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