in research work and consulting on remote sensing of water
quality.
JOHN F. VAN DOMELEN, born 1942, obtained a B.S. in Applied Physics from
Michigan Technological University in 1964, served 5 years as a Photo and Radar
Intelligence Officer in the Regular USAF followed by a year as manager with a
paper products company. He obtained an M.S. in Water Resources Management
from the University of Wisconsin in 1972 and is presently engaged in a PhD
program, in Civil and Environmental Engineering which he expects to complete
in May, 1974. His major focus is in remote sensing determination of water
pollution parameters.
ABSTRACT
Aerial and Satellite Photography can be a very valuable tool for water
quality investigations. Such aerial imagery correlates with the water quality
parameter of turbidity, which in turn under specific circumstances, correlates
to other water quality parameters such as suspended solids or apparent color.
The advantage of aerial imagery is that it provides an overall view of tur-
bidity possible by no other means. One striking example of this potential use
is in Lake Superior where an $8,000,000 water intake was located in turbid,
unpotable water. The turbidity and its location was clearly visible on aerial
and ERTS imagery. Another example of its potential use is where apparent
reflectance of various lakes in northern Minnesota as obtained from ERTS
imagery correlates very well with the turbidity, solids, and the U.S. Forest
Service classification of eutrophication for these lakes. For correct analysis
of aerial imagery for water quality, one must understand light penetration into
VIII - 2
-------
water and the corresponding bottom effects, as well as sky light reflection
from the water surface. If these effects are understood and accounted for in
a workable and practical manner, aerial and satellite imagery can indeed be a
valuable tool for water quality investigations and should be used as such.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the following: Professor James L.
Clapp and the staff of the Institute for Environmental Studies, Environmental
Monitoring and Data Acquisition Group (EMDAG) for their administrative assis-
tance; Kenneth Piech of Cal Span Corporation for his help in the planning of
early work designed to determine the truth about depth penetration and bottom
effects; the city of Cloquet, Minnesota which provided the financing for much
of the field data acquisition relating to Lake Superior; to Dr. Kenneth
Holtje and the U.S. Forest Service for their support in gathering field data
near Ely, Minnesota; to Steven Klooster for the contributions he made towards
the development of a practical means of analyzing remote sensing imagery of
paper mill pollution, and to William L. Johnson for his help in the gathering
and analyzing of data obtained from the lakes near Ely, Minnesota and his help
in conducting settling tests on water obtained near Duluth, Minnesota. We
also wish to acknowledge support received under the National Aeronautics and
Space Administration Grant No. NGL 50-002-127 for Multidisciplinary Research
in Remote Sensing as Applied to Water Quality.
INTRODUCTION
Remote Sensing is a valuable tool which should be used by anyone working
with water quality.
A striking example is in Lake Superior near Duluth, Minnesota and
Superior, Wisconsin, where an $8,000,000 water intake, without filtration
facilities, was located in water that was turbid and unpotable over 53% of the
time. Numerous aerial photographs on file in Superior, Wisconsin showed
vividly this turbid water and its extent. These aerial photos were never used
in the location process and the intake was constructed in the turbid water. In
addition to aerial photography, ERTS imagery also shows this turbid water, its
extent, and its movements. (13)
VIII - 5
-------
Not only is turbid water shown quantitatively on these aerial images
but a court-related investigation into the matter revealed that the water
quality parameter of turbidity correlates with the imagery at all times.
Other water quality parameters such as suspended solids sometimes can also
correlate with turbidity. To the extent that they do, they can also be mapped
by aerial photography. Similar results have been arrived at by looking at
industrial pollution sources and at the clearness of some lakes in various
stages of enrichment. (5)
In addition to the data gathered at Lake Superior relating to this
court case, this paper also presents data pertaining to industrial pollution
and lake eutrophication. These examples are included so as to more clearly
show how remote sensing can be effectively used as a quality monitoring tool
for various waters.
In all cases, "noise" effects, are present and must be properly dealt
with. These include: Surface reflection from the water surface, and bottom
effects (14). Scatter in the atmosphere, and effects from the camera lenses
and film development also contribute, but to a less significant degree. This
paper considers these effects and suggests how they can be understood and
dealt with so that remote sensing can be a valuable tool in water quality
investigations.
BASIC RELATIONSHIP
Aerial photos have long been effectively used by professional people
in areas such as forestry, agriculture, and soils. On the other hand aerial
photographs have not been effectively used by most practicing sanitary
engineers whether they are working with water pollution or locating drinking
water intakes. Perhaps one reason why photography has not been effectively
used by people actively engaged in the water quality and supply field is
because the interaction of the sun's energy, with a water body, is a complex
process and this process, in the past, has not been widely, nor fully under-
stood. However, it is now felt that these interactions are sufficiently
understood, so that aerial photographs can and should be used for water quality
investigations.
The sun's rays interact first with the water surface. A portion of
the light from the sun and the sky (including clouds) are reflected from this
-------
water interface (see Fig. 1). This reflected energy is called surface
reflection. A portion of the light enters the water, interacts with the mat-
erial in the water and is scattered upward to the camera. This is herein
called volume reflectance. The volume reflectance is the critical item, as
it indicates the quality of the water. In shallow, clear water, a portion of
the light may pass through the water and reach the bottom where it can be
reflected back through the water body and up to the camera. Those rays which
strike the bottom, pass back up through the water body, create bottom effects
and are undesirable noise in the system; unless one desires to map bottom
vegetation.
Because surface reflection, volume reflectance, and depth penetration
and bottom effects are all wavelength (color) dependent, the whole question
becomes even more complex. For example, the short wavelengths like (ultra-
violet and blue) are reflected primarily from the surface of the water, while
the longer ones are not. Infrared energy penetrates least into the water
while green energy penetrates the most; the light returning as volume re-
flectance is entirely dependent upon the type of material in the water. (12, 8)
The light reaching the camera is therefore, the sum of the surface,
volume, and bottom effect phenomenon. Normally one is only interested in the
volume effects and the other effects must be separated out. However, the
magnitude of surface reflection can be altered by oil spills and this then
becomes the desirable factor to monitor. In all other cases, the surface
reflections are. undesirable. Since sky-light is always reflected from the
water surface, and since the reflection from a cloud is much brighter than
from the blue sky, for water quality investigations one should use aerial
photography taken only on a clear day or on a uniformly overcast day in order
to avoid local hot spots of high surface reflection caused by scattered clouds.
The sun's reflection can be eliminated by simply picking an angle where sun
glitter does not show.
The depth penetration and bottom effects can present a problem de-
pending on the turbidity of the water. Since a secchi disc really is a field
means of determining turbidity, it can effectively be used to ascertain if the
bottom effects are significant. (15) The white secchi disc is lowered into
the water until it disappears from sight. The depth at which this occurs is
called the secchi disc reading. The aerial camera will essentially see no
further into the water than the secchi disc reading. (14) As a general rule,
VIII - 5
-------
Fig. 1. Components of light that the camera
captures caused by various interactions of light
on, in, and through the water.
A = Surface Reflection of the Sun
B = Atmospheric Scatter
C = Surface Reflection of the Sky
D = Volume Reflectance of the Water
E = Bottom Effects
VIII - 6
-------
if the reading is less than the depth to bottom, one can assume that no signi-
ficant amount of energy returns from the bottom to effect the aerial image.
Where bottom effects are present, it is still possible to analyze aerial photo-
graphs of the water if the infrared wavelengths are used because infrared
energy penetrates only a few inches into the water.
The central question therefore, is how are volume reflectance .effects
associated with water quality parameters, and how can aerial photographs be
used to obtain these parameters in a reliable manner?
EQUIPMENT
A. Aerial Cameras
The remote sensing program at the University of Wisconsin makes use of
9X9 inch mapping cameras where high metric accuracy is of utmost importance.
However, experience has shown that in most water quality investigations, the
location of a pollution plume to within a fraction of a foot is of little impor-
tance. It is the spectral response of the water that is crucial. For such work,
35 mm cameras are used, because of their economy and the simplicity of their
data storage and retrieval characteristics. (7) The data gathered for this
research was done with a two camera bank-using 35 mm single-lens reflex cam-
eras - normal color film used in one camera and color infrared red film was
used in the other. The normal color film has three layers sensitive to blue,
green and red energy. The color-infrared film has three layers sensitive to
green, red and infrared energy. Each layer, of each exposed film, can be
analyzed to ascertain the reaction between the different energies and the objects
during exposure. These films are analyzed on a microdensitometer which is
attached to a spectrophotometer. This will be referred to as a color micro-
densitometer. (4)
B. Microdensitometer
Figure 2 shows the color microdensitometer apparatus used to analyze the
film. With this apparatus it is possible to analyze the energy at wavelengths
or colors of the film. In other words, it is possible to measure the color
transmittance of the film at any point on the photo. (4)
C. Spectrophotometer
If the microscope part of the microdensitometer is replaced by a larger
telescopic head as in Figure 3, we have a spectrophotometer which can analyze
VIII - 7
-------
ELECTRONICS
AND
READOUT
FILTER
OPTIC
3=
PHOTO
TUBE
MICROSCOPE
COLLECTOR
MONOCHROMATOR
(DEFRACTION GRATING)
FILM
ILLUMINATING
LIGHT
Fig. 2. Block diagram of color microdensitometer
for analyzing film. The illuminating light directs
light through the film into the microscope collector.
Then a. fiber optic directs the light through a de-
fraction grating into a photo tube. In this manner
the intensity of any color can be analyzed for any
point on the film.
VIII - 8
-------
ELECTRONICS
AND
READOUT
TELESCOPIC
HEAD
PHOTO
TUBE
MONOCHROMATOR
(DEFRACTION GRATING)
ILLUMINATING
LAMP FOR
REFLECTANCE
WORK
BARIUM SULFATE
STANDARD PLACED
HERE FOR
STANDARD READING
FOR REFLECTANCE
1 gal.
WATER SAMPLE
2 FEET DEEP ^~
4" IN DIAMETER
\\l//
ILLUMINATING LAMP
FOR TRANSMITTANCE
WORK
Fig. 3. Spectrophotometer block diagram for analyzing
water samples.
VIII - 0
-------
the transmittance or reflectance of a water sample. Light is placed under
the sample tube so the rays penetrate through the tube, enabling one to com-
pare incidental energy to transmitted energy. When the light is placed such
that it shines down on the surface of the sample, energy is scattered back,
and one can measure the reflectance of the water. This reflectance is impor-
tant because it is essentially the volume reflectance that is desired in the
water analysis work already mentioned. To get the absolute value for reflec-
tance from a water sample, a reading is first taken on a barium-sulfate
standard of known reflectance, and the water reading is compared to it. The
ratio of the water reading to the reading of the barium-sulfate is herein
termed the volume reflectance of the water. Assumptions are that there is no
reflection of bright ceilings from the water surface and that the reflection
from the bottom of the sample tube is negligible.
In the field, a styrafoam panel of known reflectance is placed in the
water and photographed. When the microdensitometer is used to analyze the
film, readings are taken from the image of the panel and then from the image
of the water. The ratio of the readings from the water and the panel are
herein called apparent re fleetanc e. Apparent reflectance includes sky-light
reflection and is always larger than the laboratory derived volume reflectance.
For correct analysis of the photo, there must be no bottom effects present.
LAB WORK TO FIELD WORK CORRELATION
For aerial photography to be effectively used in water quality investi-
gations, there is a certain necessary balance between laboratory and field
work.
A. Water Quality Using Aerial Photos
Anyone observing a photograph of an industrial plant spewing waste into
a stream intuitively realizes that there is a relationship between the relative
brightness of the waste on the photograph and the concentration of that pol-
lution. He might ask, "What water pollution parameters correlate with the
photo and how is this correlation possible?" Figure 4 shows a waste discharg-
ing from a paper mill in Wisconsin. In the summer of 1972 hundreds of water
samples were taken within the plume. At the same time, an aircraft flew over-
head taking color and color infrared photographs. Standard reflection panels
were strategically placed throughout the plume. On the resulting photos, color
VIII - 10
-------
1
M
Fig. 4. Photograph of a waste discharging from a paper mill
-------
microdensitometer readings were taken of the standard reflection panels and
other readings were taken of water at the sampling points. Ratioing the two
yielded the apparent reflectance which included the sky-light reflection.
This process was repeated for many wavelengths and Figure 5 shows the resultant
apparent reflectant curves.
Figure 6 is a plot of apparent reflectance, at 0.55 microns, versus tur-
bidity and solids for various points within the plume shown in Figure 4. The
apparent reflectance was obtained from the photographs. Figure 6, also illu-
strates the relationship between the reflectance, as obtained in the labora-
tory, and values obtained for turbidity and solids. The higher values obtained
from the field occurred because the photographs also contain sky-light reflec-
tion, while the lab samples do not. Figure 6 emphasizes the fact that there is
indeed a straight line relationship between apparent reflectance, turbidity
and solids. For low concentrations of wastes, the bottom of the lab sample
tube begins to interfere which causes lab results for these lower concentra-
tions to be in error.
From the above, it appears that one needs to obtain only a few water
samples at the instant of an aerial overflight to make positive correlation
between the water and the photograph. A suggested method would be to take a
sample in the dirtiest part of the plume, a sample of the receiving water and
a sample in the middle of the plume. The apparent reflectance from these
samples can then be used to ascertain a field curve for mapping water quality
parameters. If desired, the dirty water can be diluted with the receiving
water to obtain a lab curve, similar to that of Figure 6. The difference
between the two is a measure of sky-light reflection and other variables.
there must be no bottom effects present in the field. If there are, then any
analysis must be made only in the longer infrared wavelengths such as 0.75 microns.
Unless the camera and film conditions, altitude and lighting conditions,
and cloud and atmospheric conditions are identical on both days, it is not
possible to use reflectance-turbidity curves from one day to analyze the water
conditions on a second day. Figure 7 shows three reflectance-turbidity curves
from the paper mill in Figure 4.
The curve for 8/19/71 was obtained from photography taken at 1800 feet
on a uniformly overcast day. One will note that high values of apparent re-
flectance are obtained due primarily to the high reflection of the clouds from
the water surface. The curve from 8/11/71 was also obtained from photography
VIII - 12
-------
100 -.
<#>
w
u
Z
IS
U
W
•J
t,
w
2
W
10 --
1.0--
O.l-i
Sample Point A
Blue
Green
Red
.4 .5 .6 .7
COLOR AND WAVELENGTH IN MICRONS
Fig. 5. Apparent Reflectance of water at various
points in the pollution plume shown in Fig. 4.
These values are obtained by analyzing color film
with the color microdensitometer and comparing the
reading on the water with that of a standard styra-
foam panel. The values of turbidity and suspended
solids associated with the water sample points are
as follows;
Water Sample
Point
A
B
C
D
E
Turbidity
JTU
110
75
40
18
11
Suspended Solids
mg/L
172
100
50
23
8
One will note that there is indeed a good correlation
between apparent reflectance and the values of turbidity
and suspended solids.
"Til -
-------
E-i
Z
w
u
w
U 2
lyp jt
e H
u u
W 2
W U
« W
D
A
W
a
100
TOTAL SUSPENDED SOLIDS, MG/L
0.1 1 10 100 1000 104 105 106
10..
1.0 ..
0.1 .._..
Field Conditions
Prom Analyzing
Photographs \
Laboratory
Conditions
Prom Analyzing
Samples With The
Spec tropho tome te r
10
100 1000 104 105 106
TURBIDITY, JTU'S
Pig. 6. Volume reflectance for laboratory samples
and apparent reflectance for field conditions plotted
against turbidity and total suspended solids. The
wavelength used was 0.55 Micron (color green). The
fact that the apparent reflectance from field conditions
is higher than the lab conditions is primarily due to
sky-light reflection. The aerial photos show sky-light
reflection, while the lab analysis does not.
VIII - K
-------
shot at 1800 ft., but the values are lower because the cloud cover was much
thinner and the surface reflection of the clouds was less. The reflectance -
turbidity curve for the photography taken at 2800 ft. is drastically different
than the one for 1800 ft. It is clear that changes in altitude, especially
on a hazy day are extremely important.
Therefore with imagery obtained by aircraft where the altitude, time,
and cloud conditions vary, if one wishes to obtain absolute values of turbidity
and suspended solids from the photo, a few simultaneous water samples must be
taken somewhere in the strip of photos to establish the reflectance - turbidity
curve for that particular set of circumstances.
EH S
g 3 100
u t,
W
H
55 PS
H O
^
w o
u u
E"
U
W
£3
O
« u
EH §
Wj in
P^ fH
75 •>
50 -•
25 •'
8-11-71
1800 Ft.
8-19-71
1800 Ft.
Uniform Overcast
High
Scattered
Clouds With
Haze
8-11-71
2800 Ft.
10
20
TURBIDITY (JTU)
30
40
Fig. 7. Change in Reflectance-Turbidity Curve due to
changes in cloud conditions and altitude.
VIII - 15
-------
B. Water Quality From Satellites
The most important correlation to be made, is between the water itself
and the aerial photo. Once this is done, the correlation between aerial photos
and satellite imagery is not that much more difficult. The most significant
difference is the amount of atmosphere between the plane taking the aerial
photo and the several hundred mile height of the satellite.
There are really less variables when working with ERTS satellite imagery
than with aerial photos. Usable satellite imagery exists only for absolutely
clear days, so the sky-light reflection on a satellite photo is a minimum, and
is always the same. The altitude of the satellite is essentially always the
same. The angle of the sun and the time of photography are also approximately
constant. Also the ERTS imagery is internally uniformly calibrated so there is
no significant changes due to camera and film factors. With ERTS satellite
imagery there is very likely only one reflectance - turbidity curve that must
be established for a particular water. Once established the ERTS imagery dan
be analyzed in conjunction with this curve to ascertain water quality for any
day.
Figure 8 is an ERTS image of the western end of Lake Superior, near
Duluth, taken on 12 August 1972. There is obviously some dirty water in the
bay near the cities of Duluth and Superior. An $8,000,000 water intake for
the city of Cloquet, Minnesota was located in the middle of the turbid water.
A lawsuit against the engineer resulted. The ERTS imagery showed the extent
of this turbid water in a manner better than could be ascertained by any other
means. There was therefore much interest in establishing the relationship
between the water conditions and the brightness of the ERTS image.
In conjunction with the law suit, water samples were taken near Duluth
immediately after a storm in November 1972. Simultaneous water samples and
aerial flights were accomplished. The suspended solids of the water samples
ranged between 40 to 400 mg/1. Turbidity ranged between 10 and 100 JTU's.'
Secchi disc readings varied from 9 feet in the clear water to 8 inches in the
turbid water. The depth of the bay was about 40 feet, so bottom effects were
not significant in this case.
Figure 9 shows the lab-derived volume reflectance versus turbidity curve
for synthesized muddy water and the apparent reflectance versus turbidity curve
obtained from analyzing the photographs. The difference between the two is at-
tributed to the surface reflection of sky-light and other noise effects. There
-------
Fig. 8. ERTS Satellite Image taken 12 August 1972
showing the south west end of Lake Superior near
Duluth. Note the dirty water. An X marks the spot
where an $8,000,000 water intake was located which
when put into operation produced turbid unusable
water.
- 17
-------
M
I-1
CO
M U
u ss
E-i U
U W
H .J
PL, pq
E-i
W S
100
10 ••
1.0 -•
0.1
TURBIDITY, JTU'S
50 500 ^.
_i 1
FIELD
CONDITIONS
Dry
Wet Clay*<>
LABORATORY
CONDITIONS
BOTTOM EFFECTS
FROM SAMPLE TUBE USED
10
100 10J 10
TOTAL SOLIDS, MG/L
10-
Fig. 9. Laboratory volume reflectance and also apparent reflectance from color
and color infrared photographs plotted against turbidity and total solids. The
wavelength used was 0.65 Microns (color red). The location is Lake Superior near
Duluth. The material in the water is red clay which shows up on aerial photos
and ERTS images. On the laboratory analysis between points B and C the effects
of a shiny bottom of the sample tube are overriding, making this data erroneous.
These effects have subseqently been reduced by using a flat black bottom for the
tube. The vertical displacement between the curve for field conditions and lab
conditions is primarily due to the fact that sky-light reflection is present on
the photos, but not present in the lab analysis.
-------
is a good correlation between turbidity and apparent reflection. Similar field
tests were run two and three weeks after the storm. The dirty water remained
in the bay during this time but its characteristics changed, due to settling of
the heavier particles. When all the laboratory data from the 3 days were com-
bined the plot between lab-derived volume reflectance versus turbidity and
secchi disc readings hold well for all three days (see Figure 10).
It was obvious that it was the parameter of turbidity that correlated
at all times with the reflectance and therefore the brightness of the ERTS
imagery. An analysis was then made to see what other parameter correlated to
turbidity. As the secchi disc reading is just a rough field means of measuring
turbidity, it provided a good correlation to turbidity, which is also shown in
Figure 10. Suspended solids were the only other parameter that came close to
correlating with turbidity. Even then there was a distinctly different sus-
pended solids - turbidity curve for each of the three days (see Figure 11).
On any one day there was a very good correlation between suspended solids and
turbidity. Therefore, for a given day, if one had the proper ground truth,
suspended solids as well as turbidity, could be mapped from the ERTS imagery.
The reflectance - turbidity curve holds for all days so on any day for the
dirty water near Duluth, one can obtain the reflectance from analyzing the
ERTS image, and from it can calculate the turbidity. Also from the reflectance
and the reflectance - secchi disc reading curve in Figure 10, one can obtain
the secchi disc reading for water anywhere in the ERTS image. The aerial image
sees no deeper than the secchi disc reading. (14) Therefore by comparing the
calculated secchi disc readings with a hydrographic chart of the area it is
possible to tell if bottom effects are significant on the ERTS imagery.
The ERTS image, Figure 8, was analyzed with a microdensitometer, to
obtain values of apparent reflectance at various points in the lake. These
apparent reflectance values on Figure 9 resulted in turbidity values for the
points ranging from 4 to above 100 with the calculated turbidity at the intake
location being about 80. An exact check of the satellite's reliability and the
effect of the intervening atmosphere would have required water sampling simul-
taneous with the ERTS overflight. This of course, was not accomplished in
August 1972, but can be done in future tests.
Another potential use of satellite imagery presented itself in an in-
vestigation conducted with the U.S. Forest Service. They were interested in
the potential use of satellites for categorizing the clarity of lakes near
¥111 - 13
-------
w
u
2
EH
U
W
W
2
|
W
2
0.5 .
0
*&
H
w
u
I
u
5 10 20 50 100
TURBIDITY, JTU'S
Reflectance vs. Turbidity
Water Samples Collected 4 November 1972
(2 days after a heavy storm)
Water Samples Collected 17 November 1972
Water Samples Collected 23 November 1972
5 ..
1 ••
w
S 0.5
•4-
I i l
8 20 50 100
SECCHI DISC READINGS, INCHES
Reflectance vs. Secchi Disc Readings
Fig. 10. Laboratory Reflectance versus Turbidity and
Secchi Disc Readings for all of the water samples col-
lected during the 3 days in November, 1972. The location
is dirty water in Lake Superior near Duluth as shown in
Figure 8.
VIII - 80
-------
100
1-1
l-l
I
ro
w
a
u
z
H
to
C
2
M
Q
U
cn
H
Q
H
SC
U
U
U
tfi
10--
-- 100
November 1972
November 1972
50
100
TURBIDITY, JTU'S
Fig. 11. Plot of Secchi Disc Readings and Suspended
Solids versus Turbidity for all of the water samples
collected in dirty water in Lake Superior near Duluth
in November 1972.
-------
Ely, Minnesota, which is located in the northwest corner of the full frame just
off the illustration in Figure 8. Three lakes were sampled and the samples
analyzed during the summer of 1973: (1) Lake Shagawa classified as eutrophic
(enriched) (2) Lake Ensign, somewhat more clear and classified as mesotrophic
and (3) Lake Snowbank, a very clear lake classified as an oligotrophic lake.
Reflectance values were obtained from the water samples from these lakes and
an analysis was made for solids and turbidity. These values are presented
graphically in Figure 12. It is apparent that there is a very good correlation
between volume reflectance and turbidity and solids as well as the lake clas-
sification used by the Forest Service. When the ERTS image was analyzed with
the microdensitometer, the apparent reflectance data yielded a straight line,
about 5% higher than the laboratory data. This case provides a direct compari-
son of lab results with satellite data and the results are essentially the
same as that shown on Figure 9 (a comparison between the lab data and low
flying aerial photography). In both cases, the field curve is about 5% higher
than the lab curve with a somewhat steeper slope. It appears that the sky-
light reflection effects in both cases are essentially similar. Further work
by John Van Domelen is focused on better ascertaining the exact magnitude of
the skylight reflection for all situations.
C. Effects of Oil on Sky-Light Reflection
In Figures 6, 9 and 12 the vertical shifts between lab and photo (field)
reflectance values are attributed to sky-light reflection. If the sky condi-
tions are uniformly clear or overcast this effect will be essentially constant
over the photograph. We also assume that the surface reflection of the water
itself is uniform. This is generally the case, but if oil is on the water sur-
face, it can significantly alter locally, the amount of reflection. Assuming
all else remains constant, the amount of surface reflection varies almost
linearly with the amount of thickness of the oil spill. Oil film detection is
most pronounced on overcast days because the sky-light, which the oil reflects,
is much brighter on such overcast days. So if all else remains constant, one
can also monitor oil thickness changes by the apparent reflectance as shown on
the photo. This area is presently being studied by John F. Van Domelen.
CONCLUSIONS
Many experts in water quality work are reluctant to use aerial photos in
VIII - 22
-------
SUSPENDED SOLIDS, MG/L
W
u
10 ..
5 ..
2 ..
1.0 ••
0.5 ••
0.2
1
-t-
Apparent Reflectanc^
FIELD DATA
D
Volume Reflectance
LAB DATA
5 10
TURBIDITY, JTU'S
20
Pig. 12. Volume Reflectance from analyzing water samples
in the lab and apparent reflectance from ERTS imagery
plotted against turbidity and suspended solids. The three
lakes analyzed are classified in three different stages
of enrichment or Eutrophication as follows:
U.S. Forest Service
Sample Lake Classification
A Distilled Water
B Snowbank Lake
C Ensign Lake
D Shagawa Lake
NA
Oligotrophic (clear)
Mesotrophic(Middle Stage)
Eutrophic (enriched)
•Til - 23
-------
their work. However, when properly used, photography is a very important and
useful tool for water quality investigations. Anyone observing an aerial image
of industrial pollution or of siltation extending into clear water intuitively
realizes that these photos qualitatively show water conditions and the distri-
bution of any disturbing material. Further refinement of image analysis and
water analysis makes it possible to quantitatively obtain turbidity from aerial
photos. To whatever extent suspended solids and other water quality parameters
correlate with turbidity, these can also be quantitatively mapped, provided that
simultaneous water sampling is properly conducted.
The interaction between light and water is quite complex and these inter-
actions must be understood before aerial photos can be effectively used. Impor-
tant considerations are the surface reflection of the sun and sky-light energy.
Except for the monitoring of oil spills, surface reflections are considered an
undesirable noise, but if aerial photos are taken either on completely clear or
uniformly overcast days the sky-light reflection can be uniformly quantified
and dealt with. The sun's reflection can be avoided by proper choice of the
camera look angle.
Another undesirable noise factor can be bottom effects. However, if a
secchi disc is used and the reading obtained is less than the depth to bottom,
the bottom effects are not important. If the secchi disc reading is greater
than the bottom depth, bottom effects are an important factor, and the analysis
of aerial images must be done in the near infrared spectrum as infrared energy
penetrates only a few inches into the water.
The desirable factor to be analyzed is volume reflectance, as the water
quality parameter of turbidity correlates to volume reflectance. This general
correlation can be determined either in the laboratory or in the field. From
a photo, the apparent reflectance for the field conditions are higher than those
from the lab because of sky-light reflection. To ascertain the slope of the
turbidity versus apparent reflectance curve and its vertical shift due to sky-
light, water sampling must be done simultaneously with any overflight. This
is in keeping with conventional photogrammetry practices. As ground control
is to photogrammetry, so is ground truth sampling to remote sensing. Remote
sensing will not eliminate ground water sampling, it just makes it thousands
of times more efficient.
Enough is now understood about the interaction of light rays with water
so that remote sensing can now be utilized as a tool for water quality
VIII - 24
-------
investigations, and it should be so used. Had the designer of the water intake
in Lake Superior near Duluth properly looked at aerial photos showing the areas
of turbid and clear water, the intake should have been located 6 miles to the
northwest and taxpayers in that area would have a functional water intake in-
stead of an $8,000,000 blunder.
VIII -::25
-------
WORK CITED
1. Burgess, Fred J., and James, Wesley P., "Aerial Photographic Tracing of
Pulp Mill Effluent in Marine Waters," WATER POLLUTION CONTROL RESEARCH
SERIES, No. 12040, Federal Water Quality Administration, 1970.
2. Colwell, Robert N., et al., "Basic Matter and Energy Relationships Involved
in Remote Reconnaissance," PHOTOGRAMMETRIC ENGINEERING, Vol. XXIX, No. 5,
September, 1963, pp. 761-799.
3. Klooster, Steven A., "Instruction Manual for Microdensitometer and Scanning
Stage," includes design of new stage, unpublished manual, Institute for
Environmental Studies, University of Wisconsin, Madison, Wisconsin.
4. Klooster, Steven A., and Scherz, James P., "Water Quality Determination by
Photographic Analysis," Report No. 21, Institute for Environmental Studies
Remote Sensing Program, Department of Civil and Environmental Engineering,
University of Wisconsin, Madison, Wisconsin, August, 1973.
5. Lillesand, Thomas M., "Use of Aerial Photography to Quantitatively Elevate
Water Quality Parameter in Surface Water Mixing Zones," Ph.D. Thesis, 1973,
University of Wisconsin, Madison, Wisconsin.
6. Piech, Kenneth R., and Walker, J.E., "Photographic Analysis of Water Re-
source Color and Quality," PROCEEDINGS, 37th Meeting of American Society of
Photogrammetry, Washington, D.C., March, 1971.
7. Rinehardt, Gregory L., and Scherz, James P., "A 35MM Aerial Photographic
System," The University of Wisconsin, Institute for Environmental Studies,
Remote Sensing Program, Report No. 13, April, 1972.
8. Scherz, James P., Graff, Donald R., and Boyle, William C., "Photographic
Characteristics of Water Pollution," PHOTOGRAMMETRIC ENGINEERING, Vol. 25,
No. 1, January, 1969.
9. Scherz, James P., and Kiefer, Ralph W., "Applications of Airborne Remote
VIII - 26
-------
Sensing Technology," JOURNAL OF THE SURVEYING AND MAPPING DIVISION,
PROCEEDINGS, American Society of Civil Engineers, April, 1970.
10. Scherz, James P., "Development of a Practical Remote Sensing Water Quality
Monitoring System," PROCEEDINGS, 8th International Symposium on Remote
Sensing of Environment, Ann Arbor, Michigan, October, 1972.
11. Scherz, James P., "Monitoring Water Pollution by Remote Sensing," JOURNAL
OF THE SURVEYING AND MAPPING DIVISION, American Society of Civil Engineers,
November, 1971.
12. Scherz, James P., "Remote Sensing Considerations for Water Quality Monitor-
ing,11 PROCEEDINGS, 7th International Symposium on Remote Sensing of Environ-
ment, Ann Arbor, Michigan, May, 1971.
13. Scherz, James P. and Van Domelen, John F., "Lake Superior Water Quality
Near Duluth from Analysis of Aerial Photos and ERTS Imagery," PROCEEDINGS,
International Symposium on Remote Sensing and Water Resources Management,
June 11-14, 1973, Burlington, Ontario, Canada.
14. Scherz, J.P., "Final Report on Infrared Photography Applied Research Program,
Report No. 12, Institute for Environmental Studies, The University of
Wisconsin, Madison, Wisconsin.
15. Standard Methods for Examination of Water and Wastewater, 13th Edition,
Turbidity, pp. 349-355, American Public Health Association, 1015 Eighteenth
St., N.W., Washington, D.C.
15. Van Domelen, John F., "Determination of Oil Film Depths for Water Pollution
Control," Report No. 4, University of Wisconsin, Institute for Environmental
Studies, Remote Sensing Program, June, 1971.
VIII - 27
-------
LIDAR POLARIMETER MEASUREMENTS OF WATER POLLUTION
John W. Rouse, Jr.
Remote Sensing Center
Texas A§M University
ABSTRACT
Laboratory measurements of laser light back-
scattered by turbid water and oil on water have indi-
cated a potential for dual polarization laser radar
(Lidar Polarimeter) for remote water pollution moni-
toring. A lidar polarimeter system employing a 5 mw
He-Ne laser has been constructed and used in labora-
tory and daylight field measurements of turbid water
and thin oil spills. The system concept is based on
an electromagnetic backscatter model which attributes
depolarized backscatter to multiple scatter within the
subsurface volume. This component is related to the
number density and sizes of suspended particles within
the volume. The experimental system described records
both the like and cross polarization backscatter compo-
nents for either horizontal or vertical transmit polar-
izations .
The laboratory system is being expanded for
the Coast Guard to include two lasers operating at
633 nm and 441.6 nm to enable continuous monitoring of
oil spills. The four backscatter components are to
be 'processed in real time using a digital signal
processor which classifies the status of the water
according to turbidity or oil spill condition.
This paper describes the lidar polarimeter
concept* system configuration, and field measurements
conducted at Texas A&M University.
VIII - 28
-------
INTRODUCTION
In 1972 Rouse [1] published the results of a
theoretical study of electromagnetic backscatter from
rough surfaces which showed that the depolarized back-
scatter component was dependent upon a near-surface
volumetric scattering process. This work was based upon
previous theoretical studies by Fung [2] and Leader [3]
and experimental studies by Renau et al [4], Cheo and
Renau [5], and Leader [3]. As a result of these efforts,
the interpretation of both optical and microwave back-
scatter measurements have been modified, since previous
rough surface models had predicted depolarization due
solely to multiple scatter on the surface.
In an effort to establish the validity of the
model, a series of laboratory measurements are being made
using a He-Ne laser source. This activity has led to the
development of a new sensor instrument called a Lidar
Polarimeter. This device has shown potential for remote
measurements of turbid water and oil on water. The sensor
is still in a developmental stage and field confirmation of
certain effects seen in the laboratory is yet to be obtained,
The purpose of this paper is to describe this lidar polar-
imeter and summarize the preliminary experimental results.
-------
SENSOR CONCEPT
The scattering model developed by Rouse [1] Is
based upon the assumption that electromagnetic backscatter
from inhomogeneous, rough surface dielectrics can be ob-
tained by assuming a volume scatter component added inco-
herently with a surface scatter term. In the model, the
volume scatter process is described by a linear transforma-
tion of the surface fields. The volumetric scattering pro-
cess is controlled by the transmission properties of the
surface and the inhomogeneity of the volume. In Rouse's
work, the depolarized backscatter is also dependent upon
the surface statistics.
In an independent study by Leader [6], a similar
model was developed which has the same general properties,
except that the depolarized component is independent of
the surface roughness. A later study by Wilhelmi [7]
exp'anded upon these two studies and led to a formulation
based on the physical optics approach which shows excellent
agreement with experimental measurements. Wilhelmi's model
supports Leader's observation that the depolarized back-
scatter is independent of surface roughness and is strongly
dependent upon the number density of particles within the
YTTI - 30
-------
subsurface volume.
The general nature of the like and cross polar-
ized backscattered power received from a rough surface is:
P.. = Surface Component + Subsurface Component
^ Z-
P. . = Subsurface Component
13
Where -L, 3 denote transmitted and received polarization.
At microwave frequencies, the subsurface components
are generally much less than the surface components. This
is true whether the target is land or water. However, at
optical frequencies, the subsurface component can be signi-
ficant, especially if the target is turbid water.
LIDAR POLARIMETER
The lidar polarimeter developed at Texas A§M
University employs a -low-power CW helium-neon (633nm) laser
which transmits a narrow beam of polarized, chopped mono-
chromatic light. A narrow beam, dual-polarization receiver
detects both vertically and horizontally polarized light
backscattered from the illuminated target. The field of
view of the receiver is matched to that of the transmitter.
A narrow-band optical filter, small angular field of view,
and synchronous demodulation of the detected radiation are
combined to eliminate signals produced by skylight and
-------
and scattered sunlight. During operation the laser beam
is directed away from the nadir at a sufficient angle to
insure a predominance of diffuse backscatter. The basic
system diagram is shown in Figure 1.
The present system, is capable of operation at
slant ranges of about 30m over natural waterways having
depolarized scattering cross sections as low as -50db.
The measurements can be made independent of lighting
conditions and the effect of atmospheric interference
is minimized. The operational configuration is shown
in Figure 2.
MEASUREMENT OF TURBID WATER
Laboratory measurements of two types of turbid
water solutions were-made to determine the effect of the
subsurface volume composition on the depolarization process
One solution consisted of water with suspended spherical
particles of teflon (Dupont TFE 30) ranging in size from
0.04pm to 0.4ym diameter. The solution also contained
varying concentrations of nigrosine black dye. The beam
extinction coefficients due to scattering and absorption
were found to be
as = 85ps cm"1
a = 6600p0 cm
a 3
VIII - 32
-------
M
I
O1
Output Chopper
Optics
Linear Polarizer
Laser
Light Shield
Collimating Lens
Telephoto
Lens
0 1
O
Spatial
Filter
Spectral Filter
[Beam Splitting Pol?
1
I
«-Photodi<
Figure "1
Schematic diagram of Lidar Polarimeter.
-------
LIDAR Z*
POURIMETES
Figure 2 Lidar Polarimeter.
VTTI - 54
-------
where p = weight (grams) of teflon per cc of solution
p = weight (grams) of dye per cc of solution
The like and cross polarization backscattering
cross sections were measured for a range of scatterer
and dye concentrations at beam incidence angle of 23°.
The beasn diameter at the water surface was 1.5 cm. The
results are shown in Figure 3. These data confirm the
validity of the theoretical contention that depolarization
is exclusively a volume scatter process and that the
depolarization is proportional to the density of the
particles in the subsurface volume. However, it is also
clear that the absorption affects the data significantly,
and that direct correlation between the laser return and
the scattering particle density cannot be expected without
a. priori knowledge of the absorption factor.
The second solution consisted of water and an
ordinary casein solution which contained a wider size range
of scattering particles of random shapes. Dye was also
used in the solution to establish a beam extinction
length less than the depth of the container. The dye
concentration was 14.4 mg of nigrosine black dye per
liter of solution. Concentration of scatters from 23 ml
VIII - 35
-------
-20--
-30 --
-40 --
-50
W
10
p x 10" gm/cm
s
Figpre 3 - Polarized and depolarized backscatter
vs. mass concentration of scatterers.
Angle of incidence equals 236.
VIII - 36
-------
to 193 ml of scattering solution per liter of water
were measured. The results are shown in Figure 4.
These data were recorded with an incident beam diameter
of 5 cm. This is a larger illumination field than was
employed for the first solution and the measurements
represent a stronger multiple volume scatter process.
The results indicate almost total depolarization of the
incident energy and a very small surface dependent
scattering component. The data show a direct correlation
with scatterer concentration, j.ust as with the first
solution.
The lidar polarimete.r was also used to measure
backscatter from a natural'waterway. These measurements
of an area of the Brazos River near Waco, Texas were
supportive of the laboratory data. The river was measured
over a period of several days and data were obtained from
turbid water varying from 8.0 to 35.0 FTU. Measurements
of total suspended solids varied from 10.0 to 50.0 mg/L.
The transmittance varied between 70 and 94 percent. The
laser like and cross polarized backscatter measurements
were linearly related to these water parameters. The
average measurements from the river water were similar
to the data obtained in the laboratory using a teflon
VIII - 57
-------
1.2--
1.0..
.8--
rH
v_/
t>
.6--
.4..
©
e = 20
©
6 = 30
50 100 150
^ ml of scattering solution
S £ of H00
Figure 4 - Polarized and depolarized backscatter
vs. mass concentration of scatterers.
VIIT - 58
-------
scatterer density of approximately 1.0 mg/cm without
the absorbing dye.
MEASUREMENTS OF ROUGH SURFACE DIELECTRICS
The turbid water measurements described were
made for smooth water surfaces. An additional experiment
was performed to determine the effect of rough surfaces
on the volume-dependent backscatter. The targets employed
consisted of several samples of dielectric casting resin
containing varying pigmentations. The samples provided
a range of surface roughnesses, from very smooth to height
deviations of approximately 60 wavelengths. The varying
pigmentations simulated the several scatterer-dye combina-
tions used in the turbid water experiments. The pigmenta-
tions range from black, representing high absorption and
negligible volume scatter, to white, representing large
volume.scatter and negligible absorption. A comparison
of the backscatter from these samples is shown in Figures
5 and 6. Two factors are obvious from these measurements.
First, the backscatter behavior is as expected from the
turbid water measurements. That is, the depolarization is
volume-dependent and is strongly influenced by the absorp-
tivity. Second, the volume-dependent depolarization
VIII -.39
-------
•a
® "80
10 .
20 -
i
<
30 J
<
i
C
40 -
50 .
<£> "180
© "220
© "320
grit"
grit"
grit"
grit"
© "smooth"
O "80
©"180
0"220
® "320
grit"
grit"
grit"
grit"
© "smooth"
b
9 -.
a iss) _~
r **><** *®Q> &<&>& ^ ^ ,
^
>
> 0
0 0.
^> C5^> ®J> p,,, -
T^ O C0*(@si
^ ffi <^ OB g
© * § <^
©
®
® ® ® ~ ^
® ® ®
1 1 1 1 1
>& A
* ; *
^ ^. &
? $. *
@ CD
0
® ®
»
White
White
White
White
White
Black
Black
Black
Black
Black
®
£& S
*Sr B^
*
5>
§j«V
O
§
®
®
i
10
20
30
40
50
60
70
Angle of Incidence (degrees)
Figure 5 - Vertically polarized data for the white
and black samples for all five surface
roughnesses.
VIII - 40
-------
-10
-20
® "80 grit" White
3> "180 grit" White
© "220 grit" White
• "320 grit" White
e "smooth" .White
® " 80 grit" Black
© "180 grit" Black
O"220 grit" Black
® "320 grit" Black
-30 ._
-40 ...
-50
-60
fy-
®
«O
0
©
-f
10 20 30 40 50 60 70
Angle of Incidence (degrees)
Figure 6 - Depolarized data for the white and
black samples for all surface rough-
nesses with vertical transmit polar-
ization.
VIII - 41
-------
component is virtually independent of surface roughness.
DETECTION OF OIL ON WATER
The preliminary experimental results indicate a
potential for the use of the lidar polarimeter for remote
measurement of certain water quality parameters. There
is also reason to believe that the techniques can be
employed to detect the occurrence of oil spills on turbid
water.
The problem of describing the electromagnetic
backscatter from an oil-on-water medium can conceivably
be handled by combining the physical optics approach to
describing scatter with an established layered media model,
Generally, an oil layer is spatially inhomogeneous. In
addition to these inhomogeneities, an oil layer apparently
differs from the conventional layered media models in that
the common "layer effect" normally giving rise to an
interference phenomena is of secondary importance due to
the non-uniform thickness of the layer.
Lidar polarimeter measurements of 1 mm thick
layers of heavy and medium crude oil, kerosene, and gas-
oline indicate that a layered media model assuming the
layer to act as a lossy dielectric is appropriate. That
VIII - 42
-------
is, these substances contribute little backscattered
energy, rather the principal contributing source of
backscatter is from within the volume of the underlying
water. Measurements of refined motor oil (SAE 30) however,
show it to have a significant volume scatter term due to
multiple scatter within the oil layer. Backscatter from
this oil type is almost two orders of magnitude greater
than any other petroleum product tested.
The model under development for use in describing
the oil-ori'-water medium consists of three primary components
1. An oil-air surface scattering component
controlled by the index of refraction of
the oil (approximately 1.45) and the surface
roughness. This term has the same polar-
ization as the incident energy, however the
term should be small for incidence angles
above about 20°.
2. A volume-dependent component from within the
oil layer. This term exhibits depolarization
and is highly dependent upon the oil type.
The term will be small for most oil types.
3. A volume-dependent component from within the
subsurface water volume. This term exhibits
VIII - 47.
-------
depolarization; is highly dependent upon
the density of suspended particles; and
experiences an attenuation proportional
to the oil thickness and dependent upon
the oil type.
Each o£ these terms adds incoherently to the total back-
scattered power received for each polarization, that is:
P. . = P + P . + yP •
t-t. S O1 W"Z-
P. . = P . + yP .
13 03 wa
where: P.. = like-polarized backscatter
•z-t.
P.. = depolarized backscatter
13
P = oil surface term
•s
P = oil volume term
o
P = water volume term
y = attenuation of oil layer (two-way)
DUAL-WAVELENGTH LIDAR POLARIMETER
The potential for detection of oil on water using
a monochromatic (633nm) lidar polarimeter depends upon the
effect of the change in the index of refraction and the
added attenuation of the oil layer over the water volume.
Initial tests show that the presence of oil does alter the
like and cross-polarized backscattered return, but that
VIII -44
-------
oil cannot be uniquely differentiated from certain turbid
water conditions. To improve the detection reliability/
a record generation lidar polarimeter is being constructed
for the Coast Guard which employs two lasers (633nm and
441.6nm). The shorter wavelength source takes advantage
of the marked increase in the index of refraction of many
oil types in the blue region. The two wavelength system
increases the separability of turbid water classes and oil
types, and provides potential for determination of oil
thickness.
This system has the added feature of real-time
classification of the lidar polarimeter signals. The
signal processor is a special purpose digital computer
capable of immediate analysis of the water condition through
the implementation of appropriate classification algorithms.
The real-time signal classification technique was devel-
oped by Texas A§M University for the Naval Ordnance
Laboratory for use in real-time classification of sea
ice using an airborne radar sensor.
CONCLUSION
Measurements of laser light backscattered from
inhomogeneous media, confirm the theoretical prediction that
VIII - 45
-------
depolarization is a subsurface volume scatter process which
is dependent upon the density of suspended particles in
the volume. This effect has been utilized in the measurement
of the turbidity of water and in the detection of oil on
water. A new remote sensing device is under development
to employ this potential for remote, continuous, day-night
surveillance of water surfaces to automatically identify a
range of petroleum and petroleum byproducts and other
contaminants.
ACKNOWLEDGEMENT
The dual wavelength Lidar Polarimeter is being
ieveloped for the Coast Guard under contract DOT-CG-34017-A.
'The opinions or assertions contained herein are the private
snes of the writer and are not to be constructed as official
>r reflecting the views of the Commandant or the Coast Guard
it -large."
VIII - 46
-------
REFERENCES
1. J. Rouse, Jr., "The effect of the subsurface on the
depolarization of rough surface backscatter", Radio
Science, 7_, pp. 889-895, 1972.
2. A. Fung, "On depolarization of electromagnetic waves
backscattered from rough' surfaces", Planetary Space
Science, 14, pp. 563-568, 1966.
3. J. Leader, "Bidirectional scattering of electromagnetic
waves from rough surfaces", Report MDC 70-022, McDonnell
Douglas Corporation, St. Louis, Missouri.
4. J. Renau, P. Cheo, and H. Cooper, "Depolarization of
linearly polarized EM waves backscattered from rough
metals and inhomogeneous dielectrics", J. Optical
Soc. Amer., 57 (4), pp. 459-466.
5. P. Cheo and J. Renau, "Wavelength dependence of total
and depolarized backscattered laser light from rough
metallic surfaces", J. Optical Soc. Amer., 59 (7),
pp. 821-826. •
6. J. Leader, "Bidirectional scattering of electromagnetic
waves from the volume of dielectric materials", J_._
Applied Physics, vol. 43, p. 3080, 1972.
7. G. Wilhelmi, "An investigation of the depolarization of
backscattered electromagnetic waves using a lidar polar-
imeter", Tech. Report RSC-45, Texas A§M University,
College Station, 1973.
VIII - 47
-------
Progress Report
DETECTION OF DISSOLVED OXYGEN IN WATER THROUGH
REMOTE SENSING TECHNIQUES
By Arthur W. Dybdahl
Abstract
The technique of detecting dissolved oxygen concentrations in the
country's waterways with airborne remote sensing is discussed. This
technique was developed at the National Field Investigations Center (NFIC),
Office of Enforcement and General Council, EPA, Denver, Colorado. The
recording media and data processing are explained. Experimentation is
presently under way to quantify the airborne reconnaissance data so that
concentrations of dissolved oxygen to within 1 to 2 parts per million can
be readily obtained. A brief discussion of the water parameters that cause
interference with the utilization of this technique are discussed.
VTII - 48
-------
Progress Report
DETECTION OF DISSOLVED OXYGEN IN WATER THROUGH
REMOTE SENSING TECHNIQUES
INTRODUCTION
There is a great need present for a technology to be used in the rapid
assessment of water quality parameters in large and small.bodies of water
throughout the country. This technology will be developed and applied to
practical operations through remote sensing techniques. Qualitative detection
of water quality parameters, such as turbidity, suspended solids, dissolved
solids, and color, are presently at hand with the use of airborne-cameras
and passive scanners. There is a paramount need to quantify remote sensing
data in terms of the water quality parameters for enforcement and monitoring
applications. The development of a quantitative detection mechanism for
dissolved oxygen in water is a beginning in this enormous task.
APPLICABLE OPTICAL PROPERTIES OF WATER
A great deal of effort has gone into the measurement of the optical
properties of water since the mid-1940's. Samples of distilled water, in
addition to those obtained from the oceans, coastal, and bay (estuarine)
waters, have been thoroughly tested for optical transmittance and reflec-
tance properties. Distilled water has the maximum transmittance values
in the bandwidth from 400 nm to 650 nm. Ocean water is the next in line
with a noticeable decrease in transmittance from 400 nm to 550 nm. Coastal
and bay waters display significantly lesser transmittance values than that
I o
of the ocean waters (Specht ). The data for the attenuation (Mairs11) and
the extinction (Hale3) properties of water are also collectively available.
VIII - 49
-------
00 45O 500 55O SOO 65O ?00
The transmittance data for the four above mentioned types of water is
plotted in Figure 1. This data was measured for a water path length of
10 meter. Note that the loss
in transmittance is far greater in
the 400 nm to 460 nm region (blue)
than the green region (460-575 nm),
this is an important feature used
in the detection of dissolved
oxygen in water.
FIG. 1. Spectral transmittance for ten meters of
various water types.
THE AIRBORNE DETECTION TECHNIQUE
Recordi n g Te ch n i q u e.
The technique for the airborne detection of dissolved oxygen in water
was found and developed through empirical data rather than through any
deep-thought scientific formalism.
This technique was discovered in late 1970. It has been under con-
tinual expansion, testing, and field verification s.ince that time.
The recording medium aboard an aircraft is a camera, Kodak 2443
false color infrared film and a Wratten 16 orange gelatin filter. The
exposure is set on the camera at approximately 1/4 to 1/3 f-stop, less
than the so-called normal exposure. This sensor renders healthy water,
saturated with dissolved oxygen, as a bright-brilliant blue. It renders
VIII - 50
-------
septic (near zero dissolved oxygen waters) virtually black, or more
precisely, the characteristics of unexposed processed film.
The film is processed with the usual EA-5 chemicals and procedures.
Field Verification of Film Indications
In an attempt to optically explain the above film indications,
visual and hand-held photographic observations were carried out during
EPA Water Quality Surveys, in areas where saturated and septic waters were
being tested. Saturated waters were quite clear with the bottom visible
to depths of 8 to 10 meters The overall color of this type of water
was greenish-blue. Septic waters were nearly opaque as far as depth
penetration and displayed a very dark gray-green color. These waters were
known to be subjected to a very high biochemical oxygen demand (BOD).
This field data supports the water transmittance curves shown in Figure 1.
Film Interpretation and Analysis
Recalling the false color rendition properties of the Kodak 2443
false color infrared film, any natural green scene will record as blue,
assuming the absence of chlorophyll substances which photographs as red.
Waters saturated with dissolved oxygen photographed in a bright-blue color.
As the oxygen demand increases, the dissolved oxygen is decreased, the bright-
blue color of the water, recorded in the film, gradually progresses through
a dark-blue to black. By staying in the linear portion of the film char-
acteristic curves, it is a reasonable assumption that the optical response
of the target being photographed will be recorded linearly. This "blue-
to-black" indication results from the optical absorbance in the green region,
being increased as the dissolved oxygen concentrations decrease. This
technique does not work in the region from 0.6 microns (600 nm) to 1.0 microns
VIII - 51
-------
(1,000 nm) which is the red and near-infrared portions of the optical spec-
trum.
Several areas throughout the country have been flown commensurate
with the gathering of detailed ground truth. The dissolved oxygen data
from the ground truth has been used in an attempt to calibrate the
"blue-to-black" curve resulting from densitometric measurements made
upon the exposed aerial film. Initially only the data from two bay areas
having equal water temperatures at the time of flight, was used for the
calibration procedure. Septic waters appear black in the film densito-
metrically identical to the unexposed film area between frames. The task
of calibration is to identify film density data with the precise area of
ground truth. Blue, green, yellow, and red density measurements were
made on the film with a Macbeth TD-203AM Transmission Densi tometer. The
density data was compared to the ground truth. The variable color parameter
is blue (density). The data is then plotted. One such calibration curve
is shown in Figure 2. The ground truth was used to establish the position of
each straight line. The abscissa of Figure 2 is the film density. It is
commonly defined as:
D = Iog10{^-) (1)
where T is the transmittance of the particular area on the film under test,
for a particular color filter (red, green, or blue). It is easily seen
that the dynamic range in. film is greater than 1000:1. The ordinate
represents the concentration of dissolved oxygen in parts per million (ppm)
or milligrams per liter (mg/1). The range on this parameter has been
limited to 0 to 10 ppm because in the natural environment the dissolved
VIII - 52
-------
oxygen levels rarely exceed 10 ppm. To exceed this value usually indicates
supersaturation accomplished by water pouring over rapids or in close
proximity of an aerator. In bodies of water containing large amounts of
aquatic plant growth (in the summer), the dissolved oxygen concentrations
can reach, or even exceed, the saturation limit in the mid-afternoon hours,
the period of maximum oxygen production.
Any test method has inherent problems with interference phenomena.
This one is no exception. Waters in the natural environment contain finite
amounts of suspended solids and organic waste materials. The impact of
the materials upon the optical transmittance properties of various types
of natural water is readily seen (Figure 1). The above mentioned materials
contribute a yellow to light-red cast to water. When this is superimposed
with the natural bluish-green color of oligotrophic-type waters, the blue
cast is decreased and the water becomes predominantly green (Figure 1).
Since deep yellow to red materials photograph green in the false-color
infrared film, the green transmittance factor becomes an important one
as a base line for each particular area of water under test. Red is not
used in conjunction with this film because of the sensitivity of the cyan
emulsion layers to the near-infrared region (0.7p to l.Op). This would
create an additional interference factor when aquatic plant growth is
present and would not represent a reliable base line. So, through the
optical inspection and analysis of over 1,100 data sets (red, green, and
blue form a set for each point on the film) three color parameters have
been established for calibration purposes. They are:
(1) Blue
(2) Green
(3) Blue-minus-Green
VIII - 53
-------
The parameter denoted by (3) has been chosen, as opposed to " Green -minus -
Blue" to eliminate having to use negative irrational numbers. Again, the
green parameter establishes the suspended solid material influence in the
water under test. The initial density/ground truth data is plotted for
the blue and green parameters. Then the actual calibration is carried out
with "Blue-minus -Green" density value plotted with the concentration of
dissolved oxygen. This plot is usually referred to as a difference curve.
The difference curve for the data sets given in Figure 2, is provided as
Figure 3. In practice, the concentration vs. density data is programmed
into a computer. Two data sets are required. They are the Blue and Green
density values for the point on the film corresponding to the physical
location of the ground truth and for the unexposed (between frame) film.
The latter eliminates the influences of base fog if any is present. The
computer then calculates the equations of the three straight lines. This
is done by using the point-slope form of the equation of a straight line
in two dimensions. This is accomplished by the data points Pi(x],y-|) for
high dissolved oxygen, and P(x>y2) for zero dissolved oxygen, and
dx
where & is the first derivative of the equation for the line and is
equal to (y2 - y-|)/(x2 - x-|) = M. The expressions for the Blue and Green
lines in Figure 2, respectively, are:
yB]ue = 2.655xBlue + 10.462 (3)
^Green = 3-715xGreen + 10'866 <4>
where xBlue» xGreen are *'ie blue, green film densities, respectively
e' vGreen are dissolved concentrations for the blue,
green lines, respectively, in parts per
million (ppm).
VIII - 54
-------
The expression for the difference line is calculated from this data,
which for Figure 3 is:
y = 9.307(1 - XA) (5)
where XA = blue-minus-green density and
y = dissolved oxygen concentration in ppm.
To use the calibration curve in Equation (5), the blue and green data
pairs are supplied to the computer. It calculates the value of Blue-Green
value (XA) and finally the dissolved oxygen concentration, designated by
y.
The relative uncertainty of y can be calculated from Equation (5).
dxA
(6)
where the vertical lines signify absolute values, |-x| = x. The absolute
uncertainty for a given calibration curve is:
|dy| = |-9.307dxA| (7)
If dxA = 0.02, the accuracy of the Macbeth densi tometer, then |dy| is
0.186 ppm. The variance is given as |dy)2 = 0.0345.
Hundreds of data points for healthy waters, with insignificant dis-
coloration and suspended solids have been plotted against the calibration
curve in Figure 3. The points all fall within 1 ppm of the line.
VIII - 55
-------
It must be mentioned that one must be cognizant of the inherent errors
of the field equipment used in obtaining the dissolved oxygen phase of
ground truth. Some types of systems are consistently off by as much as
1 ppm. This error must be compensated for in order to produce an accurate
calibration-curve.
Significant Affects Upon the Calibration Curve
A. Film Exposure
Film exposure levels can have a significant affect upon the slope
of the calibration curve. Over-exposure with the same calibration curve,
results in a lesser dissolved oxygen concentration than is actually pre-
sent. Likewise, under-exposure results in a greater concentration than
is actually present. For this reason, a new calibration curve based upon
the inherent film exposure at the center of a photographic frame near
the intersection of the fiducial marks, is generated for each mission
location.
B. Lens/Illumination Effects
A KS-87B aerial framing camera is used on all OEGC missions. It
has a 152 mm lens cove assembly with a lens fall-off characteristic curve
of Cos12e, where e is the angle off the principal axis of the lens. A
plot of exposure vs. distance across the film format will reveal the effect
of the lens fall-off.
There is one more factor that integrates with the lens fall-off
to influence the optical irradiance across the film format. This factor
is the solar illumination function. If I0 is the optical energy level at
VTTI - 5R
-------
the center of the film format, then the optical energy distribution
across the film format is:
I(x) = I (x)A(e,x)S(0,x)de (8)
0 0
where I(x) = optical energy distribution across the film
format as a function of wavelength X.
Io(x) = optical energy at the center of the film format
as a function of x.
L(e,x) = lens fall-off function (Cos12e)
A
S(e,x) = Solar illumination function
5(0,x) can easily be normalized to one yielding a relative energy weighting
factor. Camera lenses usually possess a spherical symmetry that eliminates an
integration factor which encircles the principal axis.
One will normally see a bright spot in or near the center of the
frame of film. Densitometer data are limited to a circle whose radius
is 1 cm about the center of the bright spot for a 4.5" by 4.5" format.
Data taken anywhere else in the frame must be normalized to this spot to
render correct indications from the calibration curve.
C. Water Body Characteristics
Areas of significant discoloration in water would be expected to
significantly influence the effectiveness of the calibration curve. Many
sources of discoloration result from municipal and industrial outfalls and
high turbidity or suspended solids. An investigation has been carried out
over a submerged diffused lignin sulfonate discharge located at Port Angeles,
Washington. In the true-color ektachrome aerial imagery, the effluent was
VIII - 57
-------
dark-gray- reddish-brown in color. Ground truth indicated no dissolved
oxygen depression within the area of influence of the resultant plume.
The blue and green densitometric data and ground truth data were used to
plot the respective straight lines in Figure 4. The difference curve
was also plotted in Figure 5. This curve was compared to the blue cali-
bration curve in Figure 3. As an example, if XA in Figure 5 were 0.125
corresponding to 8 ppm, then the equivalent value in Figure 3 would be
8.25 ppm. In the range from 8 to 9.5 ppm of dissolved oxygen, the corre-
lation was less than 0.3 ppm in spite of the significant discoloration.
Future Studies Requi red
To enhance the integrity of this technique, many questions must still
be answered. Several will be discussed in the next few paragraphs.
A. First it must be determined if the observations of green absorbance
are physically due to the presence of dissolved oxygen. Testing will
begin later this fall in the Environmental Physics Laboratory at NFIC-Denver.
An optical test cell whose path length through a particular medium can be
adjusted from 5 to 50 meters, has been designed and is awaiting fabrication.
It will require only 2.7 liters of water sample. Under laboratory conditions,
dissolved oxygen depressions can be induced into a particular water sample
and optically monitored to document its behavior in the green region.
This program will establish the optical properties of many types of natural
and waste waters as a function of dissolved oxygen concentrations.
B. The effects of water temperature upon the concentrations of
dissolved oxygen in the optical recording medium will have to be studied.
This will involve effort in both the laboratory and in the field.
VIII - 58
-------
C. The function of water depth and the optically generated concentra-
tion values must be determined. This can also be accomplished, for the
most part, in the Physics Laboratory.
D. Other recording media must be examined to possibly provide more
reliable concentration data. Kodak 2443 film with a Wratten 58 green
filter has been quite successful although in limited use. This technique
must be expanded into the domain of active and passive sensors. It is
worth noting that this spectroscopic technique has been carried out on
an aerial true-color ektachrome film SO-397. The results showed that the
Kodak 2443 film provided much better color separation because of its
false-color rendition.
E. Natural and induced interferences with the detection of dissolved
oxygen in all water environments must be explored. These include the
discoloration and suspended solids produced by man-made and natural sources.
Many of these can be effectively studied in a laboratory.
F. The calibration curve must be further tested and retested in
order to statistically validate its integrity. To date most of the ground
truth data/calibration curve correlations have been carried out in the
concentration range of 7.0 to 9.5 ppm. The curve needs to be studied
further in the range from 0 to 7.0 ppm.
Finally, the technique must be adopted to night aerial reconnaissance,
which will undoubtedly involve an active light source and absorption
spectroscopy. This will add greatly to a round-the-clock enforcement
monitoring capability.
VIII - 59
-------
SUMMARY AND CONCLUSIONS
A technique for the quantitative detection of dissolved oxygen con-
centrations in water through remote sensing has been discussed. The re-
cording medium has been an airborne framing camera employing Kodak 2443
False-Color Infrared film with a Wratten 16 (orange) optical filter.
Through a densitometric analysis of the exposed film together with ground
truth, a calibration curve has been generated. This technique has provided
an accuracy of better than ±lppm in healthy bay and ocean waters.
Future efforts to improve the technique under all conditions was
discussed. This will involve studying the influences of water temperature
and depth, discoloration and suspended solids upon the quantitative
results.
VIII - 60
-------
Acknowledgements
The author wishes to acknowledge the support received from the
Surveillance and Analysis Division, Region X, EPA, Seattle, Washington;
and the State of Washington, Department of Ecology, in obtaining ground
truth for the many flights conducted in the Puget Sound area.
VIII - 61
-------
References
1, Specht, M. R., and Needier, D., Fritz, W. L., "New Color Film for
Water-Photography Penetration," Photogrammetric Engineering,
Vol. XXXIX, No. 4, April 1973.
2. Mairs, R. L., and Clark, D. K., "Remote Sensing of Estuarine Circu-
lation Dynamics," Photogrammetric Engineering, Vol. XXXIX, No. 9,
September 1973,
3. Hale, 6. M., and Querry, M. R., "Optical Constants of Water in the
200 nm to 200pm Wavelength Region," Applied Optics, Vol. 12, No. 3,
March 1973.
VIII - 62
-------
1O-
8-
7-
6-
4-
3"
i
1.O
Figure 2
Characteristic Curve,
Background Water.
i
2.0
3.O
I
4.0
Film Density, X
VIII - 63
-------
1O-1
9-
8-
7-
6-
5-
4-
3-
2-
1-
Figure 3
Difference Curve,
Background Water.
-0.50
-0.25
I
O.25
I
O.50
A
0.75
1.OO
1.25
VIII - 64
-------
10"!
9-
7-
6-
5-
4-
3-
2-
Figure 4
Characteristic Curve,
Lignin Effluent.
1.O
2.0
Film Density, X
3.0
4.O
¥111 - 65
-------
1O-1
E
3E
3-
Figure 5
Difference Curve In
Lignin Effluent.
O.25 O.5O O.75 1 .OO
— O.50
— O.25
XA
VIII - 66
-------
A SEARCH FOR ENVIRONMENTAL PROBLEMS OF THE FUTURE
James E. Flinn, Arthur A. Levin,
and James R. Hibbs*
for
presentation at
EPA'S SECOND CONFERENCE ON
ENVIRONMENTAL QUALITY SENSORS
NATIONAL ENVIRONMENTAL RESEARCH CENTER
Las Vegas, Nevada
October 10-11, 1973
* Drs. Flinn and Levin are with Battelle.
Dr. Hibbs is with EPA.
VIII - 67
-------
ABSTRACT
Just as changes in key indicators of environmental quality
require monitoring and prediction through a network of physiochemical
sensors, so too must changes of a technical, economic, or social nature
which impact significantly upon the environment be identified, Effectively
information needs to be brought to the attention of appropriate agencies
early on new and rapidly developing processes, technologies, commodities,
or events which threaten serious environmental consequences. This paper
describes the results of a program initiated by EPA to serve just such
a purpose.
To identify future problems, a search was initiated of three
broad categories of human activity: (1) technical production activities
by reference to the Standard Industiral Classification (SIC) system,
(2) areas of environmental concern selected by reference to known areas
of EPA involvement, and (3) societal trends or changes relating to
current social, economic, political, and institutional events. This
approach resulted in the development of a preliminary list of problem
statements and stressors. Using a set of ranking factors, these state-
ments were culled to produce a list of ten "most serious" problems.
These problems were then projected 5-10 years into the future with
respect to magnitude, effects, and consequences.
Conclusions are presented regarding the efficacy of this
approach toward prediction of environmental hazards for EPA program plan-
ning purposes.
VIII - 68
-------
A SEARCH FOR ENVIRONMENTAL PROBLEMS OF THE FUTURE
Introduction
At any one point in time, EPA resources are allocated across a
spectrum of identified environmental problems. The planning process, which
results in such allocations, requires a continuous input of new information
on (1) the extent that these resources are yielding effective solutions to
existing problems, and (2) changes of a technical, social, or economic nature
which will create new problems or significantly magnify existing problems.
This paper describes some of the results of a study ^ aimed at supplementing
EPA inputs of the latter type, i.e., changes occurring now which will have
significant environmental impacts in the near future requiring EPA's
attention. Some of these future problems have their origin in current EPA
programs, e.g., implementation of environmental controls, while others are
a consequence of rapidly expanding production-consumption activities, the
introduction of new products, processes, or activities, or changing social
conditions.
In this study an effort was made to identify short (5 year) and
intermediate term (10 year) problems. It was a preliminary attempt at a
systematic examination of several categories of human activity from whence
such problems might arise. Problems of a national, as opposed to regional
or local, nature were sought although the tools employed are applicable
to regional or local use. The broad scope of the potential problem areas,
time limitations, and sheer necessity precluded more than a cursory review
of the activity categories selected. Consequently, the authors make no
claim to completeness.
Ten problems, considered to be "most serious" were selected from
a list of identified candidate problems for amplification with respect
to magnitude and effects. A priority ranking scheme provided the means
to select these. The ten problems identified are felt to be of national
importance within the next 5-10 year time frame, and serious enough for
consideration by EPA in assigning agency priorities in coming years.
VIII - 69
-------
Development of Candidate Problems
Three categories of human activities formed the starting point
for identifying candidate problems. These were
(1) Sectors of Industrial Production Activity
(2) Sectors of Environmental Concern
(3) Sectors of Societal Change and/or Trends.
These three categories were screened over a five week period employing the
literature, EPA and Battelle resource personnel, and other sources. Each
category was screened independently in a systematic manner. Thus the
Standard Industrial Classification (SIC) system was utilized as a reference
for the technical production category. Broad sectors (2 digit SIC codes
at most) were examined. Environmental sectors reviewed were (1) legislation,
(2) ongoing R&D in air, water, and land media, (3) health effects and
specific pollutants, (4) pollution control technology, (5) monitoring
and standards, and (6) transport processes. Societal sectors selected for
examination included demographic, crime, medical science, education,
social, economics, international, technological, government, urban, and
labor.
Each category yielded a list of candidate problems expressed as
a problem statement with a number of "stressors" identified. These stressors
were specific pollutants or classes of pollutants—chemical compound or
element, biological agent of physical substance—which have an adverse
impact on the environment. The repeated occurrence of a given stressor
in a list of candidate problems itself suggested a pollution problem of
major concern. For example, heavy metals and toxic organics were frequent
stressors in candidate problems.
The preliminary lists generated from the three independent
category searches were culled and combined into one single list of problem
statements. It was found that these could be grouped in the following nine
general areas:
« Pollution control residues
e Industrial production-consumption residues
e Energy supply
• Toxic and hazardous substances
c Air pollution
VIII - 70
-------
• Water pollution
• Ecological effects
• Radiation and sound
• Social.
Many problems of course related to more than one of these areas.
Selection and Ranking of Serious Problems
A set of nine factors each with a value scale of 1 to 5 was
developed for use in selecting and ranking ten "most serious" problems
from the candidate problem statement data base. The factors included
(1) Physiological risk (toxicity)
(2) Persistence
(3) Mobility/pervasiveness
(4) Bulk or volume of substance
(5) Number of people affected
(6) Relative environmental/ecological complexity
(7) Relative technological complexity
(8) Relative social/political complexity
(9) Research needs.
For a given factor the perceived worst condition was given the value of
5. Thus, the scale for persistence was as follows:
1 5
I 1 1 h—f
days centuries
It is noted that the first four factors are more related to ranking specific
substances (stressors) whereas the last five are more suitable to ranking
broader problem statements. Since each problem that was ranked has a number
of associated stressors, the person performing the ranking had to make a
judgement based on his perception of the relative importance of the various
stressors involved. Obviously while such judgements could detract from the
validity of the results, the use of such factors avoids totally subjective
and biased priority setting. An attempt to weight the nine factors did not
result in sufficient discrimination to merit the use of weighted factors.
VTTI - 71
-------
The nine factors were applied in two steps: the first of which
was to rank the initial list of candidate problem statements. Ten "most
serious" problems were derived from this ranked list. It was convenient
in several cases to combine two or more problem statements which ranked
high in importance on the ranked list. For this reason the ranking
factors were applied a second time, and by a different group of Battelle
professionals, to prioritise the resulting list of ten "most serious"
problems. The resulting ten problems in order of ranking follow.
Ranking Problem
1 Impacts of New Energy Initatives
2 Geophysical Modifications of the Earth
3 Trace Element (Metal) Contaminants
4 Hazardous and Toxic Chemicals
5 Emissions from New Automotive Fuels,
Additives, and Control Devices
6,7 Disposal of Waste Sludges, Liquids, and
Solid Residues
6,7 Critical Radiation Problems
8 Fine Particulates
9 Expanding Drinking Water Contamination
10 Irrigation (Impoundment) Practices.
Projections
It is apparent that the ten problems identified are quite broad
in scope. This is a natural consequence of aggregating related problems
with the objective of defining problems of national as opposed to regional
or local importance. In projecting future magnitude and effects, time and
budget limitations necessitated selecting only one or two facets of each
problem for analysis. Thus, facets selected for projection were (1) defi-
nitely near- or short-term future problems requiring a solution, (2) those
not already extensively examined by current scientific, social, or political
institutions (e.g., AEC's study of radioactive releases from nuclear power
plants), and (3) national in scope.
VIII - 7?
-------
Examination of these ten problems in detail revealed several
commonalities which, in themselves, could be the focus of EPA efforts.
For example, the multimedia nature of several of the problems is apparent.
For specific stressors, such as toxic heavy metals and organics, the extent
to which these are transferred from one medium to another is fairly well
documented. The adoption of pollution-control measures, resulting from
environmental legislation, is a contributor to intermedia transfer.
The need for more information on sources, pathways, and effects,
especially health effects, of stressors is an aspect common to several of
the problems identified. Similarly the impact of specific stressors on
ecosystems, while recognized, needs clarification in order that tradeoffs
between man's need for material resources and environmental controls can
be balanced.
A summary of the principal findings for each of the ten "most
serious" problems follows.
Impacts of New Energy Initiatives
Problem. Current concern exists over the possibility of shortages
over the next 5-10 years in energy resources. This concern is popularly
referred to as the "energy crisis". Whether the crisis is due to actual
resource limitations or the effects of institutional policies of the past
(more probable), one fact appears fairly certain, viz., the energy demand
of the U.S. is expected to nearly double between now and 1985. Meeting
this demand will require major technological efforts, all having significant
environmental impacts. The development of new energy technologies such
as coal gasification, coal liquefaction, and oil shale and tar sand
processes, is at an early stage. The impact of these will likely not be
fully felt until a time greater than 10 years in the future. It is
logical to examine these developments now with respect to likely environ-
mental effects and incorporate controls as the technology develops. And,
in fact, this is already being done as a consequence of increased public
awareness of environmental issues.
Of more importance are the short-term pollution problems that
will arise as a result of certain initatives being instituted today in
VIII - 73
-------
anticipation of meeting actual energy and environmental needs within the
next 10-year period. These problems include: (1) oil spill incidents
from current and projected massive imports of foreign oil; (2) radioactive
emissions and waste disposal needs from increased numbers of nuclear power
installations; and (3) increasingly large volumes of solid residues from
coal cleaning—needed to help meet SO emission limitations on coal powered
X
electric plants. Of these three problems, only oil spill incidents will be
projected here. The problems of radiation effects and disposal of pollution
control residues from nuclear power production and coal cleaning are covered
elsewhere. The environmental effects of coal gasification, oil shale, and
other emerging technologies lie farther into the future than this program
is aiming and it is assumed that their associated environmental consequences
will be dealt with as the technology develops.
Magnitude. Estimates of oil import quantities to supply the U.S.
energy deficit range from 7.2 million bbl/day in 1975 to 19.2 million bbl/
(2)
day in 1985 (Table 1). Transport of this oil will be by tanker ships
into coastal ports. The development of supertankers [-250,000 dead
weight tons (DWT)] and offshore deepwater ports for their operation will
occur in this period. Transshipment of oil to land in smaller tankers
(-50,000 DWT) from these ports is a possibility. Estimates of traffic
density increases in U.S. ports (with and without deepwater ports) range
from about 15,000 tanker trips/year in 1975 to a near maximum of 25,000
in 1980. Using methodology developed by the Coast Guard estimates
of annual incremental oil spill volumes as a function of oil import level
have been made. These are given in Table 2 for three levels of daily oil
imports. Using the worst case data of Tables 1 and 2, it can be shown
that approximately 800,000 bbl of oil would potentially be spilled in U.S.
coastal and harbor zones within the next 10-year period (1973-1983) due to
increased oil imports in the absence of superports. Insofar that superports
come into existence within this period, the amount would be expected to
decrease, unless all of the oil is transshipped from offshore superports
in tankers rather than by pipeline. The Coast Guard methodology used is
based on statistical analysis of 10 years of past accidental spill incident
history (1960-1970) as a function of spill volume. In addition to spills
VIII - 74
-------
TABLE 1. RANGE OF OIL IMPORTS (Million bbl/day)
Case* 1970 1975 1980 1985
Best 3.4 7.2 5.8 3.6
Worst 3.4 9.7 16.4 19.2
* Best case: Results from maximum effort to
develop domestic fuel sources.
Worst case: Results from continuation of
present trends in U.S. oil and gas drilling,
i.e., continued deterioration of U.S. energy
supply posture.
TABLE 2. ANNUAL SPILL VOLUMES AT VARIOUS OIL IMPORT LEVELS (barrels)
Import Level
(million bbl/day)
3
6
9
Case I
No Superports
15,900
32,600
49^400
Case II
Superports*
3,400
6,500
10±000
Average number of
bbl spilled per
million bbl/day
imported 5,400 1,100
* With transshipment by pipeline.
VTTI - 75
-------
due to accidents, operational spills due to leakages, poor housekeeping,
intentional dumping, etc., can be expected to increase. In 1971, these
amounted to nearly 9 million gallons in U.S. coastal and inland waterways.
Geophysical Modification of the Earth
Problem. In attempting to supply a growing demand for both
renewable and nonrenewable natural resources, man resorts to a variety of
technological measures which often have serious and widescale effects on
the environment. Examples include (1) the use of underground nuclear
explosions to release natural gas supplies, (2) strip mining for coal and
mineral ores, (3) dredging to create deepwater ports for oil-bearing
supertankers, (4) stream channel modifications, (5) ocean floor mining,
(6) pipeline construction, (7) silviculture practices, (8) deforestation,
and (9) the construction of dams for flood control, irrigation, and power
production. In each case, major physical disturbances of the earth are
made frequently over large land areas. These disturbances often cause
such immediate and long lasting ecological and sometimes irreparable
effects as: mine drainage, erosion, release of toxic substances, unsightly
and nonproductive terrain, species erradication, etc. It is illogical
to suggest that such practices be abandoned, since so long as man seeks to
upgrade his standard of living, the demand for materials and energy will
continue. The problem is to find ways of providing these needs while at
the same time minimizing the impact of the technological activity involved.
Magnitude. Within the constraints of this study only two near
term problems were examined in any detail. These were coal mining and
stream channel modification.
In a recent study for EPA Battelle estimated the volume of
acid mine drainage produced in 1970 from bituminous coal mines of the
Appalachian mining region as 103.8 billion gallons. In that year, 294
million tons of bituminous coal were mined in the region. National
Petroleum Council predictions for coal demand indicate an increase of
70-100 percent in demand (worst and best case) for coal by 1983. Assuming
mining of bituminous coal in the Appalachian regions grows accordingly,
VIII - 76
-------
the amount of acid mine drainage there would nearly double to 200 billion
gallons in 10 years if no preventative measures are taken. Another source
has estimated 500 billion gallons of mine drainage in the Appalachian
region, containing 5-10 million tons of acid and polluting over 10,000
miles of surface streams and 15,000 acres of impoundments. When
extended to all of the U.S. and expanded to include strip mining as well,
the amount of acid mine drainage from these sources is enormous indeed.
A recently issued study prepared for the Council on Environ-
mental Quality analyzes the dimensions of environmental effects from
channel modifications underway and planned by the Corp of Engineers and
the Soil Conservation Service. This is another activity where the areas
affected are large and the potential environmental effects are of profound
interest. The length of watercourses channeled by these two agencies by
1972 amounted to about 11,180 miles. By 1980, if projects started and
planned are accomplished, this will increase to 35,000 miles. Thus, a
study of the type conducted is timely with respect to the land area to
be affected in the near future.
The projects undertaken were for flood abatement, drainage of
wet land, erosion control associated with channelization programs, and
in rare instances, related to water supply, water quality, or recreation
needs.
The conclusions appear to be that while the purposes desired in
initiating a channeling project are generally achieved, there are potential
impacts which have far reaching implications on the ecology of the affected,
surrounding and downstream areas. Some of these include
(1) Reduction in wildlife habitats and food resources
through hardwood elimination
(2) Stream water quality impairment from water table lowering
(3) Reduction in aquatic organism productivity and diversity
of aquatic life
(4) Increased erosion (among most severe effect) and sedi-
mentation affecting turbidity, light penetration, algal
productivity and hence fragile food webs
(5) Enhanced downstream flooding.
VIII - 77
-------
Analysis of the benefits which offset these effects is complex as both
direct and indirect social, environmental, and economic values and costs
must be considered. In this particular setting, benefits were found to
outweigh the cost of such projects.
Trace Elements (Metal) Contaminants
Problem. Toxic trace element contaminants (principally heavy
metals), as a class, have already been implicated as a particular segment
of concern in the spectrum of identified environmental pollutants. For
example, beryllium and mercury have been officially declared hazardous air
pollutants and national emission standards established. Cadmium is on a
proposed list. Others such as selenium, vanadium, manganese, lead, etc.,
are under study.
The pathways by which trace element contaminants are distributed
to man are complex and encompass essentially all media and their associated
ecosystems. As with many pollutants where effects are manifest at low
concentration levels, control is difficult to exercise once the metals are
well dispersed along a pathway. The biological conversion to even more
toxic forms, e.g., organometallics, and accumulation in ecosystems further
complicate the problem. Interruption of such pathways at key points through
the application of control technology is a current major U.S. effort. The
sheer magnitude of the pathways and sources, however, combined with a
lack of information on the human health and biosphere effects of trace
contaminant deficiencies and overabundance, underscores the future impor-
tance of this pollution problem.
Magnitude. Trace elements, particularly the heavy metals, disperse
into the environment from a variety of sources. Major sources include:
(1) natural sources; (2) wastewater from metallurgical processing operations—
60 billion gallons in 1970-1971 from the lead-zinc-copper industry alone and
fa\
containing 0.01 to 25 mg/1 of specific contaminants ; (3) particulates to
the air associated with commercial-production consumption activities—about
500,000 tons/year can be projected (Table 3) for 1983 for a selected 15
VIII - 78
-------
TABLE 3. PROJECTED AIR EMISSIONS OF SELECTED ELEMENTS
Element
Zn
Mn
V
Cu
Cr
Ba
B
Pb**
As
Ni
Cd
Se
Hg
Sb
Be
Totals
Tons of
Base Year*
151,000
19,000
18,000
13,500
12,000
10,800
9,500
9,300
9,000
6,000
3,000
900
800
350
150
263,300
Pollutant
1978
216,700
25,840
37,240
20,680
14,980
17,290
14,000
11,840
12,750
10,940
4,090
1,240
1,160
460
200
389,410
1983
273,000
31,720
58,370
24,070
17,800
22,860
17,690
14,370
16,990
17,500
5,050
1,560
1,560
550
260
503,350
(9)
* Data for 1968-1971 period.
** Excludes automotive sources.
VIII - 79
-------
elements assuming no reductions in pollution due to process changes or added
levels of pollution control over the 1968-1971 period; (4) contaminated
sludges and solid residues to land disposal from mining, beneficiation, and
resource extraction technologies—an estimated 3 billion tons/year in the
1978-1980 period*; and (5) trace-metal contamination from automobile exhaust,
disposal of municipal sludges, and area applications of agricultural chemicals.
Sources, pathways, and health and ecosystem effects especially
need further study in order to establish controls that have acceptable
economic and social tradeoffs between man's need for material resources and
environmental health.
Hazardous and Toxic Chemicals
Problem. Public attention has been focused often in recent years
on newly identified hazardous chemicals or classes thereof: diethylstilbestrol,
thalidomide, DDT, polychlorinated biphenyls, pesticides, phthalic acid esters.
What has been startling is that, before a warning had issued, the environ-
mental hazard had become so widespread as to seemingly preclude any immediate
or short-term remedy. Today's chemicals, initially synthesized to meet the
technical requirements of a new product, frequently become widely proliferated
among hundreds of products of unrelated uses. Thus, PCB's showed up in
paints, carbonless carbon paper, electrical transformers, coolants, etc.
The manufacture, distribution, consumption, and disposal of such products
introduce the associated chemicals into the environment along a variety of
incredibly complex pathways, many of which impact directly or indirectly on
man. E.g., in a study of street contaminants and their contribution to
urban storm water discharges , PCB's and seven pesticides were identified
as significant components.
Of all the chemical classes now in use, pesticides appear to pose
the most difficult pollution problem in the near future. Pesticides,
synthesized and applied for their lethality toward pests, are spread over
literally hundreds of millions of acres of the U.S. Many are resistant
* See description of these in later problem discussions on waste sludges,
liquids, and solid residues.
VIII - 00
-------
to biodegradation, toxic to man, and accumulate in plants and animals. The
effective control of this class of chemicals is fraught with social and
political consequences.
Magnitude. A recent study has analyzed in detail the manufacture
of pesticides in the U.S., using a broad definition of the term, viz.,
including rodenticides, insecticides, larvacides, miticides, mollusicides,
nematocides, repellents, synergists, fumigants, fungicides, aligicides,
herbicides, defoliants, dessicants, plant growth stimulators, and sterilants.
1971 production quantities were determined to be over 1.3 billion pounds.
About 275 specific pesticides are of current commercial importance and
perhaps as many as 8000 individually formulated products are marketed for
end use applications. Projected 1975 quantities are only slightly higher
than 1971, perhaps reflecting the current environmental concern and
legislation. There is little reason to believe 1978 will be greatly
different.
Crop dusting as a method of application is a big business. In
1972 agricultural crop-dusting aviators flew about 1.6 million hours, up
(12)
11 percent from 1971, covering a record 120 million acres. Justification
for this method includes factors like: faster application—one airman can
cover 100 acres in an hour compared to a day for a ground rig; heavy rains
sometimes keep tractors out of fields; one-twentieth less fuel is used in
air application; soil compaction which hurts crops is avoided. Unfortunately,
this method has environmental hazards: some error with respect to placement
of the chemical is likely; drift of pesticides to adjacent fields is trouble-
some; wildlife in the area can be affected; and the question of human effects
is always present.
In addition, pesticide usage in urban areas has grown rapidly.
The variety of product formulations available for application by the home
owner is large. Aerosol containers for this purpose alone amounted to 100
million units in 1970. Presumably, such uses contribute to the pesti-
cide concentrations observed in street solids from eight major cities in
the U.S. which were on the order of 0.00125 Ib/curb mile (including PCB
which was of higher concentration and the major constituent in most cases).
VIII - 81
-------
The specific pesticides identified were the more persistent ones such
as the chlorinated hydrocarbons (p,p-DDD, p,p-DDT, and dieldrin) and
organic phosphates (methyl parathion).
In addition, major spills from highway, rail, and river transport
vehicles have occurred wherein pesticides were involved. Fish kill data
from such spills are available which illustrate the geographically wide-
CIS)
spread nature of the problem.
It is concluded that the new methods of identifying hazardous
substances before they are widely proliferated in commercial use categories
are needed, along with criteria for establishing priorities for testing and
research.
Emissions from Automotive Fuels, Additives,
and Control Devices
Problem. The automobile as a major source of air pollutants is
currently receiving national attention. Strict standards have been legis-
lated on automotive exhaust emissions and these are leading to significant
technological developments ranging over new engine design and modifications,
development of catalytic converters for exhaust treatment, formulation of
new fuels and additives to reduce levels of legislated emission components.
The adoption of certain of these technological alternatives has already
occurred and others will be introduced within the next 2 to 10 years. The
number of automobiles involved, their relationship to major population
centers, and the complexity of the impacts that result from each technolo-
gical alternative suggests a need for careful evaluation before-the-fact
of the consequences which are likely to occur. Some indicators of
potential problems already exist.
Magnitude. Since 1960, new automobile purchases as a percentage
of registered vehicles have been at a rate of 9 to 11 percent a year. These
purchases have exceeded replacements by 2 to 4 percent per year, indicating
a. significant growth in the number of autos on U.S. roads. At the present
rate, by the year 2000 the number of autos will about equal the number of
people in the U.S. (-250 million).
VIII - 82
-------
Gasoline consumption has on the average risen steadily for
automobiles since cars became available. The rate of rise has been sharper
since about 1968. Total gasoline consumption by cars in the U.S. is
projected to be 25 percent higher in 1980 than in 1972, a fact attributable
both to a greater number of cars (and hence road mileage) and to increased
gasoline consumption per car.
The contribution of auto exhaust to smog in large cities has
been experimentally verified and, in fact, was apparent as early as the
late 1950's in the U.S. Public reaction to this problem led to the
enactment of emission standards for hydrocarbons and carbon monoxide
beginning in 1968 and nitrogen oxides beginning in 1972. These standards
in turn have resulted in the rapid development of catalytic control
devices and other approaches for use in meeting the standards.
There are some 300 fuel additives in use for highway vehicles
and the expectation is that the number will increase significantly in the
near future. Removal of lead from gasoline, since it interferes with
effective operation of planned catalytic control devices, requires reformu-
lation of the gasoline to compensate for the resulting octane rating and
antiknock property reductions. Inorganic additives being considered include
manganese, boron, nickel, and phosphorus. Changes in the organic makeup
of gasoline involve increases in branched chain alkanes, olefins, or
aromatics. Increases in aromatics appear to be the simpler approach, but
if this course is followed an increase in emitted polynuclear aromatic
hydrocarbons (PNA) can be expected.
Current catalytic converter designs (platinum and noble metal
catalysts) remove between 50 to 70 percent of the PNA. The remainder
would contribute to exhaust emissions. It has also been suggested that
the partial oxidation of PNA by the catalysts may actually form a more
(14)
toxic carcinogen than PNA itself. Phenol emissions would also
increase from the added aromatic components. It, too, has carcinogenic
activity, especially with respect to skin and the lungs. Other reported
increased emissions would be benzene, aromatic aldehydes, and nitrogen oxides.
The incorporation of catalysts as a means of meeting standards
is planned by most American and foreign automobile manufacturers. As with
VIII - 83
-------
most pollution-control equipment, there is a potential disposal or recycle
problem and perhaps emissions associated with the devices themselves. With
EPA's required durability of 25,000 miles for each catalytic converter and
an assumed annual travel of 10,000 miles per automobile, the need for
large-scale continuous disposal or recycle of catalytic converters will be
reflected 2-3 years after their installation, i.e., 1977-1980. An estimate
is ,that by 1985 nearly 120,000,000 Ibs per year of spent catalyst will
require treatment or disposal.
Emission of very fine metal-containing particles has recently
been observed during the operation of a vehicle. This would represent
a new contribution of fine particles to air pollution.
This problem is going to develop within the next 5 years as
control approaches are applied to over 10,000,000 new cars/year in compliance
with existing Federal statutes.
Disposal of Waste Sludges. Liquids, and
Solid Residues
Problem. Future problems related to the disposal of waste sludges,
liquids, and solid residues are a direct result of the relatively recent
passage of environmental legislation aimed at cleaning up the air, water,
and land environment. Two aspects, are apparent: first, with the addition
of pollution control devices to air and water emission points of
industrial, utility, and municipal processes, large volumes of liquid and
solid residues are or will be generated. Much of the material collected has
some potential value, although it may be some time before technology for
recovery of the values will become available with current institutional,
economic, and political constraints. Examples of these residues current
and foreseen include sulfate sludges from power plant SO scrubbing,
X
sludges from increased application of secondary treatment to municipal
wastewater treatment plants, inorganic dusts from add-on particulate
control devices required by industry and utilities (fly ash, e.g.), and
ultimate residues remaining from the processing of industrial solid
(frequently hazardous) wastes by contract waste disposal firms. Second,
VIII - 84
-------
large quantities of industrial, municipal, and utility residues that were
generated and routinely disposed of before the advent of environmental
quality control efforts have come under scrutiny with respect to their
potential for polluting (trace metals, BOD, toxicity, etc.) various media.
Past practices of dumping into watercourses, lagooning (near watercourses),
burial in uncontrolled landfills, etc., can no longer be employed. Affected
are sludges from drinking water treatment, dredgings, coal production
residues, and slimes from metallurgical and inorganic chemical production.
Because of these and other factors, the search for suitable
ultimate disposal routes has become a major preoccupation of the generators
of these wastes. However, the options for such disposal have diminished.
Magnitude. The magnitude of this problem is best illustrated
by reference to a limited number of specific waste items within the two
categories of (1) new bulk residues directly resulting from recent
pollution control legislation and (2) residues from continuing past
practices which present acute disposal problems as a result of new environ-
mental concerns.
Pollution control residues include these items:
(1) Sludges from throwaway processes developed for SO
removal from combustion flue gas—48 million T/yr
by 1980
(2) Increased sludges and brines from the adoption of
secondary and tertiary treatment of sewage by muni-
cipalities—11 million T/yr (dry solids) by 1980 for
secondary treatment sludges alone
(3) Flyash from increased coal consumption by utilities—
26 million T/yr by 1978
(4) Residues from treatment of hazardous military, industrial,
and nuclear power plant wastes—3 million T/yr by 1978-1980.
Other high volume residues historically a disposal problem will increase in
proportion to material needs of man. Examples are:
(1) Sludges from drinking water treatment—0.5 million
T/yr
VIII - 85
-------
(2) Mineral ore tailings from coal, copper, phosphate rock,
iron ore, lead-zinc, alumina, nickel, etc., extractions—
2-3 billion T/yr by 1980
(3) Dredging spoils from major port construction, stream
channel alteration, and pipeline development activities.
Controlled land disposal, as an acceptable alternative, requires more study
if the needs of the next decade are to be met.
Critical Radiation Problems
Problem. Exposure of man to radiation sources which are not of a
natural origin is increasing. The growth of nuclear power and electronic
technology raises this issue as one which will have an impact within the
5-10 year period.
Radioactive releases from nuclear materials required for military
and peaceful uses have been widely recognized as a serious problem.
International nuclear test bans have been formulated to deal with the
former. Releases associated with nuclear power generation, while signi-
ficant, have received much study - to the extent that these sources,
quantities, and effects have been projected and assessed before-the-fact.
Less well studied are the radiation sources associated with
accelerating use of a whole spectrum of consumer, medical, industrial,
military, and commercial devices and systems based on electronic technology.
Near term increases in the numbers of devices in use and in the power
output levels suggest a need to further evaluate the problem. Thermal
effects are largely known. The extent of nonthermal biological effects
is largely unknown.
This report focuses on electromagnetic radiation in the radio-
frequency range as an area within the general problem statement where
information is needed to keep pace with future technology trends.
Magnitude. There is a constantly increasing number of electro-
magnetic sources in this country in the radio-frequency range. [Radio
frequency (RF) is defined very broadly as extending from the extremely low
VIII - 86
-------
2 10
frequencies through the microwave range, i.e., from less than 10 to 10
Hz.] With increasing numbers of sources there are also increasing power
outputs per source at many frequencies from the proposed Navy Sanguine
Communication antenna at less than 100 Hz to radars at 10 Hz.
In general RF radiation is used by man in four ways: (a) as a
heating source, (b) as a detection method (radar), (c) as a communication
method, and (d) as a power transmission method. The relative importance
of the hazards from RF radiation sources used in these ways is probably
in the order listed.
Residential microwave ovens are probably the current most
important potential hazard source for the general population. This is
true because of the total number in use now and projected in the future.
They are becoming more popular as a food heating source each year.
Table 4 gives the number of units in use from before 1967 to the
present and the projected sales for the next 4 years. Although the power
level of residential microwave ovens is in general 600-800 W, much lower
than industrial units, there is a potential for 20,000 to 100,000 people
being affected by just a few percent of defective units among the present
-1/2 million installations. This is projected to increase more than 10
to 1 in the next 5 years. The commercial microwave oven market, industrial
microwave processing units, and medical diathermy equipment are other
sources of exposure.
Radar and other detection devices employing RF radiation con-
stitute a health hazard only under special circumstances (except for
possible low level effects). Radiation levels found in the vicinity of
high-powered broadcasting stations in most practical instances are
considerably lower than those usually associated with biologically hazardous
fields. However, close to high-powered broadcasting antennas it is
possible to atttain field strengths approaching the safe minimum.
The growth of radio broadcast stations appears to be nearly
linear. Table 5 shows the number of TV and radio broadcast stations in
theU.S.(17>
Power transmission, based on solar energy sources, while
theoretically feasible, will not become a factor in the next 5-10 years.
VIII - 87
-------
TABLE 4. RESIDENTIAL MICROWAVE OVEN INSTALLATIONS
pre-1967
1967
1969
1971
1973
1975
1977
Retail Sales
10,000
5,000
30,000
120,000
375,000*
840,000*
1,750,000*
Home Installation
10,000
15,000
65,000
235,000
860,000
2,260,000
5,260,000
* These estimates considered to be conservative.
TABLE 5. TOTAL BROADCAST STATIONS IN U.S.
Year
1945
1950
1958
1960
1965
1971
TV
6
97
421
562
674
892
Radio
930
2832
3310
4256
5537
6976
VIII - 88
-------
It is evident from the present knowledge of the RF radiation
environment, that there are specific hazards that present current dangers.
With safety devices and education no widespread danger to the public will
occur in the next 5 years. Low level effects should, however, be investi-
gated to remove the uncertainties expressed by some persons and agencies.
Fine Particulates
Problem. Chemically active and inert fine particulates emitted
to the air constitute a potentially serious health hazard due to their
retentivity in the human respiratory tract. This problem ranks high in
terms of (a) direct health effects on man, (b) multiplicity of sources,
(c) the relative persistence and pervasiveness of fine particulates once
they are emitted, and (d) the difficulty of control before, and mitigation
after, emission. Even with the best available control technology, which
will result in a significant reduction in total particulate emissions,
a major fine particle fraction will be emitted. The health hazard can
be quite out of proportion to the mass involved, whether the particulates
are chemically active or inert.
Magnitude. Fine particulate matter is defined as a material
that exists as solid or liquid in the size range of 0.01 to 2 microns in
(18)
diameter. The lower size limit of 0.01 micron is based upon consi-
derations of potential adverse effects of particulates on human health.
The upper limit is based upon the fact that the collection efficiency of
present control equipment deteriorates significantly below 2 micron
particle size.
A comprehensive review and assessment of various environmental
effects have been documented in an EPA report entitled "Air Quality Criteria
(19)
for Particulate Matter". The environmental effects described included:
(1) effects on health, (2) effects on visibility, (3) effects on climate
near the ground, (4) effects on vegetation, (5) effects on materials,
(6) effects on public concern, (7) suspended particulates as a source of
odor, and (8) economic effects. The first two are probably among the most
serious effects.
VIII - 89
-------
Projections of fine participate emissions by quantities and
/I Q\
sources have been made for the years 1975 and 1980. Two methods of
projections were used. Method I assumed that there will be no change in
the net control for each source, which would result in increases in emissions
in proportion to increases in production capacity. Method II assumed that
all sources will be controlled by 1980, and that increased utilization
of the most efficient control devices will continuously increase the
efficiency of control for fine particulates so that by the year 2000 the
control efficiency will reach the efficiency of baghouse control, which
is the best currently available method for fine particulates.
The projections indicate that the fine particulate emissions
from industrial sources through 1980 will remain at a significant level
(better than 3 million T/yr) even after the application of control to all
such sources.
A substantial portion of fine particulate emissions originates
from nonindustrial sources for which no practical control method is available.
These sources include mobile sources, such as automobile exhausts and tire
wear, forest fires, cigarette smoke, ocean salt spray, and aerosol from
spray cans. The emissions from these sources will continue to remain
substantially at the existing levels through 1980. The fine particulate
emission from these sources, excluding forest fire, in 1968 was estimated
at 2.46 x 106 tons/yr.(19)
Expanded Drinking Water Contamination
Problem. Drinking water for human needs is derived from the same
sources that supply water needs for human enterprises - industry, power
production, irrigation, and the like. These sources unfortunately are also
the receptors of pollutants from the same enterprises. Current USPHS drinking
water standards (1962) provide impurity limits for only a relatively few
pollutants under the classifications of bacteriological, physical, chemical,
and radioactive characteristics. Water meeting these standards is generally
accepted by the public as safe.
VIII - 30
-------
Several events of the past decade have raised new concerns about
the relative safety of- drinking water. One is that methods of detecting
constituents in water have become more sophisticated permitting even lower
concentrations of contaminants to be recognizable. Another is the
increasing awareness of the extent to which the environment is polluted,
i.e., new recognition of the sources, amounts, pathways, and effects of
specific pollutants. Coupled with the latter is the realization that the
numbers of chemical entities synthesized, produced, and utilized in the
U.S. have been increasing dramatically. Since these substances in most
cases eventually find their way into water supplies, a reexamination of
drinking water safety appears to be a necessity. The problem is inti-
mately tied |