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,  D.C. 20460
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $6.25

-------
                    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.
                                   1) . X o ,C
                             John D. Koutsandreas
                             Conference Chairman

                             December 10, 1973

-------
                    TABLE OF CONTENTS
                                                          Page
Preliminary Report Of The Second Environmental
  Quality Sensor Conference
Conference Agenda
                                                            IX
Conference Key Note Address
  Willis B. Foster 	xv

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
  Long Path Optical Measurement of Atmospheric
    Pollutants
    Andrew E.  O ' Keef fe ..... . .............. . ..............  HI  _  j_y

  A Review of Available Techniques For Coupling
    Continuous Gaseous Pollutant Monitors to
    Emission Sources
    James B. Homolya .....................................  jjj  _  22

 Session  IV -  Water Quality Sensor Development
   Insitu Sensor  Systems  For Water Quality
     Mea surement
     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 Major 	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
              NUV 4
                              1373
                                                   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
          Equipment and Techniques Division

          Deputy Assistant Administrator fo;
            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.  SOn 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.

                               7sk$  Voucfo
                               Irohn 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.

II.  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.

VII.  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.
                         Vll

-------
                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. Earth - 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
                               IX

-------
II.  10:30 a.m. - 11:40 a.m. - NASA-Langlev 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,  "Marquet'te 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

-------
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.  Holmes,  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
         Region II
         Region III
         Region IV
         Region V
         Region VI
         Region VII
         Region VIII
         Region IX
         Region X
         NFIC- Denver
                 - Dr. Helen McCammon, OR&D
                 - Mr. Francis  P. Nixon,  S&A Division
                 - Mr. Edward Cohen, OR&D
                 - Mr. Edmond T. Lomasney, OR&D
                 - Mr. Clifford Risley, Jr., OR&D
                 - Mr. Ray Lozano, S&A Division
                 - Mr. Alex Alexander,, OR&D
                 - Mr. 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
                             Xlll

-------
   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 of 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
                               xvi 11

-------
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 sollution.  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-IAS VEGAS  PROGRAMS




       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

-------
                          CONTENTS
Figures




Abstract                                           I~'4
Sections




  I         Introduction                             ^~->




  II        Airborne Data Acquisition                 I—a/




  III       Data Processing                          1-4.9




  IV        Aircraft                                 1-23





  V         Appendices                               I—*24
                             1-2                     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

-------
                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

-------
5322
                                     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 future 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

-------
     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







                              1-6                     NERC-LAS  VEGAS

-------
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

-------
222
                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

-------
532
            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 radiometric 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 'B' 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

-------
3EZ
               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
\™,'
(•••••^••MflH^B^BNMHI
1 AERIAL AND/OR
j SURFACE CATA
I ACQUISITION
.1
flEAR REAL
TIME ASSESS-
MENT. VISUAL
OBSERVATION'S,
VIDEO TV, ETC.
A
PROCESS, REDUCE
ENHANCE, ETC.
CURSORY INTER?.
r
J,
COMPREHENSIVE
INTERPRETATION,
ANALYSIS,
CORRELATION
A,
FINAL REPORT j
GENERATION j
1 I DOHM* I .$ Region |
°F
 Bi/i
M Bi-> 
,3 i. ft
> •• 0 S-l S-
^- : Q. CJ O
4-> CJ I— O
-Loorai nation '.JLUUKU i :,'/', !L'L'|
/
Report v.
1

Report

. , ^jti^A.^ • 1 J™
\v
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  MRS
                                                          SANTA MONICA. C ALIFORNIA
IA^«^_J
               NERC-LV

                                       12OO  6OO  O     12OO

                                             -J      I-
                                              SCALE IN FEET
Figure  2  - Santa Monica Tlutne

-------
                                FACING NORTHEAST
DATE: 8/8/73
TIME:~1330 MRS
SANTA MONICA. CALIFORNIA
                                         / •* •>
                                         (m:
Figure 3 - Canyons entering Santa Monica Bay

-------
                             SANTA MONICA
                                                     FACING SOUTHEAST
                                           DATE: 8/8/73
                                            I IIVI t .'Nl/ loOU  MHO SANTA MONICA. CALIFORNIA
Figure 4 - Construction activity in canvon entering LIICO
         Santa Monica bay.

-------
        DATE 7-4-73
       TIME ~1300 HRS
                                                                                                A) KAISER REFRACTORIES
                                                                                                B) PACIFIC GAS & ELECTRIC
                                                                                                \) OUTFALLS
Figure 5  -  Outfall  discharges  in Monterey Bay

-------
322
                                       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.






                                         '1-17                   NERC-LAS VEGAS

-------
                             Figure 6
H
I
H1
00
             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
/O 7CC
-I    MAGNETIC
\O TAPE
                                               ^o
                                                   14 TRACK
                                                   ANALOG
                                                     "QUICK
                                                      LOOK"

-------
5322
                                      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 zones in thermal plumes.   The




            7- and 9-track tapes then  go to the central computer facility where the




            data are  processed in the  CDC 6400.  From the information on  the 9-track




            tape an atmospheric  effects  table is  generated which corrects all  data






                                          I_19                   NERC-LAS VEGAS

-------
                                 Figure 7
          GROUND DATA INTERPRETATION STATION
            14 TRACK
                 TIME
                 VIDEOs,
NJ
O
ANALOG
 TAPE
                          A/D &
                        CONTROL
                                     I S I
                                     LEVEL
                                     SLICER
SEQUENCER
    &
COMPUTER
 CONTROL
                                     5 INCH
                                     FILM
                                     RCDR
            9 TRACK
            COMPUTER
             TAPE
                                     FILM
                      TAPE TO TAPE/TAPE TO FILM

-------
5322
            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.
                                          I _  21                  NERC-LAS VEGAS

-------
                                   Figure 8
 CONTOUR PLOTTING SYSTEM FOR  I R  SCANNER DATA
        B-26 AIRCRAFT
                                            GROUND STATION
H
I
to
                                                                      V(t.e)
      NOVA 1200
      COMPUTER
  7-TRACK
   TAPE
           PARAMETERS
 DATA POINT
CALCULATION
 PROGRAM
                   CDC 6400
                   COMPUTER
                     TAPE
 /DATA\T(x-y)
 (  POINT ]
 V TAPE J
                     GRID
                  CALCULATION
                   PROGRAM
                           PARAMETERS
 NUMERICAL
  APPROX.
 PROGRAM
                                                        PLOTTER
                                                       CONTOURING
CONTOUR
  PLOT

-------
$322
                                       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.
                                           I -  23                  NERC-LAS VEGAS

-------
                           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

-------
5322
                                    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

             Communications Equipment

             Control Equipment
70 feet

54 feet

18 feet

34,000 pounds

21,139 pounds

31,000 pounds

R-2800-71 ZOOO HP

800 gallons

240 knots

100 knots

Up to 25,000 feet

1200 nm

2

2

Includes DOPPLER, VOR,  ADF,  DME

Includes VHF, 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      (rnilliradians)

    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

5.75" L x 5.32" W x 3.00" H
2.7 Ibs.
       Power Supply
(Unit  3)
       Film Magazine     (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

-------
3EZ
                                 APPENDIX.B  (Cont.)
             Characteristic
             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 j»10
2.30 in.
                                           1-27
                   NERC-LAS VEGAS

-------
                        APPENDIX B  (Cont.)
2.   Aerial  Camera  (KA-?6)

    Primary Use
    Film Information
         Format  (in.)
         Width  (in.)
         Length  (ft.)
         Frames  (per roll)
         Magazine
         FMC 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   3
       5.6
4.5
  o
218
  o
1.5
 , o
12
3.5
  , o
       105~  74   41"  41   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 (40DO-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°
                        . o
                  @ 300 A  intervals
                    400 A  intervals
'I  - 28
                                                           NERC-LAS VEGAS

-------
 A
53S
                                 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 MonJ.tor 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,  16K  Memory

Phoenix 8BIT A to  D Converter

Sangara 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
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 (NOj 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 o"f
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 RAPS 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
 photomultiplier  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-NOX  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-isotqpe 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; N0xj 0-p CO and non-methane hydrocarbon analyzers receive a
daily zero and single-point span calibration.  Logs are kept of
instrument drift and mechanical problems.  In addition to these
routine and rather ordinary procedures, instrument zero levels are
measured in flight during each data gathering mission at
approximately 45-minute intervals.  Measured zero drift can
then be subtracted from the recorded levels.

SUMMARY
The state of air pollution instrumentation now permits the fabrication
of an aerial air pollution monitoring system to measure concentra-
tions of NO, NOx, 03, CO; non-methane hydrocarbons^ and other
parameters from an aerial sampling platform.  Special considerations
must be given to temperature and altitude stability in the selection
of this instrumentation.
Mention of trade names or commercial products in this report does
not constitute an endorsement or recommendation for use by the
Environmental Protection Agency.
                          1-37

-------
ParaiMtat


Particles > .3 fa
Visibility
NO
hu
X
OZ-IUC
CO
so2
rifsc
and
W.
Temp
Ucv Point
Altitude
# of Digital'
Minimum Ji Characters In
Measurement Inltiunxnt Instrument Detectable Lag Tine & Magnetic Tape Calibration Calibration Sampling Wara-Vp
Method Manuf scturer/Hodel Ranges Ronrcntratlon Response Tim Outnut Method Concentration Flow Rato Tlmi*


Individual
Particle
Counter
light scattering
Chcmllunlne scent
Cheml luminescent
Cheml luminescent
HDIR

Flame
'hotom.trtc
FID
rhermoelectric
Thermoelectric
treasure
Kee Industrles/ll
Hoyco/220
HRI/1550B
Thermo Electron
Corp. /I An
Thermo Electron
Corp./l«B
SEM/612B
Andros 7000
Heloy SA 160
HSA 11-2
Cambridge
Systems Mdl
137-C1-SIA-TH
Cambridge Systems
Hdl 137-CI-SIA-TH
Computer
Instrument Corp-
Hdl 8000
Continuous,
1, 2, 5, and
10 second
Intervals
10 channels


.05, .1, .25,
.5, 1.0, 2.5,
5.0, 10 ppm
full scale
. 05 , . 1 , . 25 ,
.5, 1.0, 2.5,
5.0, 10 ppm
full scale
0-200 pphm
0-20pphm
U-2pp!uo
20,50,100,
200 ppm
lOppro
0-5ppm
0-20ppm


0-30,000 ft


"seat - "l
.0005 ppm
.0005 ppm
•01 pphm
.Ippn.
.005 ppm SO,
SOppb


+40' to
20,000'
40.4% above
20,000'
250 IBB RT per particle

Variable R.T.
• 1 - 200 sec. "*"'
<2 sec
with modifications
(manual mode)
<2 sec
with modifications
(manual mode)
< 1 see
6 sec. < R.T. •< 10 sue.
15 see Lag time
4, 30 sec response time
20 second lag time
15 second response time


+20' dynamic
error
4
20
6
6
6
4
6
6
6
6
6
6
6
Zero Output
for zero source
InLcrn.il
calibration or
latex particles
of known siio
Internal
calibration
or Frcou 12
Sample gas
Bendlx
Calibration Inst-
Internal
Calibration
Sample Gas
Internal
Calibration
or Sample Gas
Sample Gas


Airport
Altitude


Pure Freon
1 ppm NO
in N2

1/J of full-
scale reading
lOppm CO in
N2
. Ippm SO-
in air
lOppn, C,H
In N2 J 8
LOprm CH, In ni



.75 l./«oc
1 ft3/mln
0 ft3/
-------
                         LIDAR FOR REMOTE MONITORING
                                 S. H. Melfi
                    U.S. Environmental Protection Agency
              National Environmental Research Center, IAS 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 L'.s Vegas, Nevada (NERC-LV).  This paper will discuss tlie research pj.ojjram
being initiated at NERC-LV.  The areas for which remote monitoring research
will be conducted include air, water and possibly terrestrial pollution.
     The remote  onltoring technique to be discussed in this paper is LIDARs
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 LIDAR 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

-------
r.
ROLE OF REMOTE MONITORING
                 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 X mercury line,  this background absorption does
not change significantly over 1 cm"1.   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)
                                           on/
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 o~  component shifted to longer wavelength,
a o  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 TT component corresponds accurately to the
peak of the absorption profile of  natural mercury, while the a components
are both well off on the wings of  the  profile.   Consequently we may use
the differential absorption of the IT  and a components as a measure of
the quantity of mercury present in the absorption tube (the TT component
becomes the probe beam, and the a  components taken together become the
reference beam).
     Another feature of the Zeeman effect provides a convenient means of
separating the TT and the a components.  At right angles to the magnetic
field,  both o components are linearly  polarized perpendicular to the
field,  whereas the IT 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
 OJ
z

 c j

 o» '
    f\ C
-^  a 0.5
 o
 «-  e
   o
5s
6 «*-
V)
C
O
.u
            204
    a.o
Hg ff- (I5KQ
          -54.0  -36.0   -18.0     0     18.0   36.0
                          Figure 1
                        1-47

-------
H

I

^
CD
           PHOTOMULTIPLIER
           INTERFERENCER
           FOR 2 537 A
FILTER
 VARIABLE PHASE
RETARDATION PLATE?
  LAMP MOUNTED IN 15 KG
V MAGNETIC FIELD
          LINEAR \ SAMPLE I f
          )LARIZER   PORT ^vJP
                                       AUDIO

                                      AMPLIFIER
    1 ABSORPTION
     TUBE, 800-1000'C
                   POLARIZATION
                   COMPENSATOR
                                       °2

                                      VALVEl
                                       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
                                v
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 a 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 OP THE SIGNAL
     The reproducibility of a measurement is perhaps the most important factor
in selecting any analytical instrument.  Various concentrations of mercuric
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 repxoducibility
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
8

16

28
Microliters
Sample
Size
20
20
40
40
70
Chemical
Form
Mercuric
Chloride
Thimerosal
Mercuric
Chloride
Thimerosal
Mercuric
Chloride
Signal
Average of
10 readings
0.513
0.471
0.766
0.741
1.089
Standard
Deviation
0.024
0.019
0.022
0.011
0.020
Standard
Error
0.011
0.008
0.010
0.005
0.009
70
Thimerosal
                          1.081
                               0.028
                                                        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 phenylmercurie 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 IT 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 af 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 Calculated
Average Signal
10 readings
0.766
0.741
1.089
1.081
0.957 0.754
1.341 1.085
1.306 1.085
%
Difference




+27%
+24%
+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-
 CM
^ ^
.E o>
 C" o>
2 £
   o
   -
£
to
c
o
L_
h-
     i.o
     0.0
             204
                 HgO"~(l_5KG)
                                             204HgO"»- (15 KG)
          -54.0   -36.0    -18.0
                                     0      18.0    36.0
                                                    XBL T31-105A
                              Fig. 3
                    1-58

-------
H
       r—PHOTOMULTIPUER
         .—INTERFERENCE FILTER
       \ \  FOR 2537A
                  VARIABLE PHASE
                  RETARDATION PLATEy
          LINEAR  \   SAMPLE
          POLARIZERA   PORT-
L ABSORPTION
 TUBE,800-IOOO°C
     -    MOUNTED IN 15 KG
   \ MAGNETIC FIELD
                                           \
                                                                                     AUDIO  AMPLIFIER
                                                                          \_
POLARIZATION
COMPENSATOR
                                                                   02
                                                                 VALVE
                                                                                                   XBL 736-857
                                                     Fig. 2

-------
         SESSION II




NASA-LANGLEY  SENSOR PROGRAMS




         CHAIRMAN




     MR. JAMES  L.  RAPER




        NASA-LANGLEY

-------
           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
Presented at the Second Environmental Quality Sensor Conference
                        Las Vegas,  Nevada
                       October 10-12,  1973
                              II -  1

-------
               MULTIWAVELENGTH LIDAR FOR REMOTE SENSING OF

                     CHLOROPHYLL a IN PHYTOPLANKTON



                            Peter B. Mumola

              Olin Jarrett, Jr., and Clarence A. Brown, Jr.

                     NASA Langley Res.earch Center
                       Hampton, Virginia  236.65



                           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 1969 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 2k years' experience  in
aerospace research.   Mr.  Brown received his  B.S.A.E. from Auburn
University in 19^8 and his B.S.M.E.  from Auburn University in 191+9.
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 LLDAR 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 a_ 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
AA
 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
 ex-itation and emission spectra of over U5 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
 Parkin-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 360 nm upward through the visible
 spectrum.   Both monochromators were set to 5 n111 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"' molar solution of Rhodamane B in
 ethanol.   Cross sections were then computed using these spectra and
 accounting for instrumental effects such as monochrociator 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 nm.   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
         2R
                                                 For single fluorescent
                                                 scatterer
or

 4

3=1
                                                 For four different
                                                 fluorescent scatterers
where
         I = optical efficiency of receiver
                                                          p
         A = effective area of telescope primary mirror (m )

         AA  = detector bandwidth (nm)

         AA  = fluorescence bandwidth (20 nm)

         8  = receiver field of view (rad)

         9  = laser beam divergence (rad)
          is

         n = density of chlorophyll a (molecules/m )
                                                     2
         a  - effective fluorescence cross section (m )

         P  = laser output power (W)
         A  = laser wavelength (nm)
          6
                                                -1,
         a = attenuation coefficient of water (m  )

         P    = fluorescent power received (w)
          X GO
and subscripts  f  and  I  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  tff) 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   (V
 rec  1
                                      0
                                            afl(V
n,
or
                                   =  C   nX
Therefore
                                                -\
                                             OO
                                             v  1;
                                         P   (A,
                                          recx
The elements contained in the  X  matrix are obtained from the cross-
section measurements previously described.   The laser power at each
wavelength can be measured and controlled.   Accurate knowledge of  a
at all the appropriate wavelengths is essential for quantitative deter-
mination of chlorophyll a concentration.  Since  af > a^  for all laser
wavelengths under consideration (^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  ct  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.

                    MULTIWAVELENGTH 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.k--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.  Experiments 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
VIMS.   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 shorn 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 Wo.  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 run) 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  -

-------
H
H
FLUORESCENCE DATA
     RECORDING
            oo
                                  EMISSION MONOCHROMATOR  EXCITATION MONOCHROMATOR
                                                                                     LIGHT
                                                                                     SOURCE
                      Figure 1.- Laboratory apparatus used in fluorescence excitation and
                                emission studies.

-------
         EXCITATION SPECTRA
  EMISSION  SPECTRA
 o


~0

 X

 z:
 o

 H
 O
 UJ
 CO

 CO
 CO
 o
 o:
 o

 UJ
 o
 z:
 UJ
 o
 CO
 UJ
 cc
 o
        .6
        4
        .2
        0
           .8
     r-Rhodosorus (Red)
Aphanothece /
(Blue-green)/
           400 480  560 640  540 580620 660 700 740 780

    EXCITATION WAVELENGTH,nm  EMISSION WAVELENGTH,nm
                                Eutrepia marina—
                                   (Green)
        0L
Chaetoceros
                                 (Golden Brown) /
           400  480  560 640  540 580 620 660 700 740 780

    EXCITATION WAVELENGTH ,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


C\J

o
~O
X
0
H-
Ljj
CO
CO
CO
O
o:
o
LL)
o
LU
O
CO
LJ
a:
O
Z)
_l


3.0
2.8

2.6

24
2.2
2.0
1.8

1.6

1.4
1.2

1.0
.8

.6

4
.2
0
                                        Monochrysis
                                           lutheri

                                        Chaetoceros
                                        Gymnodinium
                                           simplex

                                        Chrysosphaeropsis
                                           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.
                          IT - 11

-------
H
                     BUBBLE FILTERS
                    HEAT EXCHANGER
                        Figure  4.-  Schematic of the airborne multi-wavelength LIDAR system.

-------
                         POLISHED
                         ELLIPTICAL
                         CYLINDER
                                   LINEAR
                                   FLASHLAMP
                                  / 75 ST ALUMINUM
                                  /  f  s   f  /  f  S
Figure 5.- Cross-sectional view of the four-color dye laser used for
         fluorescence excitation.
                         II - 13

-------
H

I
                         Figure  6.-  Photograph of complete LIDAR package prior to helicopter
                                    installation.

-------
H

I

H
(Ji
               —   5
               ro
               !   4
               o
3   3
ol
 o.
 o
.o
o
                                                  "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.

-------
H
H
FLIGHT DATE - 25 JULY 1973
 ALTITUDE-100 m
 AIRCRAFT - BELL 204B
 LASER NO.  WAVELENGTH(nm)
              598.7
              454.4
              539.0
              617.8
       1045 EOT
       1200 EOT
                                2
                                3
                                4
                             START
                             FINISH
NORFOLK?/
                              o
                                        O
                                         A
                                 A
                                 O
                                                                    O
                                            o
                                                                          O
                                              A
                                                  A
                                               FLIGHT  LEG
                                                                                      O NW FLIGHT
                                                                                      ASE FLIGHT
CHLORO
O — rv>
A £j
12 ,11, 10 .9 8 68.7 54321
13 '14 15 16 17 18 1
9 20 21
                            Figure 8.- Flight 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:  AN 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 work at Virginia Polytechnic
Institute.  From 196l 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
1-921 ^
                              II  -  18

-------
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 (Refs. 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
Doming 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 field-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 h, were generated  from
the MOCS data.  The algorithms to use are:


     Green Band:          Y, =  11?J  — 2juL
and


     Red Band:             Y.  =  20"J  "  19 'J                        (2)
                           •J       T
                                  10, j


where  1^ i   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  attempt to
extract a spectral signature of the algae.   A  linear regression
equation was assumed of the form

                            Y. =  A.I.  .  + B.                         (3)
                            J     i 1,0    i
where  Yj  is the value of a given algorithm for spectrum  j , and  A.j_
and  Bi  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  Ai , the  slope  of the regres-
sion equation for each of the  20  spectral bands.  In this example, the
algorithm
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   (H) 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 h 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 nm.  The  algorithm  in Equation (U)  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 1^
spectrophotometer.  The spectral  signature  in Figure 5 looks very
similar to the inverted spectrum  of Anacystis marinus , a blue-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 lj-5-l) 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

-------
ur-welling signal from blue vater is  equal  to zero relative signal.  Each
data roint in Figure 10 is an average  of four spatial elements which
reduce the resolution to 70 feet by  lUO  feet.   A change  of one frame
count corresponds to an advancement  of lltO feet along the flight line.
For purposes of discussion, the  ocean  along the track has teen divided
into regions, designated Rl,  R2, 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 run)  remains rela-
tively constant.  In Region h 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.
ncwever, 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 Regions 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 fee-
starting about ^-0 n. mi. offshore and ending at the shoreline.  The simi-
larity between Figures 13 and 1^ 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 1^, could be uniquely associated with certain distributions of
the suspended matter.  If known concentrations can be associated with
this signature, then the distribution of 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 a 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 shoved 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-view
Swath width
Ground resolution
Aircraft
17,500 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
                        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
7.5 watts
.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 NASA Convair 990 and field-of-view of  MOCS
             (2 by 18 miles) over Clear Lake at 37,^00 feet.
                           A.  GREEN BAND
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
                408   438   468   498    528    558   588"  "618   648    678
                                   WAVELENGTH,  nanometers
                       i  i   i   i   i   i   i   i   i   i   i   i  i    i  i   i
      I	L
I	I
                1  2  34  5  6  7  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
      ISOCHRYSISGALBANA

 11 11111111 11111 11 i 1111111 11111 1111 11111 111 1111111 i I
300        400      500       600        700       800
                  WAVELENGTH , nm

 Figure 6.  Absorption spectra of sample phytoplankton species.
                      II  -  31

-------
                              \

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 vaste dump in New York Bight.
  The outer boundary  is 50% of the acid strength in the center of
  the dump.
                          II - 33

-------
                                  LONG  ISLAND
                       ^ROCKAWAY BEACH
                                   ACID WASTE

                                       TRACK 45-1
•*•!? >\ '••'£3f<&-f*f <•••,}* >C
•^^'.^.-^•^^•.•./N
^^•^1^
                                                ATLANTIC
                                                  OCEAN
Figure 9.   Sketch  of the ERTS I imagery (Fig. 7)  showing approximate
  boundaries of the plume, acid waste, and one track of the NASA C-5U
  aircraft.
                              II - 34

-------
200
180
160
140
120
100
RELATIVE 80
SIGNAL
60
40
20
0
-20
-&(]
r- GREEN
0 468 nm BAND DYE -.
D 543 nm BAND MARKER \ rROCKAWAY
O 603 nm BAND ;' , BEACH
PLUME . ? !; •
T BOUNDARY
- , '(,, ,,;,;,! ; 1
•; >ROCKAWAY
'•'•,, ... >,;?«- .' INLET
- ,;' ^ ,V| v .,,,.. u..(_ . , /" '"'\ ,'• ';
. ii' •' i i
-'• 1
^'''K,,.., , &
; ; •'':>- ^ i
^|W^^ ^/f^'
1 PI I R2 I R3 I R4 i P5 iR6i R7 I
i i i i 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 ^5-1  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
         160
RELATIVE
 SIGNAL,
 603 nm   12°
  BAND

         80
              40
             -40
                           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 U68-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, 12Q
     603 nm BAND

                     80

                     40
                         20 POINT AVERAGES
                    -40
                             I
I
I
I
I
                      -40    0     40    80   120   160   200
                             RELATIVE SIGNAL,468 nm BAND
                           240
Figure 121   Relative signals of 603-nm band versus U68-run band for
  data along track 1^5-1 in New 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
              -.4
                            EVERY 3nd POINT
                               J.
                               V
                                 +
 .4       .8       1.2      1.6
RELATIVE SIGNAL,468 nm BAND
2.0
2.4
  Figure 13.   Relative signals of 603-nm band versus U68-nm band  for
    data along 36° parallel track off Cape Hatteras.  (Compare with
    Fig- 11.)
                                 II -  38

-------
                 1.6 r—
                         20 POINT AVERAGES
                 1.2
                 .8
RELATIVE SIGNAL,
  603  nm BAND
                 .4
                  0
                 -.4
                         I  I
                                                             !  !  I
                             0        .4        .8        1.2
                              RELATIVE SIGNAL,468 nm BAND
1.6
   Figure lU.  Relative signals  of 603-nm band versus U68-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. 12.0
                              II  - 39

-------
SUMMARY OF NASA LANGLEY RESEARCH CENTER REMOTE SENSING ACTIVITIES

UNDER THE ENVTRONMENTAL 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 Langley's 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 Flight
 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
Langley'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.k-GEz 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,  1.kO mg/rn^.    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 IkO mg/m3 of chlorophyll.   Figure 7 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
197^.

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 1J 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 1^ 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 2if-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 16 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 optijnum 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 NERC) 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 VjO-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

-------
IE scanner are planned.   The potential for application of remote  sensing
to this problem is fairly obvious} 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.

                            ACKNOWLEDGMENT S

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
      FROM 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
      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

I

Ul
o
  CHANGE IN
 BRIGHTNESS
TEMPERATURE,
    deg,  K
                           10
                           15
                                                                         1.4 GHz
                                                                         t-BAND
           Figure 2.  Microwave radiometer, change in brightness temperature with change in salinity.

-------
     CHESAPEAKE BAY       3
     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

     • 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
I

CO
                                 Figure 5-   Patuxent River multispectral scanner data.

-------
                  V
                            .. .r
H
                                                                     .,. ,•....     .,.•   .,.
                                Figure 6.   Patuxent River — training set taken at boat.

-------
H
H
Ul
Ul
         \  ,y  v''
           T^-,.1
.r
              Figure 7.  Patuxent River - training set taken at entry of upper  left  stream  into river.

-------
U1
CTi
                                      -•'  *-
                               V.
                                                I..
                                        .. ->•*
                                            r >
                                                  ••'	^'^V::...*.,. ••••-"'"
              Figure 8.  Patuxent  River - training set taken in inland body of water behind boat.

-------
Ul
-J
                              . ••• ..-•»
                 Figure  9-   Patuxent River — training  set  taken in the river to the left  of the boat.

-------
                                      •K
H
H
I '
on
              Figure 10.   Patuxent  River — training set  taken in the water at  the left  of the river bend
                                                       sand  bar.

-------
Ul
             Figure 11.  Patuxent River — training set taken in the water in the center of the river  at
                                                    left of boat.

-------
     OBJECTIVE -
     STATUS  -
H
H
     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.  ERTS — 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   - 10%
           •  SIZE -  10 ACRES

PROBLEMS - •  EDGE EFFECTS
           •  NON-HOMOGENOUS
           «  UNIQUENESS

OUTLOOK -  •  FURTHER  LARS  TERMINAL ANALYSIS
           •  PROGRESS REPORT  - APRIL ' 74

COORDINATION-EPA, REGION III HEADQUARTERS

                     Figure 14.  ERTS - landfill classification.

-------
      OBJECTIVE -
H
H
cr>
co
       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 SENSING - AIRCRAFT AND SPACECRAFT
     b) IN SITU  MEASUREMENTS
     c) LAND RECORDS

  TASK 2. DATA ACQUISITION AND ANALYSI S (4 MONTHS)
     a) IDENTIFY LAND USE ACTIVITIES
     b) LOCATE AREAS AND POINT SOURCES OF POLLUTION
     c) TOPOGRAPHY DISTURBANCES
     d) DRAINAGE PATTERNS AND POLLUTANTS

  TASK 3. DEMONSTRATION PROGRAM PLAN (2 MONTHS)

     a) OPTIMUM SURVEILLANCE PROGRAM
     b) CONTRACTOR'S REPORT

              Figure l6.  Contract effort.

-------
H
H

I


Ul
W. VA.
                                             SCALE IN MILES
                                         5    0    5    10   15   20
Figure 1?.   Mine drainage pollution study — north branch Potomac River.

-------
       OBJECTIVE -
   TO DEFINE AN APPROACH FOR AUTOMATING THE CONVERSION OF
   IN-SITU 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
       OUTLOOK -
CTi
(T.
•  CONTRACT AWARD BY JANUARY ' 74
•  COMPLETION OF SYSTEM  DEFINITION  TRADE  STUDIES
•  START SYSTEM CHECKOUT NOVEMBER ' 74
                                                                    MARCH ' 74
       COORDINATION  -  EPA,  RESEARCH TRIANGLE  NERC
                       Figure 18.  ERTS — data automation and transmission.

-------
H
SENSORS
   •   SEVERAL SENSORS. ARE REQUIRED TO  DETECT POLLUTANTS AN& MONITOR
       METEOROLOGICAL CONDITIONS
DATA VALIDATION
   •   ROUTINE CALIBRATION
   •   DETECTION AND CORRECTION OF  SENSOR FAILURE
CALCULATIONS
   •   CONVERSION AND AVERAGING FOR EPA  STANDARD
DATA AVAILABILITY
   •   ON SITE
   •   AIROMETRIC 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
(Ti
03
                           r
                                MANUAL
                                CONVERT. REPORT.
                                  VALIDATION,
                                 ZERO/SPAN CAL

SENSOR


SIGNAL
rriMn i
L-UIMU 1
TIDMIMP.
1
1
1

CHART
RECORDER
               PRESENT SYSTEM
PROPOSED SYSTEM ADDITION
                                           I	
                                               A/D CONVERTER.
                                                  SCANNER
                                                  CLOCK
                 LIFO
              DATA STACK
        TRANSMISSION
            MODEM
                                       DATA
                                      HANDLING
                                      SYSTEM
                        r
                   24v
                   in
          ELECTRONIC
             UNIT
                        L_
                                                                           OFF
CLOCK

                         Figiore 20.  Data flow at  camp and proposed system addition.

-------
            THE  USE  OF WEAR-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  2lU01

                         BIOGRAPHICAL SKETCH

Mr. Bressette is a Senior Research  Scientist vith 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 vater  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 a_ concentrations that varied from k to >3000 yg/&,  revealed
a phytoplankton  "bloom" threshold in  the near infrared between the
concentration of 3^ and 51 yg/£.

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 l6 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-51* 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 summed 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 Hasselblad 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  AND 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 k.   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 17
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
1^ as chlorophyll a_-containing phytoplankton.

The positive prints that are seen in figure k 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 h with figure 5.  In figure 5 are presented  essentially
the same pictures as in figure h, 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 h, 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
                                   - 73

-------
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 kO yg/5- which  you  can see fits
the data reasonably well.

Observation of the faired values of R between the  values of  chlorophyll a
concentrations of h and 31* Ug/£ 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  a_ concentrations
over the range of k 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
of sun altitude angle">° 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 U to 3^ yg/£  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-"-0 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 HO yg/£ could provide information at one of the two wavelengths
required for the measurement of suspended sediment .I'-'  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^ yg/& in figure 7, R increase
and the S decrease, if any, is slight.  Thus with chlorophyll a_ concen-
trations above 3^ yg/£, 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 3^ and 51 yg/& 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 H-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 a concentrations that are  below the threshold concentration for
producing phytoplankton "blooms."  The largest percent of "bloom" area
is between the chlorophyll a increment of kO  to 60 yg/Ji 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 a_ concentrations that varied from h ug/& to >3000 yg/£.

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 yg/£.

U.  The upwelling of NIR light through the  surface of  the  water is inde-
    pendent of chlorophyll a_ concentrations below a value  of 3^ Ug/£.

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/&,  and remained  nearly constant
    with further increase  in chlorophyll a_ 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

-------
T.   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

l-OcLum, 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.

•^Spiess, 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, irth Annual Earth Resources Program 'Review,  Volume IV,
Presented at the Manned Spacecraft Center, Houston, Texas,  January  17-21,
1972.
      , 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,1 for Scientific and
Technical Use," Twenty-second Edition, First  1966 Printing.
     , J. E. , and Martin, M. J. , "The Photographic Process,"  First
Edition, McGraw-Hill Book Company, Inc., page 211.

°Piech, 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.

^Payne, R. E. , "Albedo of the Sea Surface," Journal of the Atmospheric
Sciences, Volume 29, July 1972.

10Williams, Jerome, "Optical Properties of the Sea," United States  Naval
Institute, Annapolis, Maryland,  1970.
                                -  II - 77

-------
                              Table 1.  EPA,-1- Mid-Potomac Estuary Phytoplankton  Surveillance
                                        Data, Surface Samples, October 2,  1972
H
H
I

00

Sta. No.
1
2

3
k

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
Time3
I*tl5
llt08

1^03
1355

13U6
1337
1328
1320
1306
1253
Cond.
Tide umhos
Slack 12.1
Slack 10.5

Slack 10.5
F 10.3

F 9-8
F 8.6
F 8.0
F 7-1
F 5-7
F It. 2
Salinity
°/oo
7-5
6.6

6. U
6.7

6.0
5.U
U. 8
it.lt
3. H
2.7
Temp.
°C
21.0
20.5

21.lt
21.H

21.3
21.it
21.2
21.2
21.7
20.9
DO
mg/£
5.8
6.0

6. it
	

6.0
—
6.5
6.9
7.0
7-0
Secchi1*
Disc-in.
l.itit
1-52

1.80
1.18

-67
1.23
-93
-95
.77
.82
Chloro-
phyll a
ugm/£
007.5
930.0

3>t.5
16.5

10.5
it.O
10.0
9-0
332.0
6.0
Wind
dir .
N
N

N
N

N
N
N
N
W
S
Wind
vel. ,
mph
0-2
0-2

3-6
3-6

3-6
3-6
3-6
3-6
3-6
2-5
See
state
Calm
Calm

Ripples
Ripples

Ripples
Ripples
Ripples
Ripples
Ripples
Ripples
R5
0.9^
• 7U

.87
• 78

.90
• 77
• 78
.80
It. 58
• 9U
        EPA,  Annapolis Science Center, Annapolis, Maryland  2lit01.
       2Coast and Geodetics Survey Navigation Map #559-.
       •^Eastern daylight time.
        The visibility depth of a 30-cm-di-ameter 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
Location^
Buoy 22
Off
Time3
1242
1230
Tide
F
F
Cond.
ymhos
3-20
2.75
Salinity
°/oo
2.1
1.7
Temp.
°C
20.7
21.3
DO
7-5
7-6
Secchi
Dis-m,
0.82
.46
Chloro-
phyll a
ugm/£
57
>3128
Wind
dir .
S
S
Wind
vel. ,
mph
2-5
2-5
Sea
state
Ripples
Ripples
R5
„ 	
18. ho
Potomac Cr.
13
14

15
16
17

18

19
20
N26
Off
Acquia Cr
BT30
N34
Off Va.
Shore
N40
Mallow ' s
Chi
Buoy hh
1222
1210
.
1201
115*1
1146

1137
Bay
1130
1120
F
F

F
F
F

F

F
F
2.45
1.85

1.80
1.25
1.00

• 75

.ho
.25
1.5
1.2

1.0
.85
.6

.5

.2
.1
20.5
21.3

20.2
21.0
21.1

20.8

20.7
20.65
7-5
7-5

7.5
7 « h
7 • h

7-1

6.7
7-3
.59
• 72

.62
.67
.6h

• 72

-72
.59
783
1192

79
328
"51

559

76
274
S
S

S
S
s

s

s
s
2-5
2-5

2-5
2-5
2-5

2-5

2-5
2-5
Ripples
Ripples

Ripples
Ripples
Ripples

Ripples

Ripples
Ripples
6.2h
5-80

2.38
4.30
1.56

6.19

2.80
4.52

    -'-EPA, Annapolis Science Center, Annapolis, Maryland  2lit01.
    ^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
H
CO
O

Camera
1. Hasselblad

2. Hasselblad.

3- Hasselblad

h. Hasselblad
Focal
length
(mm)
ho

ho

ho

ho
Filter1
58 (green)

12 (yellow)

89B (NIR)

HF-5 (haze)
Film
format
(mm)
70

70

70

70
Film type2 AEI3
2U02 Black & 6
White
2h02 Black & ho
White
2h2h Black & 28
White NIR
SO-397 Color 12
Speed1*
(sec)
1/250

1/250

1/250

1/250
f^
number
5.6

11

8

5.6

      1Kodak Wratten filter number.
      2
       Kodak film number.
      -%odak recommended aerial exposure index.
      14
       Actual exposure.

-------
Chlorophyll a_ concentrations,          Area covered,
            yg/£                         percent
1000
23
36
18
6
5
3
6
                      II  -  81

-------
H
H
CD
M
                           1/e ATTENUATION IENGTH OF DISTILLED WATER
                                            (REF. 3)
      REFLECTANCE
                                              REFLECTIVITY OF
                                               CHLOROPHYLL
                                               PLANTS (REF. 2)
         COLOR-I V I  B   G
,6      .7     .8
Y   0   R   «-|R
       WAVELENGTH ,
                                          30
                                          20

                                           ATTENUATION
                                            LENGTH , m

                                          10
                                                    .9     1.0    1.1     1.2
                                                                           0
      Figure 1.   Comparison of the reflectance of chlorophyll-containing plants with the attenuation
                              length of sunlight in distilled water.

-------
          N
H
i
oo
CO
/-WASH.
\ D.C.
/>
/
/
A

^!H
>G^/
5UNSTON COVE
*YLAND POINT
^
UPPER REACH
•
A
B

r
v/

E
F
G
H
|
•
J
"SALT WEDGE
   AREA"
                                LOWER REACH
       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

I

03
       TRANSMRTANCE,
               %
100

80
._ FILM
CUT OFF~*J
"1

/

                           60
                          40
                          20
                           0
                            .4      .5
                   COLOR - I V I  B   G
                                         12 FILTER
58 FILTER
                                               A
                                                       89B FILTER
               I
                                                                     IR FILM CUT OFF
.6    .7      .8
Y   0   R   «-IR
      WAVELENGTH ,
                    .9     1.0     1.1     1.2
              Figure 3.  Filter selection and film cutoff for detection of phytoplankton.

-------
H

I

00
cn
                                                                                       22
        Figure h.  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.

-------
                                                                                            •     \
                                                                                            R   '*
00
en
10
II
12
13
14
15
           17
        Figure 5.   Essentially the same series of pictures as in figure h  only the development time
                  for the  positive prints is reduced in order to define the  land  features.

-------
H

I

00
        14


        12


        10


         8
       R
         2


         0
    N __
      ^B
B
 PHYTO-x
PLANKTON
 RADIANCE
                          BOTTOM
                          RADIANCE
                   0
                                        	 DENSITOMETER TRACE
                                        	RIVER BOTTOM DEPTH
                                         A  UNEXPOSED FILM
                                         B  LAND AREA
                                         C  WATER AREA
                                      DEPTH,
                                      meters
                                                             89B NIR
                                                             FILTER
                                          DENSITOMETER TRACE
                                X, km
                                                            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
CO
00
         10 r-
     R   1
         .1
                        D..

                           D
   D
                                 D
                    	fl.__n    :
                             r-tL-T   	r-|-=
IT
        H   h-
THRESHOLD FOR
PHYTOPLANKTON "BLOOMS"
                                                                            10
                                              METERS
                                10                   100

                         CHLOROPHYLL a CONCENTRATIONS,  pg/*
                                        1000
        Figure T.  Relative film transmittance, R, from densitometer traces and 30-cm Secchi disk
                        depth, S, versus measured chlorophyll a concentrations.

-------
H
H
00
         DISTANCE Y, km
                                                                      CHLOROPHYLL a. ,
                                                1         2        3
                                                  DISTANCE X, km
                                                                                500
                                                                              r-  100
                                                                                  50
                                                                                  10
      Figure 8.  Distribution  of chlorophyll a_ concentrations  over l6 square kilometers of the Potomac
                   River obtained from densitometer traces of picture 6  in figure U.

-------
            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,  participate, meteorological,  and radia-




tion parameters.  Distribution of  the stations  is on four concen-




tric circles,  centered on the arch,  and at  distances of h, 9, 20,




and kO 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 -  3

-------
                            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 vill 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.
                ft + FIUC  ^+Rc,T)+S (x,t
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,




                              III  - 4

-------
H
H
I

(Jl
       SOURCES
Fuel  Combustion

     Industrial

     Steam Electric

     Residential

Transportation

     Eoad Vehicles

     Aircraft

     Railroad

     Vessels

Solid-Waste Disposal

Process Losses
                              TRANSPORT
                                    Diffusion

                                    Portage

                                    Turbulence
                              TRANSFORMATIONS
                        -=>
Solar Radiation

Chemical Reactions

Decay
                             AIR Q.UALITY
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.

      7.  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 h,  9, 20, and Uo  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.

     ^.  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 tower 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

-------
      U.  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



81HO.



      Each station will have a NO  box capable of direct NO and NO
                                 x                                x


determination and NOp inference.   The probable instrument is the Moni-



tor Labs 8UUO.



      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

-------
HI - 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, ^75, 530, 570, 630, 695,




and 780 nm cutoffs.  Pyrogeometers measuring 3-50 pm 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 ASE-33 tele-




type.  Addition of new equipment is easy because 
-------
      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 Research 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 Air Pollution Study - An Overview.   APCA
      Annual Meeting, Paper No. 73-21.

Pooler, Dr.  Francis.   Network requirements for the  St.  Louis Re-
      gional Air  Pollution Study.
                          Ill  - 15

-------
RAPS/RAMS  PROPOSED  STATION  LOCATIONS
              ST*T1D*S AHt WDABGEIt  ri CONCENTRIC CIRCLES
              »»CH«D  STATIdH  lot  AT DISTA1CCS  OF «,  9,
              W.  AND 40  rlllWTEftS.  THt STA«i  ART  AT
              THE  AppsoiiPWTt cnntM or  nt  ARIA WDER
              CfWSIDCUTlOH WIG] KG FROM  LES5  THAU A
              lllO*Tr*'S M01US AT  THt  CUTML  STATIC"
              TD MOsn  *  S  mO»CTEl RADIUS  AftOlMD  THC

-------
              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 value 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

-------
svstem 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 ©Deration, the water



condensate is continually removed and the probe filter is



periodicallv back-flushed with compressed air to remove ^ntrained



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 (4) leakage of sample Hnes.  The



Environmental Protection Agency has contracted an investigation




                             III - 23

-------
into these problem areas.1  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 samp IP 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* S0? and NO  analyzer.  Stack gas is passed
                ^       JV


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 SC^ measurement with the 436 nm wavelength selected for



N02 measurement.  Since nitric oxide has little absorbance in



the visible and ultraviolet, conversion to N0? 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~, NO-, 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.





                            Ill - 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-f low/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 S0? 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
                           o /
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 permeatine 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 rceirVrane materiel for SCL and NO dilution.  In  addition i,



polymer tubes have been substituted for the membrane.  In  tl iis



manner, the tube is enclosed in a temperature controlled c''h. amber



and the sample stream passes over the outer surfaces of the



tube with carrier gas flowing through the tube.  The desirt id



dilution ratio is dependent upon:  (1) the permeability of  the



membrane or tube to the component of interest; (2) the sui .-face



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  manufac :turers.



However, the permeation samplers still require; extractio n of a



source sample which must be filtered and held at an elev fated



temperature prior to entering the diffusion chamber.



     Recently, a mechanical device has been developed i n our



laboratory to quantitatively dilute a source s.ample 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.  Ifhe disc  is sand-



wiched between two stationary discs having sample  inle.'t ports to



allow gas exchange between the sample chambers and tho  st;ack 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 dil.uted samp If.» stream



is then analyzed by an ambient air analyzer.  The  dil .ution




                            III - 27

-------
ratio  'depends upon: (1) the number and volume of sample chambers;



(2) t h<2 rotational speed of the sample disc; and (3) the volume-



tric lr'Low of the diluent gas.  Operation of the sampler has been



demons  trrated in the field by the continuous analysis of the S02



emissic 5ns from a 190 megawatt pulverized coal boiler.



     Fi gure 7 represents a typical 24-hour segment from a week's



continui ous operation during which the "disc diluter" was coupled



to a cor  iductometric ambient SO^ monitor.  The resultant SQ^



emission s, based on a dilution ratio of 1600:1 are plotted



against 1 -he net load,  in megawatts,  from the boiler turbine



generator .  The diluter/analyzer combination appears to follow



the trend, s in power output quite consistently.  Further develop-



ment of tl:  ie dilution system is being carried out under contract



to determi ne its range of applicability.   Also, design modi-



fications 1 have beeni incorporated to allow in-situ calibration of



the diluticm head.





IN-SITU MON ITORING



     In-sta(:k measurement avoids any extraction of sample by



utilizing tl ie sample? stream itself as an analysis chamber.  These



instrument  systems  employ electro-optical detection which can be



arranged in  three differing configurations.



     A foAded-path  design places the energy source and receiver



at the sam.e  location.   In this manner the energy beam enters the



emission st ream through a slotted probe and is reflected back



into the ins'trument.   For large stack or duct diameters, the path-





                                Ill - 28

-------
length of measurements might 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 pathlength 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 SCL 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 Power Company's



River Bend Steam Station in Charlotte, North Carolina, from



January 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 S02, NO, and


NO •  the CEA Mark IV in-situ S09 system;  and the Bailey Meter
  x'                            z

Company S09 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 Ester line-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 SO,, 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 "read"


position,  light is directed into the probe and the probe mirror


directs the light beam back into the spectrometer.  If S02 is


present in the probe slot, it will absorb energy in regularly


                            III  -  30

-------
spaced bands at 3025A.  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 SC^ 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 SC^ 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 bv 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 SO^ 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 pine is provided with a metered flow of purge air to



each housing to maintain a definite optical pathlenpth 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 closely



spaced such that the extinction coefficient of particulate matter



remains the same.  Therefore, the output sienals represent the



averaee SO- 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 SC>2 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 SC>2 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 NO,, 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 1573.

2.   Rodes,  C.  E.  "Variable Dilution Interface System for
     Source  Pollutant  Gases," Proceedings,  Analysis Instrumenta-
     tion, 11,  125-128 (April 1973).

3.   McKinley,  J. J.   "Permeation Samp ling--A Technique for
     Difficult  On-Stream Analyzers,"  Proceedings,  Analysis
     Instrumentation,  10. 214 (Mav 1972).

4.   Rodes,  C.  E.} 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  Research 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





Fipurp ].   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.   SO^ Concentration,  ppm vs Net Load, Mw, March 17,



            1972.



Figure 8.   CEA-MK IV In-Stack  S02 Monitor.



Figure 9.   Bailey S0« Source Analyzer.



Figure 10.  SO^ Concentration,  ppm and Stack Opacity, 7<>,



            0900-1400, March 1, 1973.
                            Ill - 36

-------
 VENT
             MONITOR
             •  ooo
                                                                                    MOUNTING
                                                                                    HARDWARE
                                                          PURGE GAS INLET
     FLOWMETER
         FUNCTION
         SELECTOR
          VALVE
                                    ZERO GAS INLET
FLOW
REGULATING
VALVE
                     m^^^m
                             SPAN GAS INLET
                                                 SELECTOR VALVE
                                                                 X
-------
RECORDER
                             In.	     r~7
                 PHOTOMETER
                                       SAMPLE CELL       I   | H
                                      	i,	i i •
I       "" I   I                PRfl! I   \f  j  i  \' /•—
I        I   J. Irr^	—ii» BALL --^r-^D  Jr-€:

PROGRAMMER-TT VACUUM BREAKER J VALVES Y    XT  ^

i        i i       \        /-rv|(PNEUMATIC OPERATED)

i~j-—Jj  »      \r~'i   vS|       i        \
	i      i  I       >v  ?    i .       A        \ n
  CONTROL

  STATION
il
I

J
  fill
HXJ-W-!-
                                        lx*-sj-
X

              AIR
02 CYL.
           ASPIRATOR
                    I
AC POWER
                                   r-
                                   i



•» «a> ••

T
JL_ jsUi_J

POWER
SUPPLY


.« _J
IX XJ


V
dsi 1 ,
l^-Ni
PRESSURE REGULATOR
© OXYGEN GUAGE




TRAP
	 1
                                                                LAMP
                                                     HEATED  I


                                                   COMPARTMENT
                                                             a
  FLOW !

INDICATOR

      I
                                                                               STACK
                                                                            «	I


                                                                            FILTER PROBE
                                                                                             CO
                                                                                            H

                                                                                            H
                                                                          ELEC SOLENOIDS
                                                                          HEATED TUBING
                                                                 —h
                                                                  SAMPLE IN
                                                                      J
                  EXHAUST

-------
             ACID MIST FILTER
—7
FLOWMETER
H
H

I


OJ
                     SECONDARY FILTER
                                                                            ANALYZER
                                                            VENT TO
                                                          ATMOSPHERE
                                             t
                                        CALIBRATING GAS

-------
                PRESSURE
                REGULATOR
                AND GUAGE
                                                                                                        HYDROGEN
                                                                                                        INLET
                                                                                                         V
AIR INLET
SAMPLE INLET
    NO. i
                                                        EXHAUST TO
                                                        EXTERNAL AMB
                                                                                         COLD PIPE AIR
                                                                                         SOLENOID
           SAMPLE SOLENOID
                                                                                                 VENT TO
                                                                                                 CABINET
                                                                                  1/min
                                                                                 ORIFICE
                                                                                 NEEDLE
  HEAT
EXCHANGER
                                                                               NEEDLE VALVE ^  COLD PIPE
                                        NEEDLE VALVE
                                                           VENTURI
                                                           ASPIRATOR
                                                                                     HYDROGEN
                                                                                     TRANSDUCER
                                                                                                    PRESSURE
                                                                                                    REGULATOR
                                                    AIR TRANSDUCER
                           CALIBRATION
                           SOLENOID
                                                                  PRESS TO
                                                                  TEST SOLENOID
                                                                                                               AUTOMATIC
                                                                                                               HYDROGEN
                                                                                                               SHUTOFF
                                                                                                               SOLENOID
                                                          HYDROGEN
                                                          SHUTOFF
                                                          VALVE
                                                                                                            v-1   VENT TO
                                                                                                            J—'— EXTERIOR OF
                                                                                                                  CABINET
CALIBRATION:
   INLET
       A
                                                                         80 °C OVEN
                                 PUMP EXHAUST    EXHAUST TO   V   JGAUGE
                                 TO CONTROLLED  EXTERNAL AMB. ^-^
                                 AREA
o

 I
H
H
H

-------
SOURCE SAMPLE
STREAM
H
H

I
                     MEMBRANE
    VACUUM PUMP
             DILUTER CARTRIDGE
  CARRIER STREAM
                                            FILTER {-->-
                                            _.,_,,_.,, .. .^.^^Jjj
    i=£
VACUUM PUMP
     CARRIER BLEED LIME
     TO ANALYZER
                              ANALYZER

-------
H
H
                      STACK GAS FLOW
                                            DILUTED SAMPLE
                                             RETURN TUBE
                                                                                TO STACK WALL
SAMPLE INLET PORT
                                                                 AIR INLET TUBE
                              SAMPLE INLET PORT
                                                               DISC DRIVE
                                             X"
                                            'SAMPLE DISC

                                       'SAMPLE CHAMBER

-------
           200
           190
           180
                 I   I   I    I   I    I   H_t_t_f_H_j   i   i   1_i_i_u-i  !.
        o
          160
          150
          140
          130
          120
          110
                 — HOURLY AVERAGE LOAD
                  O  INSTANTANEOUS LOAD
H
H
i
U)
        o
        <:
        cc
        S
900
800
700
600
500
400
300
200
100
                                        DILUTION RATIO -1600:1
0
 12   1
                    5  6  7   8   9
                                               10  11 NOON 12
                                               TIME OF DAY, hours
                                                                345
9  10   11  12

-------
          SAMPLE SLOT
H
H
H
              STACK GASES

1.
2.
3.
4.
5.
6.
7.
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
 PHOTOMULTIPLIER

-------
SOURCE
  J
                           FLUE GASES

                            \\\\\
                            tit
     REFERENCE
     SENSOR
                    \\\\\\
                 FIRST HALF CYCLE
                                       S02 MEASURING
                                       SENSOR
H
H

I


Ul
         SOURCE
           JL
                   FLUE GASES

                  \\\\\\
      REFERENCE
      SENSOR
                SECOND HALF CYCLE

-------
   tft
&
I M
   »>

                                                                             l!/
                                                                             Mi
                                                                    '
    m
                                       UN
                                                         I2N
                                          TM OF MY. honn
                                                                          UM
           CXIfOKT
O
§
                                       UM                iro
                                           TIME OF DAY. iKMi
                                       1BO               1TO
                                           TDK OF DAY kM<
                                                                          1300
                                                                           I3W
                                                                                            I4M
                                                       OPACITY RANGt, p
                                       TK Of OAT. torn
                                                   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

-------
          INSTRUfBTATION UTILIZED DURING STUDY
A,  S02 MEASUREMENT
    1.  DuPaNT 460/1 SOURCE Man TOR ING SYSTEM
    2,  CEA/BARRI.-IGER f-K-IV S02 STACK MONITOR
    3,  BAILEY METER Co,   S02 STACK MONITOR

B,  SUPPORTING MEASUREMENTS
    1.  EPA METHOD 5-S02
    2.  EPA METHOD 5-f-lASS PARTICULATES
    3,  TRANSMISSOMETER-OPACITY
    4,  BETA GAUGE-MASS PARTICULATES
    5,  PMC AUTOPITOMETER-CONTINUOUS STACK GAS VELOCITY,  TEMPERATURE

C,  DATA ACQUISITION
    1,  EA 2020 + TELETYPE
    2,  STRIPCHART RECORDERS
                                     III -  48

-------
S/M£
A
B
C
D
E
F
OPACITY
2,5%
3,0
15,0
17,5
45,0
47,5
fCTHOD 6
46QPPM
475
675
619
661
682
DuPONT
451, - 1,9%
438, - 7,8
577, -14,5
632, +2,1
664, + 0,5
678, - 0,6
CEAMKIV
424, - 7,8
416, -12,4
550, -18,5
599, - 3,2
624, - 5,5
638, - 6,4
BAILEY S02
590, +28,2
570, +20,0
680, + 0,7
720, +16,3
730, +10,4
750, + 9,9
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                                IA7-5
       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) result




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,  chlorophyl




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 expressions




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 neasured 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.  Potentiometric 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 timS 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 electrode 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.  Potentiometrtc
         Membrane Electrodes
       n
L " K  >  C. \. z
     C i
E  = constant
pH « -log a +
                zF
                                /
                RT .   r  ^   zi/z
                - In la fK a
                          -1,
pE - -log a  • Eu/(2.3 RTF  )
(c) Voltammetric Membrane Electrodes-       -
    (Dissolved Oxygen)                i, = jzFAP
Cationic = pM  = -log
Anionic = pA  = -log £
            11
          m b JV
(2)
(3)
(A)

(5)
(6)

(7)
      L = specific conductance
      K  •= cell constant
       c
      C. a ionic concentration
      X, = ionic equivalent conductance
      z  = ion valency
      E  B measured electrode potential
       in
      F = the Faraday constant
      K. = 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 hydrodynamic 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 show selective




permeability to various gases and vapors.  Gases reduced at the potential of the




sensing electrode (e.g., S0_  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., CO  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 xd.th the use of these electrode




systems have been the effect  of  mixing in the test solution on the electrode




sensitivity arid short term stability which necessitates frequent calibration.




In monitoring operations, the accumulation of inert or biological material on the




 cmbrane surface has caused a let of nuisance.  In addition, it has been practically







                                        IV -  10
m

-------
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  JL, is
                                                                          d


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"effect.  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  £] eKsI C                (10)
                           d   L     m bJ




Equation 10 can be reduced by introducing the proportionality constant (t>  in
                                                                        3.


replacement of term in brackets in equation 10,



                          1  =  A   eKsI C                      (11)
                           d     a
                                         IV  - 11

-------
Ionic strength values can be approximated by measurement  of  electrical conductance,




using Kg as a proportionality coefficient,
                                                                   (12)
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
                                          '02




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,








                                               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 oldes type, developed by Pungor  et.  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~, l~ and CN~~  ions.   Main interfering




ions are sulfides  and thiocyanates.
                                        IV -  14

-------
          The potential determining ion in these electrode systems is the Ag  and


the Nernst relationship can be expressed as follows



                          E  = constant + — In [Ag ]
                           m              r



and since



                          [Ag+] =  -4§§r                               <15)
Where K   ..  is  the silver chloride solubility product, then
       AgGl


                                           T?T             T?T       —
                             = [constant +   - In KJ - ~ In [Cl ]      (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



                                          RT       H~
                         E  = constant + — In [Ag ]                    /-U\



and since




                                                                         (18)
                                         IV - 15

-------
where K     is silver sulfide  solubility product, then
       AP . s
                                            PT           -prp       _
                           Em =  [constant + ~ In KA§ gj— In  [S2- ]         (X8)
          This electrode is also responsive  to  other forms of free sulfides,

and H S, through the following equilibrium relationships
                                                                               HS
                          Kl    -  K2


where Kj 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~], [H2S] and total sulfides  [ST],

                                             rn+i2     TH+I
                   log [S  ] = [S2  ]  +  log  ^y~  + -Ljr-J- + 1              .  .
                                              12       1                    *  '
                                                r,,+ -,2
                   log  [Hs] = log  [s2-] +  log
                         2
                                      2
                   log  [HS  ] = log  [S   ] + log   --                        (22)
Another type of the silver sulfide  electrodes  include those electrodes responsive to

Cu  , Pb   ot 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

     +                                      2-
to Ag  activity which is dependent  on the  S   activity, which in turn depends on the

activity of the divalent metal in solution according to the following equilibrium

relationships


                          [Ag+]  [S2~] = K.  _                               (23)
                           [M2+]  [S2"]  =  K                                  (24)
from which
                                             t^lj
                          [Ag'] - 1(^-1   [M2^]/                          (25)
                                  L • %is
                                          IV  - 16

-------
Since




                                 RT        "f~
                   E  = const. + •=-  In [Ag ]





it follows  that


                                 •nrp     Ap S    T?T       -f-

                   E  = const. + •££ In -^	f- ^ In  [M2 ]                 (26)
                    m            /r     K-,_     /r




                                                  2 +
         This electrode system is selective for M   provided  that
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 JM to 10 5M.   The main interfering ions are Hg2  and Ag   for  the cupric ion



electrode  and Hg2  , 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




                   Ag2S + 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



                                   •n«ri      _

                   E   = constant - ^ In [F ]                              (28)
                   m               i




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 (A1F,3~) 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 Mg   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,




                      E  = constant + f£  In  Ca2+)             (29)
                       m              /r     L    J



The operative concentration  range is 10  M 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 arid 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



pH range is A to 7.  Major interferences are HCO~ , Cl~, SO2"  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



In addition, the electrode sensitivity to commonly present interfering ions, such

   -     _       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




C0_,  NH., and  S0~.  The  C0? eltctrode 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.  C09 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 C00 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, VIII, 135, Ann 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, j48, 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-/}

           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
Fluorimeter
Flame spectrophotometer
Atomic absorption spectrophotometer
Electroanalyticn! 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


                                  Beckman Solid State Membrane Electrodes
     Addresses: 2500, Harbor Boulevard, Fullerton, Calif., 92634, U.S.A. (.P.O. Box 1, Glenrothes, Fife, Scotland).
Ion

F.
ci-
Br-
I-
sa-
Cu3+

Model

39600
39604
39602
39606
39610
39612J

Molar Range

io° — io-8
10° — 5 x IO-6
10° — 10-'
10° — io-°
10°- !0-°
IO*8 (lower
limit)
pH Range Resistance {Megohms) Response Time Principal Jnterferants

0-*13f
0-* 14
0—14
0—14
0—14
0 — 14


<5
< i
< 1
<0-25
< 0-25


{Seconds)*
<3
<2
<2
<3
<3



OH-
Br-;I-;S2- and CN~
7T;S2- andCN-
sa-

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 coni response to a stepchange from 10"1 — 10~3 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.
  J Like the Coleman solid state electrodes contains no internal filling solution.
                            Orion 94 and 96 Scries Solid State Specific Ion Electrodes
             Addresses: 11  Bhckstone St., Cambridge, Mass., 02139, U.S.A. (E.I.L., Richmond, Surrey)^.
Ion
CNS-

CN-
F~/La'*
F'/La'*
Na*
ci-
ci-
Br
r
s>- j
<


Cda»

Pb'*

Model Total Molar
Concentration
Range"
94-58

94-06
94-09
96-09
94-11
94-17
96-17
94-35
^94-53

94-.I6
"94-29

94-^8

94-82

10°

IQ--
10°
10°
10°
10' —
10° —
10° —
10° -
10°

10°
10°

10"

10'

— k

1 -J
-•.
— >
-*
5
5
5
5
-»

~*

—

-^

-io-B

"IO-6
10""
10"°
io-8
X 10-"
X 10-°
x 10~6
X 10-"
io-7

io-7
io-7

io-7

-.o-1

pH Range]'
0-

3
0
0
3
0
0
0
0
0

0
0

1

2


-*
-»
_+
_,
— >•
— >
-»

_

->•

—

14

14
11
11
12
13
13
14
14
)4

14
14

14

14

Resistance {Megohms) Membrane Material Principal Interferant ions
AgSCN + AgaS S--, Hg3* and Cu3+ must

<30
< 1
<30
<200
<30
<30
< 10
l-i-5

< 1
< 1

<1

< 1


Agl + Ag2S
LaF3 + F.ut
LaF3 -f- EuJ

AgCl + Ag3S
AgCI -f- Ag,S
Agflr + Ar2S
Agl 4- AgaS


AgjS
Ag2S + Cus

Ag,S i- CdS

Ag3S -f- f'bs

be absent
S3" must be absent

j>OH~ only intcifercnce
Ae+
>S3" must be absent
S2" must be absent
S2" must be nlvcnt
None as lai as examined

Hga * mist be absent
S'-. Agl and Hg2* must
absent









be

Ap\ Hg3' andCu3^ must
b-ji absent

Ag\ Hg** ami Cu2' musl
be ubsc.:t

                                               IV -  24

-------
                                -  LIQUID  ION EXCHANGE  ELECTRODES
                                       Some Corning Liquid Ion Selective Electrodes
               Addresses: Coming Glass International, Mcdlidd, Mass., 02052, U.S.A. (3 Cork St., London W.I)
                                                                                (EEL, Hallstead, Essex)
Ion Model
Number

Cl- 476131

NO3- 476134

Caa+ 476041

Ca2+ -Mg2+ 476235

Working
Concentration
Range
IQ-'-lO-'MNaCI

10°-10-6MKNO3

]0°-10-°MCaCla

10°-10-sMasCa2 +

Operating
pH Range

1-12 in
10"1 MNaCI
2-5-10 in
10-3MKNO3
5-10 in
10-3 M CaCI2
5-10 in
JO-3 M CaCIa
Temperature
Range (°C)

10-50

0-50

10-60

10-60

Principal Interferants
(^"MN > 1)

i- > cio4- >
NO3- = Br-
cior > r

None given

(Ba2 + --=Sr2+ >
Ni2*=Ca-=Mg2+)'
    Overall size (length X diameter) cm :  12-3/12-7 x 1-58.
                             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 Ca3 + and Ca2* — Mg3* electrodes only.
             Minimum sample size (ml) :  10 quoted for Ca2 + and Ca2+  - Mg2+ electrodes only.
                                       (Upright in air for short  periods.
                              Storage : |jon excnangcr )jquid 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          Model          Molar        pH    Resistance    Principal Inlerferants    Mobile Exchanger
                               Activity Range    Range    (Megohms)      (where JfMN > 1)          Sitei
cr

NO-T
BF<-
cior
Cu3+*
Pb2+*
Caa +
Divalent
Cas+ -Mga+

92-17

92-07
92-05
92-81
92-29
92-82
92-20
92-32

10- MO-6

lO'MO-6
lo-'-io-6
io°-io-5
10-MO-S
10-MO-8
io°-io-s
10MO-"

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
<10
<25
<10

Br-,r,N03-,CIOr NR4 +
andOH-
r, C1O3- and CIO,
I-
OH-
Fea +
Cu2 +
ZnaMButseeTabIe2)
Zn2+, Fe2 + ,Ni2 +
and Cu3*

-[Ni(phen)3]a +


'R-S-CH3-CO2W

•(AlkylO)aPOa-

Operative temperature range (°C): 0-50.
Overall size, lefg'h X diameier (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 licjuid.
Dollar cost: 195
  * Now withdrawn from Orion 1969 Research Guide. (Cat/961).
  t Iron replaces nickel in the ion exchanger material for (he CIO,," electrode.
                                                   IV  -   25

-------
                        El- HETEROGENEOU MEMBRANE  ELECTRODES
                             Pungor-Radclkis Heterogeneous" Ion SelcctiM? Electrodes
                fProtech, 40 Hit;h St., RicKmansworth, Hcrls (Advisor)1 Services).
      Addresses < Simac Instruments Ltd., BriJgcuay House, Bridge \Yay, \Yr.inon,  Middlesex (Technical Services).
                (_Radelkis L-lcctrochcmical Instruments, Budapest 62, Hungary.
Ion

ci-
Bf
I-
sa-
so,3-
PO43"
F"
Laboratory

op-ci-711',;
OP-Br-711«i
OP-I-711^
OP-S-71 lH
**
*•
**
Molar
Range

IQ-1 -* 1Q-6
]Q-' -H. 1Q-B
10"1 -> 10~7
lo-1 — io-17
JO-1-.-10-5
io->~io-5

Maximum
Resistance
(Megohms)%
10
10
10
10



Membrane
Material\\

AgCI
AgBr
Agl
Ag2S
BaSO4
BiPO,
LaFc-CaF2
Principal
Iiiterferants

Sa-;I-;Br-
Sa~; I"
sa~




        Operative temperature range (°C)
Overall dimension (length x diameter) cm
                               Storage
                       Response time (s)

                                 Cost
1/5^90
11-5  x !  for 711 Models.
Can  be dry stored after a distilled water rinse.
15 -•• 60 (Up to 3 min quoted by Rechnitz and co-workers for some electrodes.)
References 18 and 20.
$60
  • The terms "homogencoub'1 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-Ion-700 series carry side pin electrode connector plugs, while the OP-Ion-711
types are filled wuli shielded cabling.
  ', Pure silicone rubber > 11 x 10Johms.
  II Dispersed in s:licone rubber matrix (50 v,t per cent of silver salt).
  H Thc
-------
          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
                                          X

sensed by the detecting electrode in conjunction with a reference


electrode.




                 E(mv)  = Ea + b log (CxYx + kyCy Yy +....)           (1)




The electrode inherently senses activity,  rather than concentration,  so


that Cx must be multiplied by YX) the  activity coefficient,  in this


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  k    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


   designing an  ion  selective electrode monitoring system.  In this  regard,
in
                                      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 Cx and C  may be influenced by temperature, T, ionic




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 HF, 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 k   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).



    E , 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
                                                              A.



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

-------
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 electrode 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/32(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
                                   electrode
                                   chamber
          -0
     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 Variable
Vx
                                              T,  I,  pH,  complexing agents

                                              T,  I,  reference electrode,
                                              ions  selective electrode

                                              T,  ion selective electrode,
                                              charge on  ions sensed
V-
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/l,  Using Different Aluminum Concentrations (2)
Aluminum
concentration
(mg/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

  NH|/NH3

  ci-

  oci-/ci2
  Cu++

  CN-

  F-

  Hardness

  N03
Range (mg/1)
1-100
3.5-3550
0.05-100
0.05-1000
0.05-100
0.1-10
0.01-100
1-10,000
Interferences
Volatile amines
None
Br-.I-
None
Au"1"^, Ag+
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
          Standardization:

        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 th$  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 Minnesota 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 Water 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]  + K2



(T-.2)tM.  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

-------
                                    ANOKA CO-      |  COLUMBUS

                                 GROW    I   HAM LAKE.   I
                                                                       nl
                                                                     QHEW TRIER
                                                            I ri' ""•"•"
                                          EUREKA    |  CASTLE ROCK

                                                  I          |   HAMPTON   I   DOUGLAS
       I     NEWPHAGUEl          I       ELKO            |          I

            -^j-j                         J           ^     |     Ti   ^__ OOLF'H~

                              _        |           1,1 fRANOOLPH
METROPOLITAN     SEWER    BOARD

              Wastewater  Treatment Plants

  • Treatment  Plant in Operation During 1973
1  Anoka
2  Apple  Valley
3  Bayport
4  Blue Lake
5  Chaska
6  Cottage Grove
7  Farmington
8  Hastings
 9  Lakeville
10  Lonq  LBke
11  Maple Plain
12  Medina
13  Metropolitan
14  Mound
15  Newport
16  Oak Park Heights

        IV -  50
17  Orono
18  Prior Lake
19  Rosemount
20  St.  Paul  Park
21  Savage
22  Seneca
23  South St. Paul
24  Stillwater
25  Victoria

-------
    METROPOLITAN SEWER BOARD
   WASTEWATER TREATMENT PLANTS
           (LOCATED ON MAJOR RIVERS)
H
<

I

(Jl
NEWPORT

ST. PAUL PARK

-------
     WATER QUALITY SAMPLING STATIONS
                      1927-1973
                                                                                           ^
H
<
14
1
ft
<
                                                                               I         f
                                                                              WASHINGTON AVE. BRIDGE

         FORD DAM

FORTSNELLING BRIDGE
 FORTSNELLING PARK
                                                 CHASKA f^NESOTA\£\^
     LAMBERT LANDING    §

        ABOVE METRO PLANT
         ST. PAUL TERMINAL
          SOUTH ST. PAUL
MENDOTA W 1-494 BRIDGE
BRIDGE    m. |NVER GROVE BRIDGE

                       /
                    PRESCOTT
                                                                                                SAMPLE LOCATIONS
                                                                                                DOWNSTREAM ON
                                                                                                MISSISSIPPI RIVER

                                                                                                SMITH LANDING
                                                                                                DIAMOND BLUFF
                                                                                                REDWING DAM
                                                                                                REDWING
                                                                                                FRONTENAC
                                                                                                LAKE CITY
                                                                                                READ'S LANDING

-------
 PRESENT METROPOLITAN SEWER BOARD
   WATER QUALITY SAMPLING STATIONS
H
<

I

Ul
       INVER GROVE BRIDGE


                    <§~
                    ^
	J
                                                                                           SAMPLE LOCATIONS
                                                                                           DOWNSTREAM ON
                                                                                           MISSISSIPPI RIVER

                                                                                           REDWING DAM
                                                                                           REDWING
                                                                                           FRONTENAC
                                                                                           LAKE CITY
                                                                                           READ'S LANDING

-------
                              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) (BOD5 = 25 mg/l,
 SS = 30 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

-------
  U.S. GEOLOGICAL SURVEY
RIVER FLOW GAGING STATIONS
                 -j  JORDAN GAGE

                  L	,


-------
       PROPOSED LOCATIONS OF
    CONTINUOUS AUTOMATIC WATER
    QUALITY MONITORING STATIONS
H
<
LTI
CT>
                      —I    JORDAN GAGE
                       i
                       L	


-------
   WATER QUALITY HISTORY
 DATE  HR MN NAME TEST
 09/26 00:00 QR06
"09/26 01:00 QR06
 09/26 02:00 QR06
 09/26 03:00 QR06
  09/26 04:00 QR06
"  09/26 05:00 QR06
  09/26  06:00 QR06
"09/26  07:00 QR06
  09/26  08:00 QR06
 "09/26  09:00 QR06
  09/26 10:00  QR06
  09/26 11:00  QR06
  09/26 12:00 6R06
  09/26 13:00 QR06
  .09/26 14:00 QR06
•EST
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
nrv
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  1973  PRODUCED 09/26/73
   MONTHLY MEAN VALUES

    9f    .-    S    ,.,
~ SSSS SKȣ
                                                           03
                                                           08
                                                                 5.2
                                04
                                09
                                      5.5
                                                 7.9
                                            EXCEEDED FOR
                           11:00
                           i5:oo
                                                                                      05    12.7
                                                                                      10


                                                                                      80s
                         0.00
   CURRENT QUALITY READINGS
_ DATE  HR NN NAME TEST  VALUE
 09/26 16:00 9R02  DO     4.6
-09/26 16:00 QR03  DO     8.1
 09/26 16:00 8R06  DO
                                                             if
                                            80s
                                            7.P
                                          I;28
                                          ^7.7
                                           806
                           	MONTHLY RIVER QUALITY REPORT FOR  08/73 PRODUCED 09/26/73
                           NAME"TEST " MINIMUM DATE  HR  MN   MAXIMUM" DATE  HR MN   MEAN"  #RDG XSTD
                                QR02  CN
                                QR03  CN
                                QR06  CN
                                QR02 . DO
                                QR03  DO
                                QR06  DO
                                QR02  PH
                                QR03  PH
                                QR06  PH
                                non%  TP
  322 08/23
  261 08/18
  708 08/13
.  0.4 08/05
  2.0 08/30
  3.4 08/13
 5.03 08/14
 5.46 08/20
 7.41 08/15
 70.9 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
OS/28
08/14
08/03
08/2S
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
                      EXAMPLES OF COMPUTER PRINT-OUT FORMATS

                     AUTOMATIC WATER QUAUTY MONITORING DATA

                                       IV  -  57
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
0
0
0
3
1
0
0
0
0
0
0
0

-------
             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 projects 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

-------
      GCNEIUl. LOCATION OF AWOMS

  FbORCZYK H.:RESEARCH AND STUDIES ON THE UOCATON OF AUTOMATIC WATER QUALITY MONITOWNG
           STATIONS - AWOXS


                        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
5
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 prime considerations with respect 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

-------
                       LOKALIZACJA  OGOLNA  A5PJW MA TIE PRZEBEGU  KRZYWEJ  BZT5  WZDKUZ 8IEGU  RZEKI ODRY
                       GENERAL  LOCATION  OF AWQMS ON  THE  GROUND  OF COURSE  OF  B005 CURVE ALONG ODRA
                       RIVER WATER  COURSE
i'
It
r1

$

SC

r"


i — i —
i — i —
LJ 	
j

T

„
£


i :

| : " '~3
?' " 's
l^—ft
./>; LI 1*1

; j
CO ISC



S
r
-
— j







a
II
£

*•.
' '• 	 Si
<


03 2
1
I
|ti
"
Ir

,D


i


i
3n<



-


3



—

t;














350










---


400



	 , 	


i.50 500



~







J

55



' 	
	

C



	


600



	 ,


650 700



	


m
                       LOKAUZACJA OGOLNA  ASPJW NA TIE PRZEBEGU KRZYWEJ  ILOSCI ZAWE3N WZDtUZ  BIEGU RZEKI ODRY
                       GENERAL  LOCATION OF AWQMS ON THE GROUND OF COURSE OF SUSPENDED SOLIDS  QUANTITES CURVE
                       ALONG ODRA RIVER WATER  COURSE
                                     FIGURES   2  &  3
Znhflb
                    LOKAUZACJA OGOLNA ASRNV NA TIE PRZEHEGU KRZYWEJ  STE.ZEN FENOLI WZDIUZ  BEGU  RZEXI OORY.
                    GENERAL  LOCATCN OF AWQMS ON  T>E  GFOJND OF COURSE  OF PV€NOLS CONCENTRATIONS CURVE
                    ALONG  OORA RIVER WATER  COURSE,
ax
b^000
^? too
| acc
6 o
DO
1





i
i













—





T
I
?i
1


•^















^
UJ


i
*

«
^





1
H,








1
>

?«
x 1






i

fll

































^•^•^H
































































                    LOKAUZAOA OGOLNA ASPW NA  TLE  PRZEBEGU  KRZYWEJ  Slk/t-KI  CHLORKOW  WZDKUZ BEGU RZEKI CORY
                    GENERAL  LOCATON OF >*VQMS ON THE GROUND  OF COURSE OF CHLORU3ES CONCENTRATIONS  CURVE ALONG
                    OORA  RfVER WATER  COURSE


                                              FIGURES   4  &  5
                                                            IV  -  64

-------
                                                      1  1
8    i
H
It:
   Uonwtry rvb
                                                300
                                                                           soo
                                                                                 660
                                                                                        600
                        LOKALIZACJA OGCINA ASRWNA TIE PRZEBEGU KRZYWEJ STEZEN SIARCZANOW WZDIUZ BEGU RZEW ODRY.
                        GEhERAL IjOCATON OF AWOMS ON 1>E GROUND OF COURSE OF SULPHATES  CONCENTRATIONS CURVE  ALONG
                        ODRA RIVER WATER COURSE.




1
I 1SOO
II
P™
1




f
j
1
T



















^_ '





.
c
^
jj
,10
N
1













I



si


'">
^


















II
*

*
IS
< 3



































































































































                            1502tt>2S03D035Q«0«0500550600650TO


                        LOKALIZACJA OGdLNA ASRW NA TLE PRZEBEGU  KRZYWEJ STEZEN  ZWIAZK6W ROZPUSZC20NVCH WZDtUZ
                        BEGU RZEKJ ODRY.
                        GEfCRAL LOCATION OF  AWQMS ON T* GROUND OF  COURSE OF DISSOLVED  SOLIDS  CONCENTRATIONS
                        CURVE  ALONG CORA RIVER WTER COURSE.
                                              FIGURES  6  &   7
                  Przekrdj  pomiarowo - kontrolny
                  Monitoring  cross - section
                        CHAtUPKI
                                         Rzeka
                                         River
                                         ODRA
                 0,413.

Nr plondw bobawaych   /
No of verticals In cross -
section
                         ROZKkAD  PREDKOSCI  PRZEPtYWU WODY W PRZEKROJU  CHAtUPKI  NA RZECE  ODRZE
                         DISTRIBUTION OF WATER FLOW VELOCITY IN  CHAtUPKI CROSS-SECTION, IN ODRA 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 - 56

-------
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
Mr ptonow fiadawciych
No al vrrUcakt h cran -
HCtton
                     ROZKtAD ZAWARTOSCI TLENU ROZPUSZCZONEGO W PRZEKROJU CHAtUPKI NA RZECE ODRZE
                     DISTRIBUTION OF DISSOLVED OXYGEN CONCENTRATIONS IN CHAkUPKI CROSS-SECTION,
                     IN ODRA RIVER
                                        FIGURE  9
                  Przekrdj pomiarowo - kontroLny
                  Monitoring cross - section
                       CHAtUPKI
  » pfoocV badawczych
  No of vtftlcaJs In wo*s-
    Rzeka
    River
    ODRA
                       ROZKLAD W^RTOSCI UTLENIALNOSCI Z PROBEK SKLAROWANYCH  W PRZEKROJU CHAtUPKI
                                ODRZE
                            ,         PERMANGANATE OXYGEN DEMAND VALUES FROM CLARIFIED SAMPLES
                       IN  CHAtUPKI CROSS -SECTION, IN ODRA RIVER
                                           FIGURE  10
                                            IV  -  68

-------
                Przekrdj pomiarovvo - kontrolny
                Monitoring cross - section
                      CHAtUPKI
Nr ptondw badawuych
No of verticals In cross-
section
Rzeka
River
ODRA
                      ROZKkAD WARTOiCI UTLENIALNOSCI Z PR^BEK SKKlCONYCH  W PRZEKROJU CHAkLIPKI
                      NA RZECE ODRZE

                                                                                 M.XED SAMPLES ,N
                                          FIGURE  11
                 Przekrdj pomiarowo - kontrolny
                 Monitoring  cross - section
                      CHAfcUPKI
H ptondw badawczych
No of yfrtkals In cross -
uction
  Rzeka
  River
  ODRA
                       ROZKLAD 5TEZEN CHLORK6W W PRZEKROJU CHAHUPKI NA RZECE ODRZE

                       DISTRIBUTION OF CHLORIDES CONCENTRATIONS IN CHAtUPKI  CROSS - SECTION,
                       IN ODRA  RIVER

                                               FIGURE  12
                                                      IV - 69

-------
                 Przekroj pomiarowo - kontrolny
                 Monitoring cross - section
                      CHAUJPKI
f* pbniw bodawaych
No crt verticals In CTOM -
                       Rzeka
                       River
                       ODRA
                        ROZKtAD STEZEN ZWIAZKtfw ROZPUSZCZONYCH W PRZEKROJU CHAtUPKI NA RZECE ODRZE
                        DISTRIBUTION OF DISSOLVED  SOLIDS CONCENTRATIONS  IN  CHAtUPKI CROSS - SECTION
                        IN ODRA RIVER
                                           FIGURE  13
                                                                        KRAK6W
        WROCIAW
                                                                                                 NIEPOtD-
                                                                                                 MICE
       BLOCK DIAGRAM OF AWQMS SYSTEM
    radio rtceivtf

    radio transmitter

VTJ  telegraphic unit

;7o]  telemetric equpmerrt

VSJ  data conversion unrt

S I  performance unit

Va3  monlor W-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 permissible 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
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
Water Level
68 69 70
432
432
100
8760
7934
90.5
5448
4909
90.1
Conductivity
68 69 70
432 8760 5448
269 7311 4859
62.2 83.4 89.1
Turbidity pH
68 69 70 68 69 70
432 8760 5448 432 8760 5448
426 6358 4779 422 7306 4844
98.6 72.5 87.7 97.6 83.4 88.9

H
i
o
1. Number of
hours in
the period
2. Operating
hours
3. Operating
hours %
Redox potential
68 69 70
434 8760 5448
404 7457 4859
93.5 85.1 89.1
Temperature
68 69 70
434
327
75.7
8760
7535
86.0
5448
4899
89.9
Chlorides
(from 1.07.69)
68 69 70
4416 5448
3051 4716
69.0 86.4
Solar radiation
(from 18.04.70)
68 69 70 68 69 70
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



a heavily polluted river and the value of  ^  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

-------
                              Basic  types of curves  re-
                              presenting  concentration
                              of pollutants  and rate of
                              flow
                      type I   - for  heavilly  polluted
                                 rivers
                      type II  - for clean rivers
                      type III - for intermediately pol-
                                 luted river*.
              FIGURE  15
                                      k« (r) 20°C=QOOT7A BZlP50
                                         OA    07  10
                                            k1(r)20°C
Correlation between initial BOD,- and the   value    of
k1/r/  20°C
            FIGURE  16
                        IV -  74

-------
"I
60 1
RZEKA  ODRA W CHAtUPKACH
RIVER  OORA AT CHAKUPKI
                                         1967
                                             =0.3TO

                                            QL< 0,001
   20    40    60   80   100    120   W   160   t90  200   220   240
                                             Prztptyw m3/sek
                                             Flow m3/sec
               FIGURE 17
       10
15
                       20   25   30
                        Q nrf/sek
                                       Relationship between
                                       BOD,- and flow taking
                                       into consideration
                                       temperature effect
                FIGURE 18
                   IV -  75

-------
14
12
RZEKA ODRA  W CHAKUPKACH
RIVER OCR/ AT  CHAKUPKI
                                                               1967
                                                           =-0,750
                                                         a
-------
   300
SfH
  ooo-

- 1000
F_
||eoo
5 > AGO
f!
*
          RZEKA OORA W CHAIUPKACH   „..
          RIVER  ODRA AT CHAtJPKI     19*7
                           PrrepVyw m /sek
                           Row m^/»ec
 RZEKA OORA W CHAtUPKACH
 RIVER OORA ATCHALUPKI
1967
                       201
                                                                     Flow  m/s«
                                             1200
                                     WOO
                                 If900
                                  I!600
                                  as
                                  Is «°
                                             200
         20  40   60   80
                          W)   120
                           Prz*p
                           Flow
                                   WO   160
,   RZEKA AIRE W SEAL
u  RIVER  AIRE AT BEAL
                                          1966/ 67

                                    SUCHA POZOSTA<.o4li
                                    TOTAL SOLIDS
                                        132380
                                           200   400  500  900  1000 1200  1400 16CO

                                                             Prz*p»yw m3/l»k
                                                             Flow m^/s*c
                               FIGURE  21
           RZEKA  ODRA W CHACUPKACH
           RIVER  OORA AT CHAtUPKI
                                                      dla  aokrtwj ttmpwotur > 15 *C
                                                      for t«mp«raturt  rangt >15*C
    dla zakrtsu tamptratur 0-1S*C
•   for temparaturt rangt  0-1S*C
                     50
                                                                    Prxtptyw m
                                                                    Row m'/scc
                               FIGURE  22


                                   IV  -  77

-------
                      WYNIKI POMlARdw W - 20
                      RESULTS OF W-20 MEASUREMENTS
                                        FIGURE 23
   H
   9
   §0
   o
rt- O.
   n
p  o
3  S
13  O
•  rt-

s  I
CO  cT
   g
       II
       3e.
       Tfl
       10 8
                                            cWorkl  mg/l 0"
                                            chlorides mg/l Cl~
                                        FIGURE  24
                                            IV  -  78

-------
                             >|8S/6 poo)
                                     >|»unpDf
                              FIGURE 25
!
26000
        Przekrd'j pomiarowo-kontrolny
        Monitoring cross-section
                 25900
orides

s
s
       1
            14
            I
                               121
      7///////^/////^^
                                   M
                                          Rzcka
                                          River
                                                  OORA
                                          Rok
                                          Y*or
1966
                                                    najdtute) trwajqcy tadunek
         20  40   60  60  100  120  140 160 160 200  220
                                               drt
                                               days


                              FIGURE 26
                               IV -  79

-------
          Przekr6j pomiorowo-kontrolny CHAtupK|
          Monitoring cross-section       ^rwtu
  L-'
  oTi.
            Rive7 ODRA

            5ok  1968
            Year
£M-
225
200
175
150
125
100
50

I'
1-



21
1-



)






£





2
w//////////m//,
i-


21





'5
g







100



i
5
|b
NO
8-J
lajdh
mostl
of chl

i
41 O
ingeof chloric
ntration variat
i i
^" U
£ u
jzej ti
reque
orldes


•\vajqce st^zenie
ntly occuring values
concentration
               20   60   60   80  100     dnl
                                          days
             Frequency of occurrence curve of chlorides con-

             centrations  in Odra river at  Chalupki,  1968.


                           FIGURE  27
B
a
o
 200
      200
             PrzeknJj pomiarowo-kontrolny _.....,_...
             Monitoring cross-section     CHAUJPKI
 600r 300    30
                 Rzeka
                 River
                 Rok 1QCB
                 Year 196e
                                                                 ODRA
                         krzywa czqsow trwania
                                                  chlorkow
                         time duration curve of chlorides concentration

                       h_ krzywa czasbw trwania tadunku chlorkow
                         tim« duration curve of chlorides load

                         krzywa czasow trwania przeptywow
                         time duration curve of flow
                                          47kg/sek sredni roczny
                                                 average annuolO
104mg/l srednie roczne sWzenieCl"
•  overage annuol Crcooojit/ation
                                                 200
                                        50
                         300 dnl
                        M days
                         KX)'/.
                           FIGURE  28
                               IV -  80

-------
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




enabled the 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 Researcli 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 lias 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.J

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 FLUOROSENSOR
            FOR THE
   DETECTION OF OIL ON WATER
              By






           H. H. Kim




        Wallops Station




   Wallops Island, VA 23337






              and






         G. D. Hickraan




         Sparcora 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 nm and 390  nm 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

-------
FIGURE 1 - Laser Fluorosensor Installed in NASA's  DC-4 Aircraft




                         IV - 87

-------
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,(6) 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

-------
                   t
•p-
 I CO
E
fO
                                                   PO
                                                0  Tf
                                                «  s
                                                O  LU
Ambient Oil Levels in the Delaware River Measured
with the Airborne Laser Fluorosensor.   (Auq. 24
1973  10:40 a.m.)
                 IV - 91

-------
 and  a value of zero against ambient noise in the open sea.   Thib 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 oil 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
                      10:40 am
   50
4J
•H
CO
C
0)
-p
c
H

-P
a

-------
 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  Aquatic 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^ THEMATjC_MAPj?ING




             CHAIRMAN




     MR. JOHN D.  KDUTSANDREAS




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

-------
an
     Therefore, the growing population of this country coupled with




  widening horizon of demands being made on land resources has brought




   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 tnuch 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 prominent 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 use:




     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  centre




water supply,  and waste water treatment.




     Federal users also need current land use information.  The  asse




ment of recreational needs and opportunities requires knowledge  of  t




location and extent of urban areas and potential recreational  lands.




This information is used to forecast demand, identify potential  soli.




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 agi




cultural, urban and other types of land inundated by flood waters




would be invaluable in estimating damages, future crop losses,  and




consequent econimic impacts.




     Presently, there is no systematic compilation of information 01




existing land use and its changes on a national basis.  For detailec




planning at the local level, ground surveys, occasionally supplemem




by aerial photographs, are used.  In some cases, land use informatii




is hypothesized on the basis of data on utility hookups, school  popi




tion, building permits, and similar information.  Transportation




planners  collect the necessary information using similar techniques








                                  V -  6

-------
Some States such as Connecticut, 5   New 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
                                   / Q \
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




¥alley 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-1960's 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




rerrote sensor capability to map effectively the land uses of the entire
                                   V - 8

-------
United  States within a reasonable  time  frame  of  approximately  three




yearsf  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 correct








                                   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).
  °  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  - 15

-------
                          REFERENCES
1.   U.S. Dept. of Agriculture,  1972,  Farmland:  are we running out?:
          The Farm Index,  Dec.  1972,  p.  8-10.

2.   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.

3.   U.S. Congress, Senate, 1973, Senate Bill  268,  93d Congress,
          1st sess.,  55 p.

4.   U.S. Congress, House, 1973, H.R.  4862,  93d  Congress, 1st
          sess. ,  33 p.

5.   Connecticut, Dept. of Finance and  Control,  Office of State
          Planning, 1970,  1970  Land  use  study  for Connecticut,
          unpub.  report.

6.   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.

7.   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.

8.   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.

9.   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. Taranik
                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. Tuthill
                         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 floodplain 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  theJneated 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 unclibrated 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 just 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 S Dam 14

        4 June, 1971
    N
                               ILLINOIS
                                    CORDOVA
0   .5   Imile
I - " ^
  SCALE
 I:62,500
 ISOTHERMS
    m°C
                                 PORT .BYRON
        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-altihjde 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 the 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 icing is such that open water can result from three significantly
different conditions :

    I. First is the obvious condition of  imputs of water 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 the 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 the 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  the
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 the 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
the balance significant in any  evaluation of the unnatural  elements of the thermal
regimen of the reach of the Mississippi  River bordering Iowa.
                                        V  - 24

-------
Muscatine Co.
 Louisa Co.
 GRANDVIEW
                                                                           Patch of thin ice
                                                 Open water
                                                     Distribution of Open Water and Ice Cover
                                                      on the Reach of the Mississippi River
                                                             bordering Iowa
                                                            14 February 1972
                                           Patches of thin ice
                                           	J
                                             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 optimal,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

-------
                                     s^
                              Griswold  \
                Original Scale:
                 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
                                  20miles
                                     B.E. Hoyer
                                     G.R.Hallberg
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. Multispectral 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 inundation 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  feedlot 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 river-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.Circ. No.3,  176p.

 Tuthill, S.J., Taranik, J.V., and Hoyer, B.E.,  1973,  Thermal Remote Sensing on the
     Mississippi River in Iowa: AlChE Symposium Series  No.129, vol.69, p.391-400.

 Tuthill, S.J., and Taranik, J.V., 1972, Remote Sensing, A Tool for State Planning-
     Management in  Iowa:  Proc.Sth Internet. Symposium on Remote Sensing of the
     Environment, Env.Res.lnst.of Michigan, vol.1, 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
     Iowa; 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, Multispectral
    Flood Inundation Mapping in Iowa:  Amer.Soc.Photogram.,  Sioux Falls, South
    Dakota, Symposium, October 1973.

Hallberg, G.R., Hoyer, B.E., and Rango, A., 1973, Application of ERTS-1 Imagery
    to Flood Inundation Mapping, in Symposium on Significant Results Obtained from
    ERTS-1, Goddard Space FlightTenter, NASA.

              , 1973, Application of ERTS-1 Imagery to Flood Inundation Mapping; in
    Significant Papers on ERTS-1, Plenary Session, Goddard Space Flight Center, NASA,
    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 lnfo.Circ.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.
                                    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  of 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 - 10%.   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 207o.  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 Request.

        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
Fixed Data Inserter
Time
Mi ssion
Line
Run
      Camera

      System
Voice Intercom
Aircraft Radio
Long Range Radio
Sate 11ite Radio
Camera Control Unit
        A/C
Power Control Pane'
             Remote  Display
         PCM
                                 System
                                 Patch
                                 Panel
                                 Intercom
                                 Control
                                 System
                                Vol tmeter
                                Osci1liscope
Power
Supply


Control
Panel
                                                          Thermal Scanner
                                        Tape Speed Compensation
                                                       Signal  Processor
                                                                 1
                                                AR - 700

                                             Tape Recorder
                                                         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 on 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 2 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 thermal 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 tpnes 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

-------
CTi
ALTITUDE  3000  FEET
19 SEPTEMBER 1973
                                                                                                              B
                                                                                                              u
                                                                                                              cr>
                                                                                                              o
                                 APPROXIMATE  SCALE 1:32,000
                                                                                   ORIGINAL FORMAT FROM      |
                                                                                   TAPE TO FILM CONVERTER  J
                                                                              s
                                                                              U
                                                                                    EH
                                                                                    K
                                                                                          11.5 SURFACE TEMPERATURES °C
                                                                                           •   BOAT STATION
                                                                                          LIGHTER TONES  INDICATE
                                                                                          WARMER TEMPERATURES
                                                                                          BILOXI BAY
                                                                                                               INDEX MAP
     ALTITUDE 6000 FEET
     19 SEPTEMBER 1973
                                                    ORIGINAL FORMAT FROM
                                                    TAPE TO FILM  CONVERTER
                           APPROXIMATE SCALE 1:74,000
                                             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 conversion.  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
from 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

-------
  ALTITUDE 3000 FEET
                                                                     SC 4020 MICROFILM ENLARGED
  19 SEPTEMBER 1973
                               APPROXIMATE SCALE 1:32,000
ALTITUDE 6000 FELT
19 SLTTEMTiEr- l'J73
SC 4020 MICROFILM ENLARGED
                                                                                EACH OF EIGHT GREY LEVELS
                                                                                REPRESENTS A TEMPERATURE
                                                                                INTERVAL OF ONE °C

                                                                                        25° -  32°C
                                                                                 TEMPERATURES >  32  - WHITE
                                                                                 TEMPERATURES <  25° - BLACK
                                    BILOXI BAV
                                                       INDEX MAP
                      APPROXIMATE SCALE  1:74,000
                                    iUA'JTITATIVE  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 +0.5°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.!25°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 stems 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 //I 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 sundry,  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 tht.
overflight aircraft or data platform traverses one (1) nautical mile ot 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 150mph 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

             $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 the'use 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 o'f 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
                 U. 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 compartment a ti on 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 analysisf 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
                MOSAICS
                           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
            INTERCOM?ARISON OF ALL  INDIVIDUAL LINES
                OF REASONING - DRAWING FURTHER
           	INFERENCES AND CONCLUSIONS
       V
                                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

-------
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.  ERTS 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 - S3

-------
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, or 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                  In each,  describe the physical expression
         2.   Hills                      in terms  of form, character, extent,
         3.   Plains                     boundaries, degree of dissection.  Look
         4.   Escarpments                for indicators  of these.   Describe fully^
         5.   Basins

     C.  Geology of Region - Origin

         1.   Transported origin

             a.   Wind                    In each,  look for the indicators of
             b.   Water                  origin,, movement, agent of movement  and
             c.   Ice                    deposition - The mechanisms responsible.
             d.   Gravity                Describe fully.

         2.   Residual  origin

             a.   Igneous                Look for indicators of  origin of those
             b.   Sedimentary            deposits  formed in place.  Describe
             c.   Metamorphic            fully.
             d.   Combinations

     D.  Climate of Region

         1.   Arid                       Look for the indicators of climate in
         2.   Semi-Arid                  erosion,  vegetation and land use.  Pay
         3.   Sub-Humid                  careful attention to: Location and
         4.   Humid                      distribution of vegetation; intensity
         5.   Tropic                     of erosion and redeposition; and
         6.   Polar                      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-pattern
          2.  Pattern type and Plan

      C.  Erosional Aspects
      G.
                               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.   B rush
              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.  If
                               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.

Cultural Aspects

1.   Man's activities
              a.   Urb an
              b.   Rural

III.   Analysis -  summations

      A.   Regional

          1.   Summation
          2.   Relationships
          3.   Interdependencies

      B.   Local

          1.   Summation
          2.   Relationships
          3.   Interdependencies

  FIGURE  2 (Cont'd)
                               Description of the type use and alteration
                               of the landscape.   Describe the intensity
                               of man's  activities.
                                      - 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 (th.e -use 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.


                                      v - 68

-------
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 th.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 jja 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 inf-
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 examplts 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-

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

-------
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)
1,000
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 by three 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 - 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 overcast 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 - 30

-------
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 weather 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_MATERIAI,Si 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.




     Previous investigations by Coakley1, Fantasia, et.al.2, Thurston,




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

-------
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 nm 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
                                TTT

-------
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
                                   _ P

-------
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.

-------
                             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 Minimum Fluorescence Response
Oil Type Gravity Major in Nanometers Decreasing Mini
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
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

-------
                     figure I.
                                                NQ2FUELOIL-
                                                NQ4FUELOIL-
                                                MAR.GAS OIL -
                                                MAR DIESEL OIL-
                                                MAR. LUBE OIL
X—•
300 Mi
                EXCITATION 254NM

-------
                                figure 2.
M

I
LEGEND
NO 2 FUEL OIL —
NO 4 FUEL OIL —
MAR GAS OIL — *
MAR.DESELOIL --
MAR. LUBE OIL — <
                       1            ^r

                          EXCITATION 290*M

-------

-------
   figure 4.
                          REFNAPTHENIC OIL
                          HYDRAULIC FLUID
                          GEN. MACHINE OIL •
                          NEUTRAL OIL
                          PREM. TURBINE OIL-
EXCITATION 254NH

-------
5*-
                            figure 5*
lFGEND
MAR. LUBEO!L(weighed)
MAR. LUBE OILfestimated)
                                 V
                     EXCITATION254MM

-------
         figure 6
|  FGFNID
N0.4FUELOlUveigh$d)--
NQ4 FUEL OlLJfestimated)-
EXCITATION 254wn

-------
90% \
!&%•
Jtf.
                                 figureT
LEGEND
MAR. DIESEL OlUweighecD -
MAR. DIESEL OlLJfestirnotecJ
                          EXCITATION 254wH

-------
I
(-1
CD
                              figure 8
LEGFND
0  hours  —
                                                      48 hours	
                                                      120 hours —*-
                                                      168 ho urs —°-
                                                      2 16  hours —*•
                                                   NO. 2  FUEL OIL
                             EXCITATION 254NM

-------
       figure 9.
LEGENKD
0  hours  -
                              48 hours	
                              120 hours —x —
                              168 hours —o—
                              216 hours —*—
                            MARINE LUBE OIL
330 *n
   EXCITATION 254NM

-------
           figure  10.
                                  UNKNOWN OIL	
                                  SUSPECT OIL	
                                  SUSPECT OIL— •-
                                  ACTUAL SOURCE-*
               OIL SPILL ANALYSIS
           J3L
fUO
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 spi11 .

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  oi1.

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 REMOTE DETECTION  AND  1DENT ! r iUrjON OF
                       SURFACE  OiL  Sf-'LLS
                                Ry
                          Herbert  R   Gram

                      Spectrogram  Corporation
                         385  State  Street
                   North  Haven,  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,


-n n d, surface currents, barge length, and point of transfer.  The


buoy could be either battery powered or powered from a dockside


console, but no s^'jck  hazard should be presented to operating per-


sonnel.   They desired  that the buoys be small, compact and light
           '"       r

weight.   The system' must operate day and night, year round, with


minimum maintenance  and service, and under the most severe of weather


conditions.  The  hoice of detection technique was left to Spectro-


gram,   hfw'..>*«r',  UP'  utility c.
-------
                                          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:   infra-

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.


-------
                                          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/o

manmade debri s.



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 is 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-f1uorimeter 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   2?

-------
                                          Spectrogram Corporation
                                          October 10, 1973
spectrum that was  entirely  unique  to  No.  6  fuel  oil.   We elected

to include in 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 lorporation
                                             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

prob1 ems .



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  n-ine months.

-------
                                          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.  I.
                SYNC SIGNAL
   Q
LAMP  POWER
   SUPPLY
  LOW PRESSURE
  MLRCURY LAMP
     G4T4/I
       FILTER HOLDER'
                                                  ©  &
PHASE  LOCK
 AMPLIFIER
                                    PHOTODETECTOR
                V
                                         MONOCHROMA10R
                   .^JL--, ^' MPLE
                 --17"" "1 THERMOELECTRIC COOLER
             SAM'-'LE
 ©
POWER SUPPLY
FIG.  I.
           LABORATORY  TEST SET UP
           FOR REFLECTION AND
           FLUORESCENCE MEASUREMENT
                                                                        STRIP CHART  RECORDER
                                     SPECTROGRAM  CORPORATION
                                                     iit.
                                                                   oj
                                                             No. Haven,  Conn.  06473
                                                                APPROVED DV
                                                                             BREAD"  BOARD "A"
                                                                                                       P-GdOI-OG'.
                                               VI -  3:

-------
                                                                                             FIG.  2.
                   SOURCE LAMP
FILTER
EXCITATION  ENERGY
                   QUARTZ  LENS
                   /          .FLUORESCENCE ENERGY
                                       COLLECTED FLUORESCENCE  ENERGY
                         SAMPLE
           FIG. 2.
       DETAIL OF OPTICAL
       FLUORESCENCE SYSTEM
                                              SPECTROGRAM CORPORATION
                                                     ooj  biaie 5t.
                                                No. Haven, Conn. 06H73
                                                                 E:  6-6—72
                                                                          APPROVED BV
                                                                         FLUORESCENCE  SYSTEM
                                         VI  - 33

-------
                                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.

-------
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 of "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  -  5!

-------
(F) the discharge contains titanium dioxide,  ferrous sulphate, and sulfuric acid.
(This information was gained by ground truth  collected during an EPA sponsored
studv.)  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 - 56

-------
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 s,uch 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
                                                             f
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
                                                                Band 2
                                                                0.6 - 0.7
                                                                Band 3
                                                                0.7  0.8
                                                                Band 4
                                                                0.8  1.1
                                                                Band 5
                                                                10.4  1
FIGURE 3  MULTISPECTRAL SCANNER IMAGERY OF A RIVER BASIN
                       VI - 42

-------
           Crude Oil Storage
    QS   Q  Q
Vapor Recovery
   Crude Distillation
Vacuum and Atmospheric
                                      Low Grade Gasoline
                                      Kerosene (To Treating)
                                      Diesel Oil (To Treating)
                                      CAS-Oil
      Deasphalting    Solvent
                  Extraction
                                 FIGURE 4  FLOW OF TYPICAL REFINERY
                                              VI  - 45

-------
            Zeiss RMK 1523 Camera
 Hasselblad
Camera Array
                                                                  i 2
                                                                  o
                                                                  -i  1
                                                                                        	Film Type 2475
                                                                                        	Film Type 2403
                                                                                        	Film Type 2424
\

                                                                        Density = 0.3 Above Gross Fog  I

                                                                           I      I      I      III
              Aero Commander                Cessna 336
              FIGURE 5  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
r
                                           FIGURE 7b LUBRICATING OIL REFINING
                                                     VERTICAL VIEW
                                                        VI -  44

-------
Deasphalted
      Oil
                                                                           Impurities
                                                                           Processor
         Treating
         Column
                    Furnace
                Stripper        Flash        Flash       Stripper
                            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
                                         -
                                              «J*   ).
                                               fa       /
                                                 ^^
                                                 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 I 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 yet
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, J. C., "Polarization:  A Key to Airborne Optical Detection of Oil 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. Biberman,
    F. A. Rosell, 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.  Welch7, 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.  Wobber, 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 Ms; 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

-------
en
o>
FILTERS
CO
-<
tr
UJ
5
LJ_
SY$
ILTERS
&.. - 	 —
T.V. CAMERA 1
STEM 2

T.V. CAMERA 2A

TV. CAMERA 2B
STEM 3

T.V. CAMERA 3
^ ,f PROCESSOR 1 ~[
"L (FALSE COLOR) J

PROCESSOR 1
(SUBTRACT,
FALSE COLOR, B/W)


PROCESSOR 2
i Mnni M ATIHM
IMUUULA 1 IUI\
ENHANCEMENT)
[
i
CAMERA 1
(B/W)
MONITOR 2
]_ (COLOR)



•,

MONITOR 2
(COLOR)

MONITOR 1
J

MONITOR 3
(COLOR)
                                  Figure 1.  Schematic of basic systems.

-------
Figure 2.   Conventional TV camera, with zoom lens, and monitor.

-------
<
H

 I



00
                                      Figure 3.  Two bore-sighted  TV  cameras.

-------
                            ™"   VI  - 59
Figure 4.   Processor,  monitor,  and auxiliary  equipment used  for  system  2,

-------
CTl
O
                        Figure  5.  TV  cameras  mounted in nose of Cessna 402, shroud removed.

-------
Figure 6.   Nose-shroud of Cessna 402 modified to accommodate  TV cameras,

-------
           10 r
                             HORIZONTAL
                             POLARIZATION
                             COMPONENTS
                                y        /
1
o>
LU
O

< C
|— CL)
 i/VERTICAL
 /'POLARIZATION
/; COMPONENTS
      LU
      LU
        0)
        Q.
           0
                                  BREWSTER   //
                                  ANGLE  FOR //
                        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.

-------
<
M


I
                                         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, area!
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 ej — ;e2, 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,
                             TA =
                                 =   47r
[
                                            (1)
                                            f (6, v) dtl
                                              VI - 67

-------
                                        NRL REPORT 7512
                                      2      3     4
                                        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 (e,
-------
                               HOLLINGER AND MENNELLA
                                       Table 1
                     Measured Complex Dielectric Constant of Oil
Oil Type
No. 2 Fuel
No. 4 Crude

No. 6 Crude

Temperature
19
26

26

Complex Dielectric Constant
f = 19.3 GHz
el: 2.10 ± 0.05
+ 0.02
e2: 0.01
2 —0.01
ex: 2.4 ± 0.1
e2: 0.06 ± 0.04
ej: 2.6 ± 0.2
e2: 0.05 ± 0.05
f = 69.8 GHz
e1: 2.10 ± 0.05
+ 0.02
e2: 0.01
2 — o.oi
ex: 2.2 ± 0.1
e2: 0.07 ± 0.04
ex: 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 fj.m 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: O9:32
                                     09:36
                                                                      10:09
                                                                                                                        10:46
                                                                                                                                                         to
                                                                                                                                                         o
                                                                                                                                                         to
                                                                                                                                                         H
                                                                                                                                                         -j
                                                                                                                                                         en
                     12:02
                                                                                                              13:27
                        | METERS|
                                               100
                                                              200
                                                                              300
                                                                                             400
                                                                                                             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
                CM
                5
                     -x."
                  10'
                  102
                                                          • SLOPE 0.6
                                         ' X
                             •X.'
                                                        _- SLOPE 0.2
                                                               III I II
                    I02              I03               I04
                                        TIME (SECONDS)
                      Fig. 4 — Area of the inner thick region (dots) and 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 Fig. 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 jum 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 GHz. ATa(°K)
                                                                                           3db BEAM SPOT
THICKNESS (mm )

-------
                            HOLLINGER AND MENNELLA
area of the visible slick of 33 X 103  m2 with a uniform film to thicknesses of 2 — 4 Mm.
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  - 75

-------
                                 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 DESCRIPT ION

     The "National Environmental Monitoring System': (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 197^4.  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 communicat ions-compatible,  will form a worldwide belt of Environ-
mental  Satel1 i tes.

-------
     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  in  our  streams,
rivers, and  lakes is  directly  proportional to the volume of water available
for dibtion  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 st-ations 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 M  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 rn rungedized
form for  omnatibility  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  a 1ternative.   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 suit-e  includes
meteorological  and oceanographic parameters such as currents, tempera-
ture profi1es,  etc.    When synoptic  reporting is required, the buoy
DCP become economically  mandatory.
                                     - 75

-------
     The  National  Weather Service Automation of Field Operation and
Services  (AFOS)  Program and related programs provides for automation
of bot~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 - A1askan Pipel ine tankers as we 1 1 as to improve weather
forecasthg 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 report
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 meteorologica
and 1imnologica1,  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-
 ng .
     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 trailers
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 syste-n
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 ]k  shows  that  the major components of the 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 15.

     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-timed and interrogated con-
figurations.  The  self-timed sets automatically transmit available data at
preset  intervals while  the interrogated set transmissions are commanded
trom_the CDA station.  The specifications for these sets are tabulated
in F igure 17.
                                  VI - 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 4db.  To  preserve the  link margin on the mobile OCR, 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  reinterrogated 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^-
                               VI - 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 Ganadian/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
include:

     1.  Automatic sample acquisition
     2.  Automatic sample pretreatment
     3.  Selection of  techniques compatible with automation
     if.  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 mu1ti-discip1 ine team of systerns-oriented  engineers
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 this technique,  it  is  not
applicable to buoy platforms but is very compatible with vessels and fixed
1 and p1atforms.
                                    - 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 appl ied 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 lithium drifted
silicon diode detector is applied to a mini- computer which functions
both as a mu1ti-channe1  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
printers.

     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 Water Resources  Engineering



    Conference.



2.  Bal linger, D. G.,  May 1969,  "Analytical  Instruments  in  Water  Pollution



    Control" presented at the  ISA Symposium (15th) on Analysis



    I ns t- rumenfat i on.



3.  Maylath. R.  E.,  April 1971, "Automatic Surveillance of  New Yorks



    Waters'1 presented a* the ISA Symposium (17"h) on Analysis



    I ns trumentat ion.

-------
            NATIONAL   ENVIRONMENTAL
               MONITORING   SYSTEM
              Seismic
              Stations
  Ships
   of
Opportunity
                Hydrological
                 Stations
                   a
                                      Command
                                      and Data
                                      Aquisltion
                               n
Weather
Stations
                           REGIONAL
                           STATIONS
                                                 Moored and Drifting
                                                Environmental Buoys
                               71  63

-------
                    HYDROLOGICAL STATION
I
CD
    FWD149-11

-------
en
01
                      reat  Miami River IGMR] Pollution  Monitoring Station

                                                          GMR Monitoring Station
Well  Station Water Sample
   FUD375-37

-------
                                 Pollution Monitoring Vessel
<
M
I
                                               AUTOMATIC
                                                SPECIFIC
                                               POLLUTANT
                                               MONITORING
                                                CONSOLE
                                                             SAMPLE
                                                            DISCHARGE
                     FWD149-10
SAMPLE
INTAKE

-------
         WATER QUALITY MONITORING MOORED BUOY
73-8-3538
                         VI - 87

-------
            Chlorophyll  Detection  and  Turbidimeter  Package
                 CHLOROPHYLL DET
         FLASHLAHP

        COLLIKATING
        OPTICS
         DETECTOR
  ELLIPSOIDAL MIRROR   SCATTERED RADIATION
                    DETECTOR
                                   CONDENSING LENSES
                                            BACKGROUND
                                          /' RADIATION
                                            SHIELD
            WINDOW
         COLLIMATED
         BEAM STOP
                DIRECT BEAM DETECTOR

TURBIDIMETER
SECTION        SPECTRUM-SHAPING
              FILTERS
  SECTION
                                          ELECTRONICS SECTION
                                          (AMP.  LD, & INV)      WATERPROOF SEALED CABLE TO
                                                              ELECTRONICS IN OIL DETECTION
                                                              PACKAGE
                                   -\_JL	JL__J
             TBTO IMPREGNATED
             GUARD RING
                                          TURBIDIMETER
                                         -ENERGY STORAGE
                                          BANK
                                                                                A
                                                                               CHLOROPHYLL
                                                                               DETECTOR
                                                                               ENERGY  STORAGE
                                                                               BANK
            CALIBRATION DYE
            RESERVOIR
                                                  BOTTOM VIEW
FWD375-22
                                                     VI  - 88

-------
                              METEOROLOGICAL STATION
t
CO
       0897TM

-------
           MOORED
ENGINEERING EXPERIMENTAL BUOY
        (40' DIAMETER)

-------
M
 I

-------

VI   92

-------
                Great Lakes Environmental  Monitoring System
iO
04
                            THUNDER
                            BAY Q
                                              TAWAS CITY
                                                   o
          PORT
         OAUSTIN
EPA, REGION V
DATA ACQUISITION
AND PROCESSING
                                                                           ^•ROCHESTER

                                                                               ^
SAGINAW
      ST. CLAIR
      RIVER  *
    DETROIT
        *
    MONROE
   TOLEDO^
        *
       MAUMEE
       RIVER
                         jL--,^^T-rr
                                                                   QERIE

                                                                QASHTABULA
        FWD319-13

-------
                                USCG  - Oil Spill Surveillance  System
                                                                            # POTENTIAL OIL POLLUTION
                                                                               SENSOR LOCATIONS
0897(TM)

-------
 Data Collection System Employing Interrogated DCPRS
                                            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

-------
             SATELLITE
2094.9 MHz
g = 13.7 dB
3
1
     2034.9 MHz
VG8505-11
                                                    Collection
                                    401.7—402 MHz
                                                          401.7—402 MHz
                                     468.825 MHz
                      401.84 MHz
                      PT = 10 W/40  W
                      g = 8 dB
                     1694.5 MHz
                     P  = 20 W
                    g = 19.1 dB  468.825 MHz
                                                                                FIELD OF
                                                                                  DCPRS
                                                                         MHz
                                               401.85 MHz
                                                                                     = 5 W
                                                                                      MOD -  ±70° PSK
                                                                                  DATA RATE  100 EPS
                                                                                  RECEIVE MOD = ±70° PSK
                                                                                  MANCHESTER CODED
                     1694.5 MHz
                           TRANSMIT
                           P  = 10 W
                           gT= 48 dB

-------
                    NOAA/NESS OCP RADIOS

-------
          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 UESSaGASE)
FIXED LOCATIONS
SELF TIMED
N/A
401. 7 to 401. 85 MHz
N/A
50
+ 70°PSK
MANCHESTER
5 WATTS
N/A
INTERROGATED
468. 825 MHz
401. 85 to 402. 1 MHz
1
150
+ 70°PSK
MANCHESTER
5 WATTS
-130 dBm
HELIX RH CIRCULAR
POLARIZATION 13 dB
GAIN
132 MW
+12.5 VDC
-20° to +50°C
6" x 12" x 18"
12JO& "" |
500 MW
+12.5 VDC
-20° to +50°C
6"xl2"xl8"
&L&S
MOBILE
468. 825 MHz
401. 7 to 401. 85 MHz
1
100
+ 70°PSK
MANCHESTER
40 WATTS
-135 dBm
CONICAL
SPIRAL 4 dB
GAIN
500 MW
+28 VDC
-20°to+50°C
6"xl2"x20"
28 IBS
g

-------
 1
I
a

-------
    INTERROGATION DCPRS COMMAND FORMAT

4
15 BITS
riME
CODE
31 BITS
CnbSF ADDRESS
4

TIME
CODE
           REPLY MESSAGE FORMAT
RF CARRIER ONLY
t— • crn •.
SEC •]
»— 	
ALT.
//(]
r i
2-1/2
__ crP^j
"-i>ECH
	 ?
15 BITS
31 BITS
DATA
MLS ADDRFSS
CODE AUUKtbb
MINIITF MAY
                    UNUESS RE INTERROGATED
                    FOR ANOTHER 2 MINUTE
                    PERIOD.
VG8505-8
                   VT  ICQ

-------
                      DCPRS/SC/CDA Detailed Frequency Plan
                              vo
                              en
                              Ol
                               fi
                              o
                              in
                              CO
                    U1 Ol O
                    O VO O
                    Ol in O

                    r-- en o i—
                    Ol Ol O O
                    Ol Ol O _J
                                   f-.
                                   i-»
                                   o

                                   oo
        CDA

        DATA DEMODULATOR

        BAND
vo
01
01

o
o
CO
         in Ol O
         o vo o
         01 in o

         r-. o o
            - in i
                                                                            oo
                                                                            01
                                                                            ur>
               CDA

               REPLY  IF

               FREQUENCIES
to
Ol
Ol

o
o
         LO Ol O
         o to o
         01 un o

         r-~ 01 o
         «*• <* Lfl
         co oo oo t—
          -  -  -o
         1—I t—I f—(	I
         O O O I-H
 UHF REPLY LINKDCPRS/SC
                                    SELF TIMED
                                    50 CHANNELS
                                    SPACED 3.0 KHZ
                             INTERROGATED
                             100 CHANNELS
                             SPACED 1.5  KHZ
                                                                            CO
                                                                            Ol
                                                                            Ol
 S-BAND REPLY LINK SC/CDA
                              01
                              vo
                       01
                       IO
                                  01
                                  vo
                              in
                              r^
                              vo
                              oo
                              vo
           O
           Lf)
           oo
           vo
           •=1-
            Lf)
            CO
            00

            oo
            to
           o
           o
           01

           00
           VO
           in
           i-~
           o>
           oo
           vo
                                                   10 KHZ
        UHF INTERROGATION
        LINKSC/DCPRS
    S-BAND  INTERROGATION

    LINKCDA/SC


NOTE:

ALL FREQUENCIES IN MHZ

o
in
u
c
o

•>
0
0

1
o
o
Ol

u
r
c

1
r> o
~. in
n o
co
o
CM
ro
o
CM
•a-
ro
O
CM
ro
O
CM
ro
o
CVJ
^219-14
                                    VI - 101

-------
    Data Collection System
CDA Control And Demodulator

            VI - 102

-------
           CDA GROUND  STATION  DATA  COLLECTION SUBSYSTEM
     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 CDA 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
                      0
                      0
                                                =>
                                                    o
                                                    o
                                                    ct:
                       O
                       Q_
                       o
                         LU
                         0_
                             ;
00

-------
                                 SAMPLE ACQUISITION
O
ui
                \
          WATER
                                      D. I. COLUMN
                                 SAMPLE
                                 PUMP
                    D. I. WATER
                    FLUSH TANK
THROTTLE
VALVE
CAPTURED
TANK      (V4>*-4
                                       HOLDING
                                       TANK
                        SAMPLE
                        TANK
                                                                    TO ANALYZER
      FWD319-9

-------
                  SAMPLE PRETREATMENT
       SAMPLE






1 	
1
1
1
1 — 1
|AA|
1 	 1
I
1
1
1
1
1
	 |___
1
1
1
^1
"~FE"I
i ht i
i 	 i






— i
i
i
i
i
rLn
! ASP 1
1 	 I
                                                                 r
                                                                   ECDGC
                                                                          I
                      	L	

                      DIALYZER
                                               I	.	I
                                               I
                                                          SONIC
                                                          CHAMBER
CHEMICAL
REAGENTS
                                  I	1
                                  I COLORIMETRY *
                                  I	J"'
                                                        FILTER
                                                        PAPER
                                                TO DRAIN
FWD460-5

-------
                                  ENERGY-DISPERSIVE X-RAY
                         ED X-RAY
                         PRETREATMENT
                          III'

-------
 AUTOMATED ENERGY DISPERSIVE X-RAY SYSTEM

FWD170-3
                    VI - 108

-------
<
M
O

vo
      ENERGY  DISPERSIVE X-RAY

       (Laboratory Configuration)
   i-12-1349

-------
                   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.
                         VII - 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 multlspectral 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. mi. 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.  mi.  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  (S191)
     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 S192 scanner provides  radiance  values  simultaneous in 13 bands
in the visible, near  IR, and  thermal  IR  portions  of  the Spectrum.  Each

-------
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 milliradians).  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.R°K
     The spectral bands covered are:
               Band                          Coverage, Micrometers
                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 areal  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 evaluation, mapping and inventory information, crop identification,
acreage estimation, irrigation needs, land productivity and crop vigor,
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 o_f altitude.

COASTAL ZONES, SHOALS AND BAYS - 12 Tasks
     Mapping hydrobiological  communities, sensing bay and coastal envir-
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 evaluations.

REGIONAL PLANNING AND DEVELOPMENT - 30 Tasks
     Land use mapping, crop identification, acreage mensuration, urban
studies, land classification, effects of strip mining, effluent water
patterns, recreation site analysis, water resource development and
management, transportation planning, assess fire 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 S-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, hov/ever, considering the use of
ground computer preprocessing techniques these data will be provided to a
quality comparable to ERTS.
                                 VII  - 9

-------
    NASA S-73-284?
i
H
o
          MULTICHANNEL
             SCANNER
MICROWAVE
RADIOMETER *>
                                                      CAMERAS
               L BAND
            RADIOMETER
 CAMERAS


                                    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 l) .  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
urn (green), 0.6 - 0.7 um (red), 0.7 - 0.8 um (near infrared), and 0.8 - 1.1
um (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  |_lj .  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  are 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  halimifolia.

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 JJ3] .  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 urn 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 [4] .  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)  Qf] .   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 Prove, 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.

-------
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  outfalls.

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 .  The serpentine  curve in
the ocean south of the center of the image represents a dumping of acid-iron
wastes, containing about 8.57o 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.

 f3]  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.

 [5]  Colwell,  R. N. :  Semiannual progress report to NASA, January 1973.

 [G]  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.

 [i]  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

-------
I
                                                                       ^^^^H \  .ji J
                                                               •• '        -^•E.     •
                                                               •^       ^*r^     >
                                                                       *i    *
                                                                              4


 Figure 2.  ERTS imagery of the Salt Lake  City,  Utah area:  0.5-0.6 \m band
   upper left, 0.6-0.7 urn band upper right,  0.7-0.8 pm band lower left,
                         0.8-1.1  um band lower right.

                               VIT - 13

-------
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 volleys
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
I HIGHLANDS: rugged,  partially forested  area
  with numerous lakes
M WASHINGTON:  level valley  (rural land use)
  enclosed by Highlands Ecoxone
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

                                           Thii photomap 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.
                         - ro

-------
Figure 5.  ERTS mosaic of Lake Michigan showing inadvertent
         weather modification by pollution plumes.

                               VII  - 21

-------
ro
ro
                 EXPLANATION
              /V   Areas strip mined
              &  pnor to 1968
                   Areas mined smce
                   1968. Mapped from
                   ERTS-1 imagery
                                                   IPIKE co.-
                                                ,  WARRICK CO
                                    Figure 6.  Indiana 3trip mine map and  enlarged ERXS image.

-------
M
H

I
                A.   (Scale  « 1:125,000)
                                                                  B.   (Scale  • 1:110,000)
                 C.  (Scale  = 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
                                            AFTER A STORM
                                           OCTOBER 11, 1972
NORMAL POLLUTION LEVELS
  SEPTEMBER 23,1972
                        l> •• VERY HIGH SEDIMENT LEVEL
                               IX,8,9, BLANK)
                        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.
                            VII - °5

-------
Figure 10.  Industrial waste aad sewage sludge  dumps  in the New York Bight  area.

-------
ERTS-1 DETECTION OF ACID
 MINE DRAINAGE SOURCES
         by
 Elliott D.  DeGraff and
     Edward  Berard
Arabionics. 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 Ambionics 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
                           CaCOs
                           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 Additive 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 - 33

-------
       The land use maps based on ERTS will be studied for




appearance of areas known to be subjected to mine pollutio




and similarities noted.   Areas of similar appearance, but




heretofore unchecked or with little previous record of




despoliation will be subject to surface and aircraft inves





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 .




long intervals, and too often sampling downstream samples i




pollution from unknown sources, whether surface or subsurf.




or broken seals.




       We have had great cooperation from the West Virginia




Department of Natural Resources and the Maryland Departmen-




of Water Resources in not only obtaining stream samples am




analyses on specific sites but also maps on farmed areas,




forests and cutover lands, county roads, trails,  paths, am




wet weather springs.





       The area office of the U.S.D.A. Soil Conservation




Service has an aerial photointerpretation section which ha:




recently completed land use survey maps based on aircraft
                            VIII - 34

-------
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 relatioi




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 acic




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 Basil




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 commol




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 - 36

-------
                  WATER QUALITY ANALYSIS USING ERTS-A DATA

                                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.

Hie results show that band III is useful in determining the water 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


which are F. Hulburt^  ', J. Williams^  ' , and S. Duntley^ ' .  Application of

                                                                             (k\
remote sensing techniques to oil slick detection have been made by N. Guinardv ',


J. Aukland*1 ',  J. Munday^ '.  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 NASA'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^  ' , B. Bowker^   ^and Rugles^  ' have


studied the dispersion of high sedimentation plumes from photographic images.
C. Wezernak*1  ',  H.  Yarger^15^ R.  Mairs^   '  and J.  Schubert   '  also have  investiga-
the surveillance of areas  with high sedimentation.


     The present paper  is  an additional contribution to the effort of assessing


water quality from digitized ERTSrA imagery.
                                  VII - 38

-------
                         2.   THE REFLECTANCE OF WATEE

     The  satellites  observations were made by using the Multispectral Scanner
(lySS) which has  four spectral bands.   The bands are defined in Table I.

                                   TABLE I
                             MSS  SPECTRAL BAWDS

      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 - l.l                     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  (l6  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
where R is the reflectivity of  the  target
     H Irradiance  illuminating target
     T Transmissivity of Atmosphere
     L. Path Luminance
                                      VII  -  39

-------
     It is important  in the  coarse  of our investigation to establish the order



of magnitude of these quantities.



     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 POND

                   Sun, Zenith angle  U9°, Date: 9-28-72 (From H. Rogers)
AND 2 L
mw/cm /sr
1
2
3
U
0
0
0
0
.U76
.2U2
.lUl
.234
bits
2l|
15
10
3
.3
.3
.3
.50
/ 2, LA ...
mw/cm /sr bits
0
0
0
0
.27U
.118
.082
.1062
111
7-5
6
1.5
T
.810
.865
.909
.913
R
9
5
2
0
• 3
.5
.8
• 9
H ,
mw/cm'
8. in
8. ill
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 IV is small. This 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  ,2n wide in the visible range.



The 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. Williamsv  ;



shows that the peak is at 3u and the penetration length is approximately 100 m.



The same data show approximately that average penetration length for band I and



II of NBS is approximately 10 and 1 meter respectively.  For natural water the



peak of the window is shifted towards higher wavelength and the penetration lengtl



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 III 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 TIDAL 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 wastewate


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 ko 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 VCD) 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 - 42

-------
     The data  is  shown in Table  III  and the  location of the  stations  is  shown




in Figures I and  II.   From the results  the following observations  can be drawn.




     1.  In the Key Bridge region a  high chlorophyl reading  (5^.8  ng/l)  and  low




Secchi Disk reading (26.0") probably indicate  a high concentration of algae.




     2.  in the lUth  St.  Bridge  region  a low chlorophyl reading (21.0 ng/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 Kk, 1.90 NE,  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 MJD).



     k.  In the I^tthias Point region an interesting phenomenon is  observed.




Going from stations 15 to 15A and l6 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
                                                                  ZONE I
Fort Belvoir
 STP lU M;D
                         Pentagon
                       STP 1 MGD
                       HTP k3 MGD
                      Arlington
                        20 MGD
                      PEPCO
                      HTP 315 MGD
                      Alexandria
                      STP 20 MGD
                        Westgate
                      STP 12 MGD
            Little  Hunting Ck..
              STP 2.5
scataway C
STP 10-12 MGD
          District of Columbia  STP 300 MGD
         Gunston Cove
                                                   River Miles from Chain Bridge = 15
                                                                Andrews A .F.B.
                                                                  ZONE II
                                       Figure I

                               Wastewater Discharge Zones
                                in  Upper Potomac Estuary
                                            VII - 44

-------
Chain Bridge
                                 Testing
                               / Stations                    MIDDLE REACH

                                      Potomac River Bridge
                                                                             Chesapeake
                                                                                Bay
                                           Figure II

                                  Potomac River Tidal System
                                                 VII - 45

-------
                                                   TABLE III

                                          WATER QUALITY MEASLEEMENTS

                                                DATE:  9-6-72
Stations
River   Time   Secchi Disk.      Water
Miles            Inches      Temperature
Total P   MH--N   DO    Total C   Chlorophyl   Salinity
          mg/1   mg/l    rag/l        ng/1         ppt
                                                 O/ 00
1.
2.
2A.
3.
1*.
5-
6.
7-
15-
15A
16.
Key Bridge
Memorial Bridge
ll*th St. Bridge
Mains Point
Bellvue "N8"
W. Wilson Bridge
Broad "N86"
Piscataway "77"
Nonjemoy F113
. Mathias Pt .
301 Bridge "CN"
98.3
97
96
9^.5
93
91
88
85
52
1*7
k3
11.20
11.35
11.1*5
11-55
12.05
12.15
12.30
13^30
11.15
10.55
16.17
26.0
30.0
3**.0
21* .0
22.0
lU.O
21* .0
21* .0
21* .0
33-0
1*2.0
23.0
23.5
21* .0
2U.5
2l*.5
21*. 5
2l*.0
23.20
2l*.8o
25.90
15-20
.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-^5
26.86
18.1*7
18.59
17.1*7
51*. 8
1*3.5
21.0
22.5
21.0
28.5
18.0
1*2.0
13-5
7-5
6.5








1.1*0
3.00
6.60
  Readings were taken at midstream
  Low Tide Slack Time 13.32

-------
                         k.  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
14 .3k
VEGETATION
30.61
25.^5
38.5
22.15
                                                                MMC. READING
                                                                     12?
                                                                     127
                                                                     127
     Notice that the radiance readings for water was very close to those which
Rogers observed for Barton pond 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

-------
 n»
 1*
 3 +
i, 3 + * * * »
51+*
52 +
53 +
5<.+
55 +
56 +
57 +
5fl +
59 +
60 +
61 +
62 +
63 +
&<.+
6? +
6f- +
67 +
6F +
69 +
        Fraction of  Samples

       ..!-'*?    t.L-'V    l.L-Jl    2.f-01    i.C-Tl
 n^*¥f$*u*v*iy-?«<<*  \	water

},) + *****..*>,* + **<.><»
15 +
16 +
17 +
1H +
19 +
ZT +
Z1 +
2'^ + *
27<
2&<
29 +
30 + *
31+ +
32 + *
33+#t*******+**
34++»*M*<.-»                                  ^Vegetation

3ftt*****««»Ct    ^	Environment 
-------
                              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 n also suggests that three




distinct brightness clusters exist.   (See figures  U  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  B).   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
                                   VTT - 49

-------
                6.  INTER ERETATION 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 U and 5) areas of high




sedimentation content or algae are easily determined because of their high




reflectivity and consequently brightness.  These areas are the Anacostia 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  ih St. Bridge area and Goose Island one where it is




known that  the Pentagon thermal plant (Us MGD) and the PEPCO generating station




(315 MGD) 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  every 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 EEA (Region 3) in gathering the ground truth and in the interpretation




of results.
                                   VII - 51

-------
, ...    . Pentagon Plant
 : ::. .....  STP 1 M}D
Waterfowl-^ . "i: :!•'::::..
Sanctuary :: : 	 •••
"'••:'•; "| ; '•'••
'•.':''':'.

i-i
'•:




Arlington ,:
STP 20 M3D *'" = =••••'•„•;
•'••• 	 »••»




-°



j:



;•'


»
;;
C,
•
••:
•>-
;'

«

ii,; i -
' . . !
B't "
i::^:;;'
:!t5-i-;!L'r

-i : • o '-'5
i.i ^a- .•
i-'S -:;•:;-

ill c'o''
'« o;::i
.50 .5....
:..,:,-,....
;;;;-;|!^0
..".5 	
..........e

:::!?i:::::i:
•".":;::::"::
                           ",  Anacostia
                            iX River
                                             w
                                      S un -Angle    hk

                                      Time          9.20 EST

                                      Date          9-23-72


                                  Histogram
                               Fraction  of Samples
                               ut-ov
         PEPCO
     Generating -
     Station
     HTP 315 >GD
      Alexandria
      STP 20
        Westgate-
        STP 12
                                                u
                                                2*
                                             w  n*
                                             co  5*

                                             S^"1-
                                             4J 11*
                                             |J H»i**»«
                                             pq i«;+^ti:-v^5^v:*T + ')4v
                                               ^*-.»1»*»4» •'"••••>'
       Blue  Plains
^iiiilll"  STP 300 M}D
                        21 +
                        22*
                        23 +
                     <::::.;;;;  •;:::::!. i:!;:1::;:-  Oxon Creek
                     ofo',cc,.iB *«o..i.! '  i!!!.'
                     e-o	oo- ^O.,B.   i|
£:::£;:£!:•.  Goose Island
                             i


                Wilson Brd.   I
                                     ;.      Broad
                                     '"•'••	::-..  • Creek
                                                        Mean Brightness   . ».17.0

                                                        Standard Deviation =1.68


                                                                 Legend
                                                            16
                                                            18
                                             s 15
                                             $.17
                                                                 Figure IV

                                                         Band II.  Upper Potomac
                                    VII - 52

-------
M
                                              Port  Tobacco
                                              River
Sun Angle .

Time

Date
                                                                                              w
                                                                                              ff
                                                                                                      9.2O  EST

                                                                                                      9-23-72

                                                                                                  	^Histogram   	  -!

                                                                                                     Fraction of  Samples

                                                                                                  o     l.t-'Jl   Z.t-01    3.E-wl
            ' *
            '•
            '•
.;::::::::::::::::::::::::r::::::::::::;::;;;^x;;;;;;:::::^.:::::
:iii!!:r:;!;i:!!!i:|iiiiE:iiii!iiv§1:iir"
, .D't^o-oioiBiio
p'%; -s:
Mathia
Boint

...•-j'.^Oi
1 in* ano-o
..afffo'.t
os'^o!!3!
.s -i;i

no
(-8
aa
ii
••
i

	 	 	 »,..,,»
	 	 	 !.,»..
i ! n'oi,'!ooosani(noiooooo-aj3o^i
* r nM!"i!!!!o!ji'''.!I^*p^m
                                                                                              1  i
                                                                                             -P  n
                                                                                                  *»
                                                                              lt...*l*lili..

       jfct  l'*
       m  is.
           16*
           17+
           l-"-»
          *l»«
           2?»
           21»
           2?»

                                                                 •••JiMM!!!!l!lr'»ai!*3;otiwa"ip,;;j:::r;;:i",':.'!"::;r":


                                                 Legend

                                                 i 13

                                                 £  .  fi 17
                                                                                                     Mean Brightness      16.^
                                                                                                     Standard Deviation    2.56

                                              18 £ 0


                                                                    .Figure  V
                                                            Band II.   Mathias Point

-------
   103,000
     0,000
    20,000 —
 Q
     5,000
I
en
     1,000
       50:
       200
       100
                               HURRICANE AGKES
                               PEAK
                               6/23    189,000 mgd
                               6/2U    230,000 mgd
                               6/25    1^1,OCX) mgd
                                                                      SATELLITE OVERPASS     	
                                                                                                1910 J.GD
                                                                                                9/30/72
                JAN
FEB
MAR
AER
JUIi
.JULY
AUG
00"
^ j.
                                                           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. Munday
 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 1^6.




"Optical properties of the sea", United States Naval




Institute, Annapolis, 1770.




"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-



"Remote 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", Uth 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. Haggles
13.  D. Bowker
Ik .  C.  Wezernak
15.  A. Liud
16.  H. Yarger
17-  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




No. 1+1, 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

-------
                                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.

-------
                             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
                               VTI - 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:




      °  suspended sediments in Delaware Bay averaged 30 ppm.   During




         July-August the average sediment level was 18 ppm.




      °  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.




         suspended sediments were  silt-clay sized particles with mean




         diameters around 1.5 microns.




         the predominant clay minerals are chlorite,  illite and kaolinite.




         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.1. , 1973, Bowker et_ a^. , 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 Multispectral  Scanner  suspended sediment




oatterns 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 - G5

-------
to four neighboring shades of grey.  In agreement with other  investi-




gators (Ruggles, 1973 and Bowker et_ _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, in




our discussion of circulation patterns, we  will emphasize  imagery




obtained in band 5.
                         T7II - 6r

-------
              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 and 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  5   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,

-------
                      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, Klemas, Polis, Szekielda, 1973).  Surface slicks and foam  collected




.it 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  Januaiy 25, 1973,




shown in Figure 13, disclosed a plume 36 miles east of Cape Kenlopen



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.
                            VIT  - 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
oy 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'iboundaries 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.  Raima 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 Commerce, 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  1
                                       SUBMERGED  CONTOURS OF DELAWARE BAY
                                       TONGUES OF DEEPER  WATER RADIATE
                                       FROM THE BAY ENTRANCE INTO THE BAY.
                         Clark Point;
                         Start* B*och'
       D • I a w a r •  B 6 y
         Seal*  in  M i I •i
                Conlouri  with
        D*pthi  in Fathomi
Cap* H•n TVp * n
Al  |   1 0 to 1   A4
A2  [   1 1 to 2   A5
A3  ["~] 2 to 3   A6
                                    VII -  76

-------
                                                                               TIDAL CURRENT CHART
                                                                                DELAWARE BAY AND RIVEB
                                                          •"~*  '  N\  1 °,9
                                                              \    '* V \       o«

                                                            
-------
                                                                                     TIDAL (TRUEST (HART

                                                                                      I>KI.AWAH>. HA1 AM) KI\KK
                                                             TWO HOURS BEFORE MA«IMUM FLOOD AT DELAWARE BAr ENTRANCE
Figure 3 -
ERTS-1  image of Delaware  Bay obtained in MSS band 5  on January 26, 1973,
and tidal current map.   (I.D. Nos.  1187-15140).
                                      VII -  78

-------
                                                            TIDAL CURRENT CHART
                                                             DELAWAU BAY AND WVHI
                           ONE HOUR BEFORE MAXIMUM EBB AT DELAWARE BAY ENTRANCE
  Figure 4
VII  -  79

-------
                                                                                 TIDAL Cl'RRENT CHART
                                                                                  UlUtWAHB BAV AND MIVBB
                                                            ONE HOUR AFTER MAXIMUM IBB AT Ol IAWARI BAY (NlRANi'.f
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 - 8O

-------
a)  0.475—0.575 microns.
                                                          b)  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

-------
                                               -CAPE HENUDPCN   ^-CAPE MAY
£
S
«
               ^
                   4
       '•CAFE HENLDPEN  ^—CAPE MAY

                DISTANCE
                                            CO
                                            U)
                                            o
                                                  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
9
10
12
13
14
                                                                          STATION
                 Figures   Correlation between ERTS-1  Image  Radiance,
                 Suspended Sediment Concentrations, and Secchi Depth.
                                               SEDIMENT(lm)
                                        !5ED!MENT
                                        (surface)
                                                     MICRODENSITOMETER
                                                           SECCHI DEPTH
                                                               (cm)
                                                     t>0
                                                     e
                                                                               Q
                                                                               W
                                                                               c/5
                                                                                       Id
                                                             o
                                                             u
                                                             w
                                                             CO
                                                    1.66
                                                    1.64
                                                  --I.62
                                                  --I.60
                                                    1.58
                                                  --I.56
                                                  •-I.54
                                                  -1.52
                                               -50
                                               -40
                                              + 30
                                               -20
                                               •ID'
                                               •0
                                       --10
                                      -15
                                      -20
                                      -25
                                      -30
                                      -35
                                      --40
                                        45
                                      4-50

-------
 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 boundary 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  COLI£CTION PLATFORMS FOR
                   ENVIRONMENTAL MONITORING

                       J. Earle  Painter
               NASA/Goddard Space  Plight  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 H 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

-------
SYSTEX 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 -

-------
                                                ERTS-I
   GoIdstone
H
H
U)
                                                                    DCP Installation
                              Fig. 1   ERTS DATA COLLECTION SYSTEM

-------
THE OPERATING IIETV.'ORK








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



1? good messages  per day.
                            VII -  94

-------
VD
(Jl
       0 _
160      140      120       100
                                                                                      20
                                   Fig. 2  ERTS-1 DCP NETWORK

-------
                                             TABLE  3
                                 GENERAL LOCATION OF ERTS-1  LOP'S
                                             22 STATES
<
H
ALABAMA
ALASKA
ARIZONA
CALIFORNIA
CONNECTICUT
DE LAV/ME
FLORIDA
HAWAII
KANSAS
LOUISIANA
MARYLAND
MASSACHUSETTS
MICHIGAN
MISSISSIPPI
NEW HAMPSHIRE
OHIO
OREGON
PENNSYLVANIA
TENNESSEE
VERMONT
VIRGINIA
WASHINGTON
                                        3  FOREIGN  COUNTRIES

                                            CANADA
                                            EL  SALVADOR
                                            GUATEMALA
                                            ICELAND
                                            NICARAGUA

-------
                                              TABLE  4
                                    ORGANIZATIONS USING ERTS-1  DCS
H
H
     ORGANIZATION

FOREIGN (CANADA)

U.S. GEOLOGICAL SURVEY

BUREAU OF LAND MGMT.

FORESTRY SERVICE

CORPS OF ENGINEERS

NAVOCEANO

UNIVERSITIES

STATES

NASA

INDUSTRY
                             TOTALS
NO. INVESTIGATORS
6
10
1
1
1
1
4
1
3
1
29
ASSIGNED DCP'S
U
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  OF  ERTS-1  DCS BY APPLICATION
<
H'
00
APPLICATION
METEOROLOGY
HYDROLOGY
WATER QUALITY
OCEANOGRAPHY
FORESTRY
AGRICULTURE
VOLCANOLOGY
ARCTIC ENVIRONMENTS
NO. OF USERS
5
20
4
3
1
1
2
1
DCP'S ASSIGNED
10
75
26
9
3
3
33
2

-------
H
H
                              TABLE  6





                 PARAMETERS MONITORED BY ERTS-1  DCS





RESERVOIR LEVEL                                   WIND DIRECTION


STREAM FLOW                                       WINDSPEED


GROUND WATER                                      HUMIDITY


TIDAL VARIATION                                   PRECIPITATION


ICE CONDITIONS                                    SOLAR RADIATION


SALINITY                                          SNOW DEPTH


DISSOLVED OXYGEN                                  SNOW WATER CONTENT


TURBIDITY                                         EVAPORATION


ACIDITY-ALKALINITY                               SOIL MOISTURE


BIOLOGICAL CONTENT                               EARTH TILT


WATER TEMPERATURE                                 TREMOR EVENTS


AIR TEMPERATURE                                   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 (Fig. 3).  These data are relayed



through ERTS-1 to Goddard Space Center, then forwarded via



teletype to the USGS offices in Harrisburg, Pennsylvania,



where they are processed and distributed dally 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 reference 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.Q. monitor and the



presence of conducting material on and around" the four probe




conductivity sensor.




                      VII - 101

-------
                           Fig. 3  ERTS DATA RELAY

                         DELAWARE RIVER BASIN NETWORK
 1.   Del. River near
     E. Stroudsburg,
     PA.

 2.   Del. River at
     Easton. PA

 3.   Lehigh River at
     Easton, PA

 4.   Del. River at
     Trenton, K. 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

11.   Del. River at
     Reedy Island
     Jetty, Delaware

12.   Del. River at
     Ship John Shoal
     Lighthouse
                    39- _
                                       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 Army Corps of Engineers, New England Division deployed



three water quality DCS installations on the North Nashua



River at Fitchburg, Mass., the Chicopee River at Chicopee,



Mass., and the Westfield River at West Springfield, Mass.,



(Fig. 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 OP ENGINEERS EXPERIMENT

                          2,

                           I
                               72.5°
                                Quabbin
                               Reservoir
        Northampton
                         ^ . Chicopee
                         T  River
Westfield
  River
                           Chicopee
  Westfield
             West
             Springfield
                                    Springfield
                                Massachussetts
                                  Connecticut
      Connecticut
         River
                          Hartford
                                                              •42'
                   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 of 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


77°
76<
                                         Choptank
                                          River  i
              Rappahannock
                  River
                          VII - 107

-------
calibration tank prior to cleaning.  A thick algae coating



on the D. 0. membrane had little effect.







The Rappahannock site was primarily affected by algae



accumulation; 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 Mr. 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 ana 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 - 3

-------
       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

-------
      SUN
                SKY
CAMERA
                           B
                r\\-//
    WATER
               -  -\ -/
                     \/

                  BOTTOM
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.  Spec tropho t omet er
       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
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
                                           ILLUMINATING LAMP
                                           FOR TRANSMITTANCE
                                                 WORK
  Fig.  3.  Spectrophotometer block  diagram for analyzing
  water samples.
                          VIII - 9

-------
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 reflectance.  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

-------
Pig. 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
  2
  U
10 ••
     1.0--
  W
  05
  CU
     0.1"
                          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.
                       VTTI - 13

-------
W
U
    EH
    Z
    W
    U
    K
    W
    fc

    2
  EH U
  U U
  ft EH
  W U
  g
    r<
    04
               TOTAL  SUSPENDED SOLIDS,  MG/L


              0.1    1    10   100  1000   104 105 106
     100
      10 ..
       1.0 ..
       0.1 .....
Field Conditions
From Analyzing
Photographs \
                                 Laboratory
                                 Conditions
                               From Analyzing
                           Samples With The
                           Spec tropho tome ter
                   10   100   1000  104  105  106
                       TURBIDITY, JTU'S
Fig. 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.
                             - U

-------
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
   2  .-H
   W  H
   U  fr,
   K
   W  05
   ft  H
   H  O
      J
   w  o
   U  U

   1  2
   EH  O
   U
   H
   W
      2
      O
      U
      H
   EH  S,
   2
   W  m
100
         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 can
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

-------
Pig. 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.
                           VTTT - 17

-------
                                              TURBIDITY, JTU'S
I
M
•co
              OP

            L«  V
              W
            W U
EH U
U W
W i-3
1 *1 TV f
ft, W
            W !Z
            S W
            ID «
            ^ <
            O CU
            > CM
                  100
                   10
      1.0 -
      0.1
                                          5
                                          -4-
                                  50   500
                                  -H	1	
                                FIELD
                              CONDITIONS
                                                             Dry

                                                     A       Wet Clay*0
                                                                LABORATORY
                                                                CONDITIONS
 BOTTOM EFFECTS
  FROM SAMPLE TUBE  USED
_•	1__	1__
                                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
                                VIII - 19

-------
      w
      u
      z

      EH
      U
      W
      W
      W
      U


      S 0.5
          5 ••
                               -4-
                        -I	1  I  I I I I I
                                             100
      © =
      m =
      w
      u
      2

      IS
      U
      W
      §

      EH
      2
      W
      U
      «
      W
—t	1	
    5    10     20      50

      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 ..
        0.5 -
              t  I ' ' ' I	1	-)—i—I  i Mil	—

              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 - 20

-------
                         100
—I
M
ro
H
                      cn
                      W
                      a
                      u
                      2
                      H
                      cn
                      M
                      Q
U
cn
H
Q
a
u
u
u
en
    50- •
                          10- -
                                                          1972
                                                  November  1972

                                                   November 1972
                                             10

                                              TURBIDITY, JTU'S
                                               50
                                                           -. 100
                                                           • • 50
  - • io
                                                                                 -• 5
100
                                                                en
                                                                Q
                                                                H
                                                                J
                                                                O
                                                                cn
        P-i
        cn
        D
        cn
                           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
U
w
3
W
u
U
W
  U
  tf
  u
       10."
        5 -.
        2 ..
      1.0 ••
      0.5 ••
      0.2
                   1
                   -h
           Apparent Reflectance
               FIELD DATA
                  Volume  Reflectance
                      LAB DATA
                            5         10

                         TURBIDITY,  JTU'S
                                     20
Fig. 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
  B

  C
  D
Distilled Water

Snowbank Lake

Ensign Lake

Shagawa Lake
NA

Oligotrophic  (clear)

Mesotrophic(Middle Stage)

Eutrophic (enriched)
                         •Til

-------
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 -

-------
                                   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
                                      VITI - 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," 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.
                           VIII - 2?

-------
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
                             -ITT - 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



          P.. = Subsurface Component



Where i,s 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
                            VIII - 31

-------
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



0.04ym 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  = 6600po cm ^
                      a.        3
                         VIII - 32

-------
I

CSI
           Output     Chopper
             Optics
       Linear Polarizer
                                                        Laser
                                Light Shield
                  0
Collimating Lens
  I  Spectral Filter
  I   (Beam Splitting Polarizer
   Telephbto   Spatial
     Lens       Filter
                                                                     •Photodiodes
                 Figure '1
Schematic diagram of Lidar Polarimeter.

-------
      LIDAR
   POLARIMETER
Figure  2 - Lidar Polarimeter.
                 VTTI - 54

-------
where p  = weight  (grams) of teflon per cc of solution
       -O


      Pa = 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 beam 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



ci 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 o£ solution.  Concentration of scatters from 23 ml
                           VIII - 35

-------
-20-
-30 -
-40 --
-50
                                                   W
 4

P. *
                                                      10
                              ~
       Figure 3 - Polarized and depolarized backscatter
                  vs. mass concentration  of scatterers.
                  Angle of incidence equals 23°.
                            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, just as with the first



solution.



          The lidar polarimeter 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 - 37

-------
1.2-
1.0.
.8-
*>
i
o ,
iH . 0 -
O
-4.
.2-


Q
© 6 = 20°
•
I) *
© @
ft* ®
^

6*: 	 — r— - — 	 ••- - -•
50 100 150
mS, of scattering solution
5 i of H20
Figure 4   Polarized and depolarized backscatter
           vs.  mass concentration of scatterers.
                  VIII - .7.8

-------
scatterer density of approximately 1.0 mg/cm3 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

-------
,0
•d
® "80
<*> "180
10 -






20 -
<
30 J
(
(

C

40 -




50 .

© "220
© "320
grit"
grit"
grit"
grit"
© "smooth"
O "80
© "180
0"220
$"320
grit"
grit"
grit"
grit"
® "smooth"
g>
^ ©s> ^
•^kCSh.-, R^siA^ e.® ® &
^e^"*® t,»P &;3>
(£) rf55?!*^ X^v-r
r^$jr C,^) r*\
©^ S j58
® 0 US

©
©
® ® © ~ _
© ® ®
1 1 II 1
{ 1 1 1 1
10 20 30 40 50

>*«





**+
P Q *
@ C©
©

© ©
/
1
60
White
White
White
White
White
Black
Black
Black
Black
Black

®
«s» ^
^gsr cnf)





0 ^
^>
§o
Q,

®
®
1
1
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
-30 ._
-40 ._
-50
-60
                       ® "80   grit"
                       «> "180  grit"
                       © "220  grit"
                       • "320  grit"
                       • "smooth"
                       ® "  80  grit"
                       © "180  grit"
                       0"220  grit"
                       * "320  grit"
            White
            White
            White
            White
            White
            Black
            Black
            Black
            Black
                    © 
                                 O
                                               +
 	(	1	1	1	h	H—

     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-on-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
                               VITI - t:i

-------
              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 of these terms adds incoherently to the total back-



scattered power received for each polarization, that is:



                 P. . = P  + P . + yP  .
                  •2-t-    S    Ot     Wl



                 P. . = P . + yP  .
                  z-j    03     wj

                                                          5

where:           P.. = like-polarized backscatter
                  It- 'Z'


                 P.. = depolarized backscatter
                  *J


                 P   = oil surface term
                  •6


                 P   = oil volume term
                  o


                 P   = water volume term
                  w


                 Y   = attenuation of oil layer (two-way)

                                                 i



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
                            III - 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 o£ 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
developed for the Coast Guard under contract DOT-CG-34017-A.
"The opinions or assertions contained herein are the private
ones of the writer and are not to be constructed as official
or reflecting the views of the Commandant or the Coast Guard
at -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.
                           VITI - 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 a'nd 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
of the ocean waters (Specht1)-  The data for the attenuation (Mairs2) and
the extinction (Hale3) properties of water are also collectively available.

                                    VIII -  49

-------
    1X3
    o.s
   0.13
   0.10
The transmittance data for the four  above  mentioned types of water  is
plotted in Figure I1.  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.
   100
    so
    Z3
    Z-5
    350  40O  450 30O 55O  60O  650  70
              Wavelength, nm
   FlG. 1. Spectral transmittance for ten meters of
 various water types.
THE AIRBORNE DETECTION TECHNIQUE
Recording Technique
     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 blaqk, 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
                                      VTII - 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 Densitometer.  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 (0.7u to l.Oy).  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 P](x],y-|) for
high dissolved oxygen, and P2(x2,y2) for zero dissolved oxygen, and
where   L  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:
               yBlue  =  2.655xBlue + 10.462                             (3)
               ^Green =  3'715xGreen + 10'866                            <4>
               xBlue» xGreen  are tne b1ue» 9reen film densities, respectively
                   e' ^Green  are dissolved concentrations for the blue,
                             green lines, respectively, in parts per
                             million (ppm).

                                   VIII  - 5*

-------
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 densitometer, 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.
                                      VTII - 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 Cos126, 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 IQ is the optical energy level at
                                VTTT - 5«

-------
the center of the film format, then the optical energy distribution
across the film format is:

               I(x)  =  I (A)A(e,x)S(e,x)de                           (8)
                         0   o                               .           '
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)v
                                                       A
             S(e,x)  =  Solar illumination function

S(e,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 Required
     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 tlppm 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,  G.  M.,  and Querry,  M.  R.,  "Optical Constants  of Water in the
    200 nm  to 200um Wavelength Region," Applied Optics, Vol.  12, No. 3,
    March  1973.
                                  VIII - 62

-------
9-
8-
7-
6-
5-
4-
3-
1-
Figure 2
Characteristic Curve.
Background Water.
                  1.0
                                    2.O

                                Film Density, X
            3.0
                             4.O
                                  VIII - 63

-------
10-1
                                         Figure  3
                                         Difference Curve,
                                         Background  Water.
 2-
 1-
— O.5O
         -0.25
1.OO
                                                             1.25
                                    A
                               VIII - 64

-------
10-
 9-
           Figure 4

           Characteristic Curve,

           Lignin Effluent.
                   1.O
    2.O


Film Density,  X
 i
3.0
                                                                         4.0
                                   VIII - 65

-------
1O-I
 9-
 8-
 7-
 6-
 5-
 4-
 3-
 1-
              Figure 5
              Difference Curve In
              Lignin Effluent.
 -O 50    -0.25     O
 I
O.25
                                      O.5O
 I
O.75
                                                        1 .00
1
1 .25
                                     A
                                VIII - 66

-------
A SEARCH FOR ENVIRONMENTAL PROBLEMS OF THE FUTURE
                       by
        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
          c  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	1	1
                        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 prioritize 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.

-------
          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
fron 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 bbI/day)
Case*
Best
Worst
1970
3.4
3.4
1975
7.2
9.7
1980
5.8
16.4
1985
3.6
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          Case I           Case II
        (million bbl/day)     No  Superports     Superports*


                3               15,900           3,400

                6               32,600           6,500

                                       	I0j,000	
       Average number of
         bbl  spilled per
         million bbl/day
         imported                5,400           1,100
       * With  transshipment  by  pipeline.
                                  VIII - 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,

-------
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, e_c.,
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
                                                   (8)
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
*  Data for 1968-1971 period




** Excludes automotive  sources
                              (9)
                         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 - 80

-------
 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).
                                      YIIT - 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
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 - 85

-------
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
                                                                X
               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       TV      Radio

            1945       6        930
            1950       97      2832
            1958      421      3310
            1960      562      4256
            1965      674      5537
            1971      892      6976
                        VIII - 3fl

-------
          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 particulate emissions by quantities and
                                                   (TQ\
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.
                                    VTIT -  30

-------