EPA-600/3-77-126c
                                           November 1977
  GREAT LAKES ENVIRONMENTAL PLANNING USING

       LIMNOLOGICAL SYSTEMS ANALYSIS:

           MODEL SPECIFICATIONS
                   by
            Hydroscience, Inc.
            363 Old Hook Road,
           Westwood,-New Jersey
                  07675
             prepared for the
       Great Lakes Basin Commission
            Ann Arbor,  Michigan
                   48106
    Contract Number:  DACW-35-71-C0030
                    and
ENVIRONMENTAL RESEARCH LANORATORY - DULUTH
   OFFICE OF RESEARCH AND DEVELOPMENT
    U.S. ENVIRONMENTAL PROTECTION AGENCY
       DULUTH, MINNESOTA 55804

-------
                    DISCLAIMER
This report has been reviewed by the Office of Research
and Development, U.S. Environmental Protection Agency,
and approved for publication.  Approval does not signify
that the contents necessarily reflect the views and
policies of the U.S. Environmental .Protection Agency, nor
does mention of trade names or commercial products con-
stitute endorsement or recommendation for use.
                        11

-------
                            FOREWORD

Man and his environment must be protected from the adverse effects
of pesticides, radiation, noise, and other forms of pollution, and
the unwise/management of solid waste.  Efforts to protect the environ-
ment requires a focus that recognizes the interplay between the compo-
nents of our physical environment—air, water, and land.  The Office
of Research and Development contributes to this multidisciplinary focus
through programs engaged in

       .studies on the effects of environmental contaminants
        on the biosphere, and

        a search for ways to prevent contamination and to
        recycle valuable resources.

This report assesses the technical feasibility and economic practical-
ity of developing mathematical models to assist in defining and making
selections among alternative management strategies and structural solu-
tions proposed for solving water resource problems of the Great Lakes.
.The deliberate decision-making process reported is a milestone in pre-
application analysis of modeling for natural resource management pur-
poses.  ;
                             iii

-------
                             ABSTRACT

     The report documents the deliberate decision making process used
by the Great Lakes Basin Commission in concluding that rational modeling
methodologies could be used to evaluate the effect of different planning
alternatives on the Great Lakes and that planning for specific problems
affecting the Great Lakes system can be technically and economically
supported through mathematical mo'deling and systems analysis.   It assesses
the technical and economical feasibility of developing mathematical models
to assist in making selections from among alternative management strate-
gies and structural solutions proposed for solving water resource problems
of the Great Lakes.  The study reviews, evaluates and categorized present
and future water resources problems, presently available data, problem-
oriented mathematical models and the state of models and model synthesis
for large lakes.  A demonstration modeling framework for planning is devel-
oped and applied to western Lake Erie and the Great Lakes system.  The
report evaluates four widely ranging alternatives for future modeling ef-
forts in the Great Lakes and recommends the modeling level most feasible
to answer planning questions on scales ranging from the Great Lakes to
regional areas.  Also discussed is a proposed.Commission study which will
apply limnological systems analysis to the planning process.

     The report consists of three volumes:

        a.  Summary
        b.  Phase I - Preliminary Model Design
        c.  Model Specifications
                              IV

-------
                           CONTENTS

Foreword	.. ..	iii
Abstract	iv
Figures	. . ..	vii
Tables	vii

     I.  Specification for Main'Limnological Systems
           Analysis Programs	1
          Background	1
          Specifications	5
    II.  Great Lakes Modeling Computer Programs.	14
          Program Purpose	.	14
          Spatial Settings	14
          Program Design Philosophy	15
          Required Program Features	15
          Desirable Program Features	. ...J	16
          Phases of the Development Proj ect	17
          Technical Specifications	17
          Contractor Requirements	;	18
   III.  Great Lakes Scale Application	19
          Scope of Application	20
          Study Input Data	 . . J	20
          Modeling Framework	'.	21
          Verification Analysis		21
          Application of Model	;	22
          Final Report	23
    IV.  Lakewide Scale - Lake Erie	 4	23
          Scope of Application	— '.	23
          Study Input Data	24
          Modeling Frameworks	; i	25
          Verification Analysis		26
          Application of Model	...1.'	26
          Final Report	27
     V.  Duluth Area of Lake Superior - Regional |
           Scale Application	'.	27
          Scope of Application	*'	27
          Study Input Data	I	27
          Modeling Framework	28
          Verification Analysis	29
          Application of Model	29
          Final Report	I	29

-------
                         CONTENTS

  VI.  Southern Lake Michigan Regional Scale
         Application.	30
        Scope of Application	30
        Study Input Data	. ..	 31
        Modeling Framework	^	32
        Verification Analysis		33
        Application of Model	33
        Final Report	 .:	33
 VII.  Saginaw Bay Regional Scale Application	34
        Scope of Application	.....34
        Study Input Data		34
        Modeling Framework	 35
        Verification Analysis	. .. . .	35
        Application of Model	36
        Final Report			36
VIII.  Green Bay Regional Scale Application.	37
        Scope of Application	37
        Study Input Data		'.	37
        Modeling Framework	•	38
        Verification Analysis	38
        Application of Model. .  .	39
        Final Report		39
  IX.  Food Chain Modeling	 . ..'.	40
        Scope of Application	40
        Study Input Data			40
        Modeling Framework	..'.... 41
        Verification Analysis		42
        Application of the Model	42
                             vi

-------
                            FIGURES

Number                                                      Page

  1     Methodology for Great Lakes Limnological	2
          Systems Analysis	

  2     Water Resource Problems and Mathematical	4
          Models Included in Phase II Program	

  3     Primary Input - Output Variables	8

  4     Sys tern Diagram - Lake I Model	9

  5     A Ten Compartment Model	10

  6     A Ten Compartment Model With Spatial Definition	11

  7     A Food Chain Model	12

  8     Interfacing of Eutrophication and Food Chain	13
          Models	

                            TABLES

  1     Summary of Recommended Model	7
                              vii

-------
                        SECTION I

                  SPECIFICATION FOR MAIN  •
          •LIMNOLOGICAL SYSTEMS ANALYSIS PROGRAMS
                       Background


A study,has been completed and a final report issued on the
technical feasibility and the economic practicality of
applying- a Limnological Systems Analysis (LSA) to water
resources problems in the Great Lakes.(1)  Specific attention
was directed to an evaluation of the state of the art of
modeling as it applied to interrelated water resource prob-
lems.   The overall purpose of the study was to indicate the
degree of understanding of limnological phenomena, as affected
by both nature and man's activities, and to evaluate the degree
to which these processes can be expressed in a valid mathematical
form within a system analysis framework.

The overall methodology followed in the Limnological Systems
Analysis- of the Great Lakes is presented in Figure 1.  Two
parallel paths were followed.  The first line of analysis
evaluated :the present and future water resource problems and
water use 'interferences with their associated water resource
variables.1  The second line of analysis evaluated presently
available -data, problem oriented mathematical models, and
present state of the art of models and model building which
are useful for a Limnological Systems Analysis.  The two
lines of analysis were synthesized into a problem and model
ranking of priority from which feasibility recommendations
were drawn.  A demonstration modeling  framework was constructed
in order to illustrate the output of a Limnological Systems
Analysis in several problem contexts.

Water resource problems in the Great Lakes were identified and
grouped into seven problem categories  as follows:

     1.  Monthly Lake Water Levels and Flows
     2.  Erosion, Sediment
     3.  Ice
     .4.  Toxic and Harmful Substances
   '  5.  Water Quality
     6.  Eutrophication, Fishery
     7.  Public Health
   Hydroscience,  Inc.,  "Limnological  Systems  Analysis  of
      the Great Lakes  -  Phase  I,"  (1973).

-------
 PLANNING  n
AD1.' '. t J <::.-
  :..'.f.~ • T'S
SCIEN riFIC 3
 Af.-^.Ci <>
ICEIUiFlCuTIOtJ CF
  V.'ArER  RESO.RCE
'J?L'3 <* PRCHLrf/S

   (SECTION  IT)
    DEnr: lion
         OF
      MODEL '  .
   . BY FhCBLEt.lS
                                                                       DETERI/lliATIOII OF'
                                                                          F'hOblT.M
                                                                          l-'RIORl T (•
                                                                        (SECTION V;:P
r.EriniTioii OF
V/M-.lAUI.F.S
H-.0 = LE!.:S
(SECTION H)












p.»


EXTENT AMD
AV^ILAEILIT f CF
ri ,* T -'i
(SECTIOM HI!
                                                AVAILABLE
                                                MODELS
                                              •" ST.'.! E OF AhT
                                                 REVIE.V
                                               (SECT I Of.1 "if)' •
   MODEL
 EVALUATION
     ft(O
  SYN1 l-.tSIS
(SEC'T:.-J\ \?
                   DS!':Oi: STRATIO N    l.iODEL
                      (SECTION VTT- AI'PEii!'' X A!
                                                                                         LELHIEATIOH  OF
                                                                                           PROBLEM
                                                                                         CATEGORIES
                                                                                              FOR
                                                                                          PHASE D
                                                                                         (SECTION VTTi)
                                                                                                              OF
                                                                                                         ALTERNATE
                                                                                                          PROGRAMS
                                                                                                         (SECTIOM  lill)
                                                                                        RECOf/MENDED
                                                                                          ' PliASE n
                                                                                          "
                                                                                                              I/O t/ E L S
                                                                                                             APPUCATK ;:S
                                                                                                               FUIJC IIJG
                                                                                                               TIMIMG
                                                                                                            (SECTION VflJ)
                                     FIGURE  I
METHODOLOGY  FOR  GREAT   LAKHS  LIMNOLOGICAL   SYSTEMS   ANALYSIS

-------
For each problem category a detailed review was made of
associated water uses and water variables to provide the
link to the available models.

It was concluded that of the seven problem areas, the ice
category and a portion of the public health category (near
shore.pathogen problems) are generally not Type II planning
problems.  The Great Lakes problems associated with (a) lake
levels and (b) erosion and sediment are being analyzed and
are adequately modelled in various ways for present needs.
The ranking of the four remaining Type II planning problems
(which was subjectively established) is as follows:
     1.  Eutrophication
     2.  Water Quality
     3.  Public Health (regional and lake wide scale)
     4.  Concentrations of toxic or harmful substances

The formal modeling structure proposed for Phase II is com-
posed of a broad scale framework.consisting of seven inte-
grated modeling subsystems:  water balance, lake circulation
and mixing, chemical, eutrophication, dissolved'oxygen,
pathogens, and ecological, as shown in Figure 2.

Alternate Limnological Systems Analysis programs were
evaluated  (1973) in order to explore/varying levels of effort
and cost for a Phase II study.    •    ,

     1.  Level 1:  This alternate is•estimated to cost
         0.7 million with a two year completion time
         and represents the lowest level at which a
         meaningful Limnological Systems Analysis can
         be carried out.

     2.  Level 2:  This level is estimated at a $2 million
         cost with a three year completion time 'and repre-
         sents a favorable balance between problem contexts
         that can be approached rapidly, given the present
         modeling status, and those problem categories which
         have high priority but for which modeling frameworks
         must be significantly advanced.

     3.  Level 3:  The cost of this level is estimated at
         $3.9 million with a three year completion time and
         represents a more intensive effort than Level 2.
         Level 3 funding is felt to be the maximum amount
         that can be prudently spent. for a Phase II study of
         the use of a Limnological Systems Analysis for the
         Great Lakes.

-------
Eutrophication
VJater
 Quality
                     1
1

•VJater

.Balance.







Lake
Circula-
tion
Mixi
&
ng








Chenicc

Public
Health
                            PI1ASF. II PROGRAM

                      1.  Computer Proqram
                         Development

                      2.  Planning Applications
                         a. Great Lakes Sca]e
                         b. Lake-wide Scale
                         c. Regional Scale
  Toxic  &
  Harmful
Substances






AV












Pathogens
& Virus
Indicator
Bacteria






ecological
&
Food
	 Chain
                     MATHEMATICAL tlODKLIIIG
                              FOK riiAsi; TI
                              FIGURE 2
      WATER RESOURCE  PROBLEMS AND  MATHEMATICAL  MODELS
                   INCLUDED  IN  PHASE H PROGRAM

-------
                     Specifications
It was recommended that a Phase II Limnological Systems
Analysis study be funded at the $2.0 million level with
a three-year completion time.

Within this level, it was recommeded that:

       1.  Existing subsystem models, parameter values, and
           inputs be gathered into interactive modeling
           frameworks.

       2.  Generalized computer programs be developed and
           modifications be made to existing models to
           accommodate recently evolved numerical and
           software techniques.

       3.  Applications be made of existing systems technology
           to those problem categories for which a reasonable.
           degree of success for the application is assured.

The following specific problem contexts were recommended for
inclusion in the Phase II study:

       1.  Water Quality Problems
             a.  Dissolved oxygen
             b.  Chemical interactions

       2.  Public Health

       3.  Eutrophication - biomass problems

       4.  Food chain toxicant problems

It was recommended that the Phase II study .be directed toward
three spatial scales:

       1.  Comprehensive Great       All  lakes interconnected
           Lakes scale

       2.  Lake wide scale           Lakes Erie and Ontario  .

       3.  Regional scale            Duluth, Minnesota area,
                                     Southern Lake Michigan,
                                     Green Bay,
                                     Saginaw Bay,
                                     Lake St. Clair.

-------
Pursuant to these recommendations the following specifica-
tions have been prepared for the applicable modeling frame-
works.  The scope of applications for the models recommended
for the various spatial scales is summarized in Table 1.

Figure 3 illustrates the primary input and output variable
associated with recommended modeling frameworks.  The system
diagram for the eutrophication model is presented in Figure
4.  The food chain modeling effort is illustrated in Figures
5, 6, and 7 which diagrammatically depict a ten-compartment
model with no spatial definition, a ten-compartment model
with spatial definition, and a seven-compartment food chain
model.  Interfacing of the eutrophication and food chain
models is illustrated in Figure 8.

Specifications have been prepared in accordance with the
recommended Level 2 Limnological Systems Analysis programs.
The recommended -level of effort and these specifications do
not include certain features which were considered in the
initial project statement of work to be components of the
recommended program.  These components are not consistent
with the actual recommendations developed.  For consistency
with the scope of analysis and level of complexity of the
recommended programs, specifications have not been prepared
for the development of Administrative Models or of input data
handling'systems.  The specifications that follow cover the
modeling applications recommended for Phase II as discussed
above and summarized in Table 1.

-------
AREA AND
SPATIAL SCALE

GR£AT LAKES SCALE
LAKE ERIE
LAKEWIDE SCALE
. DILUTH AREA OF
LAKE SUPERIOR
REGIONAL SCALE
SOUTHERN LAKE
MICHIGAN
REGIONAL SCALE
SAGINAW BAY '
REGIONAL SCALE
GREEN BAY
REGIONAL SCALE
FOOD CHAIN
' ' MODELING
LAKE ONTARIO
LAKE ST. CLAIR
PROBLEM
CATEGORY
•) DISSOLVED OXYGEN
Dl CHEMICAL


















—




•




—



•

•





•




•


—





















•



•

•




EUTROPHICATION

•




•

•

•









Z
X
u
a
o
o
it


















•
•
ANALYSIS
FRAMEWORK




STEADY - STATE









\
TIME
SCALE
a
<
u
u
O
•

















—

_j
4
3
Z
X
<
•
•



•












•
•
SEASONAL

•

•


•

•


•

•

•




>-
_)
X.
»-
z
o
1










•



•

•




>-
_»
M
s

























„












--
PARAMETERS






















•
A
A
A

•

•

...





A
A


A







-

z

•


....
•
•


•

















—













z

•


—
•

•


•


























-


„
«
1

•




•

•


•









& MATERIAL
>


•


-
•

•


•













-















»•
J
O
<


PHOSPHORUS ( INORGANIC )





•













-










































-










•



•

•




•
G

•



•

•


•

-



-


•




•

•


•
































•




















•*
a
z
•




















<






A

A


A









LIGHT TRANSPARENCY

•




•

•


•









a











































Z

































































Z
O
4
e
M
«
a
































•



•

•




>•
»
<
1
O
a
tc
o
I]
0
u










•



•

•
























•
•




















•
•












































LCOEKO,
A




          TABLE  I
SUMMARY OF RECOMMENDED MODELS

-------
INPUT  VARIABLES
LAKE LEVEL
TRIBUTARY INFLOW

WASTE DISCHARGES
MUNICIPAL
AGRICULTURAL

INDUSTRIAL
                                                              r",
                                    MODELING

                                   FRAMEWORK
OUTPUT VARIABIES

DISSOLVED OXYGEN
TDS -CHLORIDES
BACTERIA
ZOOPLANKTON BIOMASS
PHYTOPLANKTON BIOMASS
NITROGEN
PHOSPHORUS
CADMIUM  *
                                   FIGURE  3
                        PPIMAPV INPUT- OUTPUT  VARIABIES

-------
UPPER TRCPHIC
  LEVEL »2
   CARBON
UPPER TROPHIC
   LEVEL * I
   CARBON
  CARNIVOROUS
 zoo PLANKTON
   CARBOM
  HERBIVOROUS
  ZOOPLANKTGN
   CARBON
 PHYTOPLANKTON
  CHLOROPHYLL
 BIOLOGICAL
SUB-MODEL
                                                   NITROGEN CYCLE
                                                      AMMONIA
                                                      M1ROGFN
                                                 PHOSPHORUS  CYCLE
  ORGANIC
PHOSPHORUS
 AVAILABLE
PHOSPHORUS
      CHEMICAL- BIOCHEMICAL
        .  SUB-MODEL
                               NITRATE
                               NITROGEN
                               FIGURE   4
               SYSTEM  DIAGRAM-LAKE  I  MODEL

-------
       FIGURE 5
A TEN COMPARTMENT MODEL

-------

AGRICULTURAL
RUNOFF
0

\

SHORE BIRDS
©


I





LAKE LEVEL
i

V
NUTRIENT EPILIMNION
              H-
                               PHYTO.
NUTRIENT HYPOLIMNION      I      HYPOLIMNION    I
	     PHYTO.
 I      EPILIMNION    |
                H-
 i©
                         NUTRIENT SEDIMENT
                           FIGURE   6
                   A TEN  COMPARTMENT MODEL
                    WITH  SPATIAL DEFINITION

-------
    FIGURE  7
A FOOD  CHAIN MODEL
         12

-------
r









cr D:
UJ UJ
5 £



.




1












SUB-SYSTEM
INTERFACE

W AT Eli.

1
0 •
L


1



1




, — S.
— ( K y*"

EUTROPHICATION






~
PHYTO
GROWTH
RATE

PMYTOPLAfJ/,1 OH

2
0

fie\

\
v_y


MODEL






. 	
~ \K /-+•
,


-/TV-
vy



i
Z 00 P L
GROWTH
RATE

LEGEND:
0- *
-KINETIC REACTION RATES

ZOOPLANKTON FISH . '" LAKE BIRDS
(CADMIUM) 	 . (CADMIUM) ' 	 ^ (CADMIUM)
,3 . — / K \+- • 4 • . — T K V»- 5 , •
(*) : •'•" 0 ;; • 00







^^ ^^
/ ., \
L
                                 FIGURE  8
               INTERFACING  OF EUTROPHICATION AND FOOD CHAIN MODELS

-------
                           SECTION II

             GREAT LAKES MODELING COMPUTER PROGRAMS


                         Program Purpose  ' '
These computer programs are to be used to do the computa-
tions involved in the LSA m6del applications discussed
previously.  Two general types of problem formulations
are involved.                            .    ;

     1.  A set of linearly coupled, linear elliptic
         partial differential equations corresponding
         to the temporal steady state spatial distri-
         bution of water quality variables.

     2.  A set of nonlinearly coupled, parabolic partial
         differential equations corresponding to the
         time varying spatial distribution of water quality
         variables.                     ...
                        Spatial Settings
The computer programs shall be quite general .in terms of the
spatial configurations that can be accommodated.  The segments
(or cells) of the finite difference scheme can be arranged
into configurations that are representative of:

     1.  The five lakes - each lake represented as a few
         interacting segments and linked toxeach other
         via the tributary flow.  The TDS model demonstra-
         tion (1) is an example of this type of application.

     2.  One lake or a basin within the lake;  fairly detailed
         segmentation in both horizontal dimensions; two
         layers in the vertical dimension representing the
         epilimnion and hypolimnion.

     3.  One lake or a basin within the lake;  full three-
         dimensional segmentation with varying degrees of
         spatial resolution in different regions of the lake.
   Hydroscience, Inc.," Limnological Systems Analysis of the
     Great Lakes - Phase I,"pp. 318-327,  (1973).
                               14

-------
                    Program Design Philosophy
This program is intended to be used by persons with a
reasonable background in computer based modeling.  There-
fore, the program should not be designed in such-a way
that its operation is invisible to the user (as would be,
for example, compilers) but rather as a generally applicable
water quality modeling program with enough flexibility and
features so that it can be applied to the different problem
settings with comparable ease.  The major difficulties to
be overcome are the large size of the problems to be computed,
in terms of the input required to specify the geometry,
transport, kinetics, boundary conditions, and external
sources; the storage required for computations; and the
execution time required for solution.


                    Required Program Features
The following is a series of features which appear to be
necessary if the program is to be sufficiently general and
useful:

     1.  Constant or spatially and/or temporally variable
         transport coefficients  (dispersion coefficients
         and velocities) parameters, reaction kinetic
         coefficients, exogenous forcing functions, and
         boundary conditions.

     2.  Easily modified non-linear reaction kinetics to
         accommodate newly developed formulations.

     3.  Program code transportability.  Coding in standard
         FORTRAN.  Highly molecular design.,' Logically
         straight-forward construction so that other pro-
         grammers can understand the code.  Detailed
         documentation in the code.  Execution of both
         IBM 360/370 and CDC 6000/7000 series computers.

     4.  General capabilities to print and/or plot the
         results on a standard line printed in the form of
         (a) tabular columns,  (b) temporal plots at a
         segment,  (c) printed arrays with boundary outlines,
         with the variables in positions corresponding to
         their locations in space for either horizontal or
         vertical planes of the three dimensional computa-
         tions and,  (d) contour plots of.the results
         generated in  (c).
                                15

-------
     5.  Non-uniform grids, perhaps as sets of uniform
         blocks,  to accommodate the requirements of
       .,  varying  spatial resolution.
                   Desirable Program Features
During the design phase of the program, the following
features are to be evaluated and included, if feasible:

     1. . Easily specified spatial configurations using the
         known geometries of the lakes and reference coor-
         dinates of latitude and longitude.

     2.  Specifications of transport parameters at some
         spatial locations with the complete detailed
         specifications for each segment done by the
         p'rogram using an interpolation scheme, which,
         .in the case of velocities, conserves fluid.

     3.  Efficient computations at full arithmetic speed of •
         the machine even for large problems that exceed
       .  memory size.

     4.  Capabilities of generating graphics output stills
         and/or motion picutres, in such a way as to be as
       .  machine and installation non-specific as possible.
          i
     5.  Graphical comparison of observed data and computed
         results for use in verification analysis.

     6.  Ability to compute and display various additional
         diagnostic results such as the reaction kinetic
         terms of the equations, the transport fluxes,
         linear combinations of variables, -etc.

     7.  Ability to modify the differencing scheme used for
         the calculation in order to investigate the sensiti-
         vity of the solution to the scheme used. .

     8.  Ability to investigate solution sensitivity to
         parameter variations in a convenient way.

     9.  Extensive error trapping capability.  Overflows,
         underflows, indeterminate quantities produce  output
         which aids in the debugging of the program, in
         addition to producing output  computed up to the
         error.
                               16

-------
                Phases of the Development Project
After an expenditure of 10% to 20% of project funds, the
contractor shall present a preliminary program design which
addresses the issues raised in these specifications as well
as those which occur during the design process.  Since this
program is not unlike other large scale scientific computa-
tions, e.g., numerical weather•codes, hydrodynamic calcula-
tions for the Great Lakes and elsewhere, the contractor shall
review these and any other relevant programs in order to
examine how they operate, what the experience has been during
their use, what the major difficulties have been, and how
this experience applies to the Great Lakes computer program.
The laboratories at which such computations are performed,
such as the National Center for Atmospheric Research, the
Geophysical Fluid Dynamics Laboratory, and the Linermore
Radiation Laboratory, should be contacted and the knowledge-
able personnel consulted for relevant experience.


                    Technical Specifications


Two sets of equations can be solved  for the various geometries.
They are: .    '           •   .    ••,-.'.

     1.  Steady state, linearly coupled elliptic partial
         differential equations of the form:



                    (-E. |S1 +.V.cj) =  f   K.vck + Wj
                      1 9xi    r   •   ;k=l   3k


         in the three coordinate  variables .,xj , x2 / x3 , with
         dispersion coefficients  E., velocity  field V., for
         the water quality, variables c1., c2,...CN, interacting
         linearly via the reaction kinetic constants K., .  With
         sources W  . , boundary conditions of the  form:-1


                cj       .     .....
          " Ei  x~7 + V.CD =  JD  + K3. (C^  - C)
                i     i               s

         are to be implemented at the boundaries.
                                17

-------
     2.   Time variable parabolic equations of the form:

                2             .

          2- +     - — (-E.  |- -l- V.cj)  = Rj (c)  + Wj
           r;   L=l   Xi    1 Xi     1 .

         with transport similar  to the steady state
         equations and non-linear kinetic coupling
         functions R-' (c) .          :'.'•••

         The contractor shall investigate the relevant
         finite difference schemes and solution methods
         for each equation, with reference to the para-
         meter ranges for the Great Lakes applications.
         Stability criteria shall be evaluated in order
         to choose the scheme that is a balance between
         computational cost and programming ease.  In
         particular, the experience of users of the
         various schemes investigated shall be considered,
         as well as a thorough literature review and
         evaluation.  Among the issues to be addressed
         are: implicit and/or explicit schemes, accuracy,
         stability, positivity,  and conservation pro-
         perties  (mass, quadratic).    ,  ,


                   .  Contractor Requirements


The contractor shall design, code, debug, validate, and
document the Great Lakes modeling computer program.

Validation shall consist of reproducing selected calculations
for the Lake Erie Western Basin coliform model, the five lake
TDS model, and the eutrophication model as .'described in the
LSA feasibility study.

Documentation shall consist of two manuals: a Users Manual
intended for the informed user,  and a Systems Manual which
describes the program in sufficient detail for a programmer to
understand and implement it at an installation.  In addition,
a report shall be produced which documents the result of the
investigations during- the design phase, with regard to the
numerical and programming issues raised.

A schedule with landmark requirements shall be established to
aid in keeping the project on budget and  schedule.
                               18

-------
                           SECTION III

                  GREAT LAKES SCALE APPLICATION


                      Scope of Application   .;


The scope of this application encompasses the five Great
Lakes.  The basin is treated conceptually as a number of
lakes in series or in parallel as appropriate.  The model-
ing effort will employ the computer programs developed under
other tasks of this project.  A total of ten to twenty
horizontal segments for the entire five Great .Lakes will be
employed.  Spatial detail will, therefore, not be considered
as a prime objective of this modeling effort.  In addition
to the horizontal segments, vertical segmentation of the
Lake systems, as appropriate, will consider., an additional
ten to fifteen vertical segments.

Two specific problem categories will be examined in this
modeling framework.  The first is consideration of chemical
water quality on a time variable basis with the use of a
time scale of year-to-year variations.  The ..consequence of
activities on the Great Lakes over decades will be analyzed.
The specific variables to be studied in this analysis are
total dissolved solids and chlorides.

A second problem category to be examined deals with eutro-
phication within the five Great Lakes.  Again, the overall
planning horizon is on the order of decades.  .The model to
be employed will consider time variable, non-linear kinetics
with seasonal and year-to-year effects.  The specific
variables to be examined in the modeling effort are nitrogen
series (organic nitrogen, organic nitrogen  associated with
living material, ammonia nitrogen, nitrite,' nitrate nitrogen);
phosphorus series  (inorganic phosphorus, organic phosphorus,
and phosphorus associated with living organic material);
phytoplankton  (as measured by chlorophyll  'a'); and zooplank-
ton  (as measured by zooplankton carbon)..  The 'objective of
this application is to examine water quality in the Great
Lakes resulting from long term projections  for municipal and
industrial development.                 '    •. ;
                                19

-------
                        Study Input Data
The geomorphology of each of the five Great Lakes will
have to be inputted to the modeling framework.  It will
be necessary to identify the volume, depths, and thermo-
cline locations for each of the spatial segments to be
considered in the modeling effort.  In addition, an
averaged hydrological balance for the entire Great Lakes
must be developed, and should include inflows, outflows,
precipitation, evaporation, and variations in lake'volume.

Input from man's activities and natural phenomena must be
considered within the modeling framework.  Specifically,
the chemical and biological input from point and distributed
sources on a lake-by-lake and segment-by-segment basis are
required.

Water quality data on the total dissolved solids, chlorides,
nitrogen, phosphorus, phytoplankton, and zooplankton under.
observed historical conditions will have to be collected and
analyzed over the time scale of the modeling analysis.  These
data should cover the greatest possible historical span.  It
is not necessary to have a complete data set in all variables
to provide a basis for partial verification of the model.
The contractor is to collect all of the available information
and to examine it- for use in model verification with due
consideration of the large time and space scale being
considered in the model and the objective of making long-
term projections of water quality in the Great Lakes.
                       Modeling Framework
The chemical water quality study, considering total dissolved
solids and chlorides, will employ a time variable analysis
with conservative variables.  The eutrophication study will
employ a non-linear time variable analysis as described in,
"Limnological Systems Analysis of the Great Lakes - Phase I,"
prepared by Hydroscience, Inc. i-U
   Hydroscience, Inc. ," Limnoloqical Systems Analysis of the
     Great Lakes - Phase I," (1973).
                                20

-------
The specific modeling framework and kinetic interactions to
be included in the Great Lakes scale eutrophication modeling
are those indicated by the kinetic diagram shown below:

Organic
Nitrogen

Organic
Phosphorus
'• '»



Inorganic
Nitrogen

Inorganic
Phosphorus'



Plankton
Chlorophyll 'a'






Zooplankton
Carbon


Sources and sinks of material to be considered in the Great
Lakes scale analysis should include not only those point
and distributed sources from the shoreline but should also
include atmospheric inputs, settling, and leaching phenomena
as appropriate.


                   Verification Analysis


The model output for both the conservative chemical and
eutrophication models should be compared to historical data.
It will not be necessary to have a consistent and continuous
set of historical data in time to obtain the desired comparisons.

Because of the large time scale of the modeling effort, it is
desirable' to include an additional step in the verification
analysis.  One of the data sets available in both the total
dissolved solids and eutrophication analysis should be withheld
for an independent check on the verified model, following cali-
bration of that model.

The sensitivity analysis for the Great Lakes, scale models should
include analysis of the sensitivity of model output to input
load, to Great Lakes hydrology and coefficients, and to inter-
relationships employed in the analysis.


                   Application of Model


A series of projections will be made under three levels of
future growth of the Great Lakes area.  The specific informa-
tion on populations and economic levels will be provided by
the GLEPS study of the Great Lakes Basin Commission.  The
                              21

-------
data on population changes and industrial growth, together
with alternations in land use, will be employed by the
contractor to develop sequences of inputs to the lakes over
time.  The effect of these sequences of .inputs will be
examined.  It will be necessary to closely coordinate these
alternatives with the Great Lakes Basin Commission.
                          Final Report
The contractor will prepare a final project report which
will describe the modeling effort, the data base employed
for model verification, the system inputs/ and the results
of the verification and sensitivity analyses.  The final
report will present the projected water .quality resulting
from the-three levels of future development in the basin.
                                22

-------
                            SECTION IV

                   LAKEWIDE SCALE - LAKE ERIE


                      Scope of Application


The study area will encompass all of Lake Erie.  Horizontal
and vertical model segmentation'will be structured so that
each sub-model is upward compatible with the largest model,
i.e., segment geometry will be such that segments do not
overlap.  The ultimate spatial resolution will be limited by
computer time and memory constraints.  It is anticipated
that a variable grid including both horizontal and vertical
segments will be employed with total segmentation ranging
between fifty and two hundred segments.

The computation framework which will be developed in other
'tasks will consider time variable and non-linear kinetics on
a seasonal time scale.  The major thrust of the calculation
procedures will be the evaluation of eutrophication as
measured by phytoplankton biomass as a function of nutrients
inputted either as point sources or as distributed sources
(including the effects of nutrients returned and removed by
phenomen associated with the benthos).  .The major water quality
effect to be considered is the anoxic hypolimnion and its
relation to primary production with respect to dissolved
oxygen.

The variables to be examined will include the nitrogen series
(organic nitrogen, organic nitrogen associated with living
material, ammonia nitrogen, nitrite, nitrate nitrogen), the
phosphorus series  (inorganic phosphorus, organic phosphorus,
and phosphorus associated.with living organic material, phyto-
plankton  (as measured by chlorophyll 'a'),,and zooplankton
(as measured by zooplankton carbon).  In addition, the organic
carbon cycle, dissolved .oxygen,  and silica are to be included,
as necessary.  Finally, chemical effects on the availability
of inorganic phosphorus may also be a significant factor.


                        Study  Input Data  .


The contractor, will review and gather data available on system
geomorphology.  Specifically,  the contractor will gather
information on the geometry of Lake Erie, including the depths,
volumes, and  location of thermoclines.  The contractor will,
in addition,  develop a hydrologic balance for  the region,
                               23

-------
gathering information as required on inflows,  outflows,
precipitation over the lake, evaporation and variations  in
lake volume.  An advective-dispersive model is to be
employed; therefore, the contractor must also evaluate
data on lake velocities and dispersion coefficients.  The
major transport variables will have to be evaluated with a
combination of calculated circulation patterns and observed
data.                            .   .

An accurate and complete quantitative assessment of all
major mass inputs of the dependent variables 'is required.
This is critical to the success of the modeling effort,  both
in verification and projections.  Estimates, therefore,  must
be generated of mass inputs contributed by man's activities
from point sources, such as industrial and municipal wastes,
as well as those associated with runoff from urban, suburban,
and undeveloped land areas.  Assessments must be developed
of the 'input of nutrients associated with bottom phenomena,
particularly with anoxic conditions in the hypolimnion of
central Lake Erie.  Finally, information on applicable reaction
and transformation rates must be obtaind for the Great Lakes
literature base.  The contractor is required to examine the
data base and gather information on appropriate water quality
variables employed in the modeling effort.  The model will
require verification at different spatial locations over the
period of several years.  It will, therefore, be necessary to
select from the data base,  information on a nitrogen  series,
phosphorus series, phytoplankton, zooplanktori, organic carbon,
dissolved oxygen, and silica.   The data base should cover at
least two annual cycles and need not extend beyong  four annual
cycles.  Data sets may consist  of information pieced  together
from a number of different  surveys and observations carried
out by interested agencies  and  groups on the Great  Lakes.

                    Modeling Frameworks  ••.•''


The eutrophication biomass  formulation will be essentially
that presented  in the demonstration model(D, suitably
enlarged and refined as required.  The specific extensions
that are to be  considered in this analysis  are:

      1.  effects of anoxic hypolimnion on  nutrient
          recycle,                            ;
      2.  the possibility of nitrogen fixation and
          denitrification,
    Hydroscience,  Inc.,  "Limnological  Systems  Analysis  of  the
      Great  Lakes  -  Phase  I,"  pp.  337-354,  (1973).


                             .24

-------
     3.   the vertical settling and decomposition of
         plankton and detritus,            :

     4.   the effect of sediment processes on the
         mineralization and recycle of nutrients,

     5.   an improved zooplankton formulation,

     6.   an explicit formulation of the significant
         inorganic chemical processes and their
         effect on silica and phosphorus,

     7.   dissolved oxygen analysis including the
         oxidation of organic carbon and oxygen
         production and respiration by plankton,

     8.   inclusion of dissolved carbon dioxide in the
         modeling framework coupled with oxidation of
         organic carbon and carbon dioxide transfer
         between the liquid and atmospheric phases.
                      Verification Analysis
Model verification of the transport regime is required and
should employ conservative ions as tracers and/or temperature
gradients.  In addition, verification and/or temperature
gradients of composite variables that are primarily affected
by the boundaries of the model can be made and would include
total nitrogen and total phosphorus concentrations.

At least two years of data for all variables, for which data
exists, should be examined.  In addition, at least one year
and, if possible, an additional year of data should be
employed in the verification.  As an example of this kind of
independent check, sparse data is available on plankton
counts and nutrient levels in 1930 in the western basis of
Lake Erie.  Comparable data from the forties, fifties, and
sixties could also be used for other portions of Lake Erie.
Finally, all kinetic constants and parameters should be in
the range of reported literature values.

Additional model output should be obtained to determine the
sensitivity of the system to variations in model coefficients
and parameters.                                         ..
                                25

-------
                      Application of Model
The contractor shall apply the eutrophication and
dissolved oxygen models to predict the effects of man's
activities and changes in natural inputs.  The GLEPS
program of the Great Lakes Basin Commission's studies
will provide data on changes in waste loads resulting
from population growth, economic developments, altered
land use. patterns, and environmental control procedures.
These data on input waste loads will be. employed in the
modeling effort to predict the anticipated level of
phytoplankton biomass and dissolved oxygen in the study
area. . Specifically, the model will be employed to develop
projections for at least three levels of growth and future
activities.
                          Final Report
The contractor will prepare a final project report which
will describe the modeling effort, the data base employed,
waste loads and inputs to the system, the results of the
verification, and sensitivity analysis.  The report will
also present the specific projected results for the three
levels .of future development in terms of anticipated
phytoplankton biomass and dissolved oxygen responses within
Lake Erie.
                               26

-------
                            SECTION V

    DULUTH AREA OF LAKE SUPERIOR - REGIONAL SCALE APPLICATION
             •f-

                      Scope of Application


The spatial extent of this, application includes the Duluth
area of Lake Superior and extends approximately twenty miles
into Lake Superior proper.  The modeling effort should
utilize from twenty to forty spatial segments.

The application will consider the specific problem category
of eutrophication.  A non-linear time variable computer
program developed under other tasks in this project will be
used on a seasonal time scale.  The variables to be included
in the modeling effort are the nitrogen series (organic
nitrogen associated with living material, ammonia nitrogen,
nitrite, nitrate nitrogen), the phosphorus series  (inorganic
phosphorus, organic phosphorus, and phosphorus associated
with living organic material), phytoplankton  (as measured by
chlorophyll  'a'), zooplankton (as measured by zooplankton
carbon).  In addition, the available data will be examined
and appropriately evaluated.to determine if silica, oxygen
and organic carbon cycle should, be included in the analysis.
The contractor is to employ data analysis and simple calcula-
tion procedures to establish the availability of organic
carbon and silica as a potential limiting nutrient for
plankton populations under existing conditions and under
future projected conditions..


                        .Study Input Data


The contractor will review data on the system geomorphology.
Specifically, the contractor will gather information on the
geometry of Lake Superior in the Duluth area.  This will
include the depths, lake volumes, and location of  thermoclines.

The contractor will develop a hydrologic balance for the
region of Lake Superior, gathering, information as  required
on inflows, outflows, precipitation over the  lake, evaporation,
and variations in lake volume.  An advective-dispersive model
will be employed.  Therefore, the contractor must  evaluate
data on lake velocities and dispersion coefficients in the
Duluth area of Lake Superior.  The hydrologic balance will be
developed for a period when observed water quality data for
the variables being modeled are available.
                               27

-------
The contractor is required to examine the observed data
base and gather information on water quality variables
employed in the modeling effort.  The model will require
verification at different spatial locations in the Duluth
area of Lake Superior over time.  It will/ therefore, be
necessary to review the existing data base and to select
from that data base water quality information on the
nitrogen series, phosphorus series, phytoplankton, and
zooplankton for one to two annual cycles. ; Information
need not be available continuously over an annual cycle
but can be obtained from periodic water quality survey
information collected over a number of years.  This relaxed
data requirement is associated with the fact that growth in
Duluth in recent years, has not been considered rapid in
relationship to the size of the area to be modeled in Lake
Superior.  This phenomenon makes it possible to piece together
data sets from a number of different years of observations.

The final item of input data required consists of the
residual inputs from point and distributed sources of the
modeling effort.  Residual inputs from natural distributed
sources may be obtained from existing data in the Lake
Superior area .or may be extrapolated from literature values
for comparable land use areas.  Modification of the literature
values may be necessary, depending, on the', specifics of the
study area in which the data was collected.  The chemical and
biochemical data from the point sources in the .area should
consider direct discharges to Lake Superior and tributaries
as point source inputs.  Checks should beimade of tributary
loads to insure that the observed data on.the tributaries
represent  the sum of contributing populations(with considera-
tion to treatment of municipal and industrial discharges), as
well as land runoff.                     ' :
                        Modeling  Framework
The contractor will employ  the  time variable non-linear
programs developed by other tasks  in  this•study.  The
modeling framework to be used by the  contractor  is  an updated
computer program which  employs  feedback  and non-linear
kinetic relationships.  The major  output variables  and
modeling variables are  as identified  previously  and are,
specifically, the nitrogen  series  (organic nitrogen, organic
nitrogen associated with living material,• ammonia nitrogen,
nitrite, nitrate nitrogen); the phosphorus series  (inorganic
phosphorus,  organic phosphorus  and phosphorus  associated
with  living  organic material);  phytoplankton  (as measured  by
chlorophyll  'a'); and zooplankton  (as measured by zooplankton
carbon).
                                28

-------
The contractor shall consider as input to the modeling
effort, estimates of the distributed and point loads from
natural weathering, erosion, and man's activities.   In
addition, the contractor shall consider estimates of
settling and leaching of material from the bottom of Lake
Superior.  The contractor shall also determine the signi-
ficance of atmospheric sources of material and, if
appropriate, include estimates of these in the modeling
effort.
                      Verification Analysis
Model verification shall consist of a comparison of computed
model output over an annual cycle against observed water
quality data.  The comparison shall be developed for all of
the model variables simultaneously, with'a consistent set
of parameters and coefficients.

In addition, model output shall be obtained to determine the
sensitivity of the system.
                      Application of Model


The contractor shall apply the eutrophication model to predict
the effect of man's activities and changes in natural inputs
to the system.  Specifically, the GLEPS program from the
Great Lakes Basin Commission study will provide data on
increases in population and economic development of the area
and changes in land use patterns within the study area.
These data on input changes will be employed in the modeling
effort to predict the level of nitrogen, phosphorus, phyto-
plankton, and zooplankton.  Specifically, the model will be
employed to develop projections for at least three levels of
growth and range of future activities.
                          Final Report
The contractor will prepare a final project report which will
describe the modeling effort, the data base, loads, and the
results of the verification and sensitivity analysis.  The
final report will present the specific projected results for
the three levels of future development in the Duluth area
of Lake Superior.
                               29

-------
                           SECTION VI

        SOUTHERN LAKE MICHIGAN REGIONAL SCALE APPLICATION


                      Scope of Application
The area of concern includes .Southern Lake Michigan and
extends approximately one hundred twenty-five miles into
Lake Michigan, in a northerly  direction from Gary,
Indiana.  The modeling effort will employ the computer
programs developed under other tasks of this project and
will consider from seventy-five to one hundred fifty
spatial segments within Lake Michigan.

There are two specific problems which are to be' addressed
by the contractor.  The first problem will deal with the
eutrophication phenomenon in Southern Lake Michigan and
will employ the time variable non-linear models for
eutrophication.  The analysis will consider seasonal
phenomena.  The variables to be included in the eutrophi-
cation model will consist of the nitrogen series (organic
nitrogen, organic nitrogen associated with living materials,
ammonia nitrogen, nitrite, nitrate nitrogen), the phosphorus
series  (inorganic phosphorus, organic phosphorus, and phos-
phorus associated with living organic material), phytoplankton
(as measured by chlorophyll  'a'), and zooplankton  (as
measured by zooplankton carbon).  In addition, the available
data will be examined and appropriately evaluated to determine
whether silica, oxygen, and organic carbon cycle should be
included in the analysis.  The second problem category to
be examined by the contractor will deal with bacterial pollu-
tion of the nearshore region of Lake Michigan.  Specifically,
the spatial extent of this second modeling effort will be on
the order of a ten mile band around the shoreline of Lake
Michigan.  This modeling effort will employ linear models and
will consider steady state conditions on approximately a
monthly time scale.  The specific variables to be examined
in this second problem category will be total and fecal
coliform bacteria with consideration given to die-away and
aftergrowth.
                               30

-------
                        Study Input Data
The  contractor will review  and gather data on the system
geomorphology.  Specifically, the contractor will gather
information  on the geometry of Southern Lake Michigan
including  the depths,  lake  volumes,  and location of
thermoclines.

The  contractor will, in  addition, develop a hydrologic
balance  for  the region,  gathering information as required
on inflows,  outflows,  precipitation  over the lake, eva-
poration,  and variations in lake volume.  An advective-
dispersive model  is to be employed for the calculation base
.in the eutrophication  modeling effort.  Therefore, the
contractor must also evaluate data on lake velocities and
dispersion coefficients. The hydrologic balance will be
developed  for a period when observed water quality data
for  the  variables being  modeled are  available.

The  contractor is required  to examine the observed data
base and gather information on water quality variables in
the  modeling effort.   The model will require verification
at different spatial locations over  time.  It will,  there-
fore, be necessary to  review the existing data  base  and to
select from  that  data  base, water quality information on the
nitrogen series,  phosphorus series,  phytoplankton, and
zpoplankton  for one to two  annual cycles.  Information need
not  be available  continuously over an annual cycle,  but can
be obtained  from  periodic irregular  water quality survey
data collected in the  area  over a number of years.   Data
sets may consist  of  information pieced together from a
number of  different years of observation.

The  final  item of study  input data required,, consists of the
residual;inputs from point  and distributed sources of the
various  chemical, biochemical, and bacterial variables
considered in the modeling  efforts.  Residual  inputs from
natural  distributed  sources may be obtained from existing
data in  the  region or  may be extrapolated from  literature
values for comparable  land  use areas throughout the  contin-
ental United States.   Modification of the literature values
may  be necessary, depending on the specifics of the  study
area in  which the data was  collected.  The chemical, bio-
chemical,  and bacteriological data from  the point sources  in
the  area should consider direct discharges to  Southern Lake
Michigan and tributaries as point source inputs.  Checks
should be  made of tributary loads to insure that the inputs
                                31

-------
represent the sum of contributing populations (with considera-
tion to treatment of municipal and industrial discharges),  as
well as the land runoff.  The information on waste inputs for
the eutrophication analysis is required, preferably, on a
time variable basis.  If this information is not available
for municipal and industrial inputs, it is adequate to
use an average municipal and industrial input to the system
which is constant over time.  The natural load'variations
should be calculated in a time variable fashion, if possible.
Constant monthly average inputs are adequate for the coliform
analysis.


                    Modeling Framework


As indicated previously, the contractor will be required to
employ the time variable, non-linear eutrophication modeling
framework developed elsewhere in this study.  The major
output variables and modeling variables are the nitrogen series
(organic nitrogen, organic nitrogen associated with living
material, ammonia nitrogen, nitrite, nitrate nitrogen); the
phosphorus series  (inorganic phosphorus, organic phosphorus,
and phosphorus associated with living organic material);
plankton  (as measured by chlorophyll 'a'); and zooplankton
(as measured by zooplankton carbon). ..  .

Specifically, the modeling framework to be used by the
contractor is an updated computer program which employs the
feedback and non-linear kinetic relationships.

The contractor shall consider as input.to the modeling effort,
estimates of the distributed and point loads from natural
weathering and man's activities and, in addition, shall
consider estimates of settling and  leaching of .material from
the bottom of Lake Michigan, and shall determine the signi-
ficance of atmospheric sources of material,xand if appropriate,
include estimates of those in the modeling effort.

The model that will be employed for the examination of
bacterial pollution in the nearshore regions of Lake Michigan
will be .the linear steady state modeling effort developed by
other tasks.  The model variables will be the bacterial input
of both total'and fecal coliform from direct point sources,
both industrial and municipal.  In  addition, the contractor
will consider the runoff from stormwater overflow in urban,
                              32

-------
suburban, and rural areas, and finally will consider the
contribution of fecal and total coliform bacteria from
natural phenomena and those entering the system from
tributaries to Southern Lake Michigan.
                      Verification Analysis
Model verification shall consist of a comparison of computed
model output over an annual,cycle against observed water .
quality data for the eutrophication analysis.  Seasonal
steady state comparison shall be made for the coliform
analysis.  The comparisons shall be developed for all of
the model variables simultaneously with consistent sets of
parameters and coefficients.

In addition, model output shall be obtained to determine the
sensitivity of the system.
                      Application of Model
The contractor shall employ the eutrophication and coliform
models to predict the effect of man's activities and changes
in natural inputs to the system.  Specifically, the GLEPS
program from the Great Lakes Basin Commission study will
provide data on waste loads resulting from increases in
population and economic development of the area and changes
in land use patterns.  These data on input waste load changes
will be employed in the modeling effort to predict the
anticipated level of nitrogen, phosphorus, phytoplankton,
zooplankton and coliform in the area under study.  Specific-
ally, the model will be employed to develop projections for
at least three levels of growth of future activities.


                          Final Report


The contractor will prepare a final project report which will
describe in detail the modeling effort, the data base employed
for model verification, the loads employed, and the results of
the verification and sensitivity analysis.  The final report
will present the specific projected results for the three
levels of future development in Southern Lake Michigan.
                               23

-------
                           SECTION VII

             SAGINAW BAY REGIONAL SCALE. APPLICATION


                      Scope of Application
The study area will encompass Saginaw Bay and extend
approximately sixty miles in a northeasterly  direction
from Bay City, Michigan, toward Lake Huron.  The modeling
effort will make use of computer programs developed under
other tasks of this project.  The analysis will employ
the linear steady state program framework.  A variable
spatial grid will be utilized with finer segmentation in
the vicinity of major waste inputs and expanded grid
sizes as Lake Huron is approached.  It is anticipated
that on the order of two hundred spatial segments will be
employed in the modeling effort.

Two specific water quality problems will be examined by
the contractor.  The water quality variables associated
with these problems are dissolved oxygen and coliform
under seasonal steady state conditions.

The modeling effort will consider oxidation of carbonaceous
and nitrogenous material (sequential oxidation of ammonia to
nitrate shall be considered), algal oxygen production and
respiration, oxygen utilization by bottom deposits, release
of materials into the water column by bottom deposits,
reaeration, aftergrowth, and die-away of total and fecal
coliform bacteria.  The modeling effort will consider steady
state summer water quality on approximately a monthly time
scale and will employ linear reactions and feedbacks as
appropriate.
                        Study Input Data
The contractor will review and gather data available on the
system geomorphology.  Specifically, the contractor will
gather information on the geometry of Saginaw Bay, including
the depths, volumes, and location of thermoclines.

The contractor will, in addition, develop a hydrologic balance
for the region, gathering information as required on inflows,
outflows, precipitation, evaporation, and variations in volume.
                                34

-------
An advective-dispersive model will be employed.   Therefore,
the contractor must also evaluate data on lake velocities
and dispersion coefficients.  The hydrologic balance must
be developed for a period when observed water quality data
for the variables being modeled are available.

The contractor is required to examine the data base and
gather information on water quality variables employed in
the modeling effort.  The model will require verification
at different spatial locations.  It will, therefore, be
necessary to review the existing data base and to select
from that data base information on the temperature, nitrogen
series, BOD, dissolved oxygen, chlorophyll, and coliform
during at least one and not more than three periods.  Data
sets may consist of information pieced together from a
number of different years of observations provided that
temperature, flow, and input are similar.
                       Modeling Framework
As indicated previously, the contractor will be required to
employ a linear steady state program as part of the modeling
effort.  The variables to be considered as output from the
modeling effort will be the dissolved oxygen and coliform
bacterial profiles in Saginaw Bay.  The phenomena to be
included in the dissolved oxygen model are oxidation of
carbonaceous and nitrogenous material  (ammonia conversion
to nitrite and nitrate should be considered), algal oxygen
production and respiration, bottom oxygen utilization and
release of material into the overlying water column from the
bottom, and atmospheric reaeration.  The coliform model will
consider sources of coliform bacteria aftergrowth and the
subsequent die-away of the coliform organisms in the bay.
                      Verification Analysis
Model verification for dissolved oxygen and coliform dist-
ributions shall consist of a comparison of computed model
output under steady state conditions and observed water
quality data in Saginaw Bay.  Monthly or seasonally averaged
information should be employed for the comparison.  The
comparisons shall be developed between all of the modeled
variables simultaneously by use of consistent parameters and
coefficients.  The verification analysis shall consider at
least one and no more than three steady state water quality
data sets.  Verification shall consider data sets obtained
under different temperature, waste loading or flow conditions
within Saginaw Bay.
                                35

-------
Additional output shall be obtained to determine the
sensitivity of the system to variation in the model
coefficients.
                      Application of Model
The contractor shall apply the dissolved oxygen and the
coliform models to predict the effects of man's activities
and changes in natural inputs.  The GLEPS program from the
Great Lakes Basin Commission's studies will provide data
on changes in waste loads resulting from population growth,
economic development, and changes in land use patterns.
These data on input waste loads will be employed in the
modeling .effort to predict the anticipated level of
dissolved oxygen and coliform bacteria in the study area.
Specifically, the model will be employed to develop pro-
jections for at least three levels of growth and future
activities.
                          Final Report
The contractor will prepare a final project report which
will•describe the modeling effort, the data base, loads,
and the results of the verification and sensitivity analysis,
The final report will present the specific projected results
for the three levels of future development in the Saginaw
Bay area.
                               36

-------
                           SECTION VIII

              GREEN BAY REGIONAL SCALE APPLICATION


                      Scope of Application


The study area encompasses Green Bay and extends approx-
imately one hundred miles in a northeasterly  direction
from Green Bay, Wisconsin to Lake Michigan.  The modeling
effort will employ computer programs developed under other
tasks of this project.  The analysis will specifically
employ the linear steady state program framework.  A
variable spatial grid will be utilized with finer segmenta-
tion in the vicinity of major waste inputs and expanded
grid sizes as Lake Michigan is approached.  It is anticipated  •
that the area will be segmented into about two hundred spatial
units.

Two specific water quality problems will be examined by the
contractor.  The water quality variables associated with these
are dissolved oxygen concentration and numbers of coliform
organisms under seasonal steady state conditions.  The analysis
will include oxidation of carbonaceous and nitrogenous material
(sequential oxidation of ammonia to nitrate will be considered),
algal oxygen production and respiration, oxygen utilization
by bottom deposits, release of materials into the water column
by bottom deposits, reaeration, aftergrowth, and die-away of
total and fecal coliform bacteria.  The modeling effort will
consider steady state summer water quality on approximately
a monthly time scale and will employ linear reactions and
feedbacks as appropriate.


                        Study Input Data


The contractor will review and gather data on the system       :
geomorphology.  Specifically, the contractor will gather
information on the geometry of Green Bay, including the
depths, volumes, and location of thermoclines.

The contractor will, in addition, develop a hydrologic
balance for the region, gathering information as required on
inflows, outflows, precipitation over the lake, evaporation,
                               37

-------
and variations in volume.  An advective-dispersive model
will be employed.  Therefore, the contractor must also
evaluate data on velocities and dispersion coefficients.
The hydrologic balance must be developed for periods when
observed water quality data for the variables being
modeled are available.

The contractor is required to examine the observed data
base and gather information on water quality variables
employed in the modeling effort.  The verification of the
model at different spatial locations is required.  It will,
therefore, be necessary to review the existing data base
and to select from that data base information on the sources
of both point and distributed man-made and natural sources
of BOD, nutrients, and bacteria.  In addition, water quality
data on the temperature, nitrogen series, BOD, dissolved
oxygen, chlorophyll, and coliform during at least one and
not more than three periods are required.  Data sets may   . . •
consist of information pieced together from a number of.
different years of observations provided that temperature,
flow, and input are similar.


                       Modeling Framework


As indicated previously, the contractor will be required to
employ a linear steady state program as part of this modeling
effort.  The variables to be considered as output from  the
modeling effort will be the dissolved oxygen and coliform
profiles in Green Bay.  The phenomena to be included in the
dissolved oxygen model are oxidation of carbonaceous and
nitrogenous material  (ammonia conversion to nitrite and
nitrate should be considered),  algal oxygen production  and
respiration, bottom oxygen utilization, release of material
into the overlying water column  from the bottom, and atmos-
pheric reaeration.  The coliform model will consider sources
of coliform bacteria, aftergrowth,  and subsequent die-away
of the coliform organisms in. the bay.


                      Verification  Analysis
Model  verification  for  dissolved oxygen  and  coliform dist-
ributions  shall  consist of  a  comparison  of computed model
output under  steady state conditions  and observed water
quality data  in  Green Bay.  Monthly or seasonally averaged
                                38

-------
information should be employed for the comparison.  The
comparisons shall be developed between all of the modeled
variables simultaneously with the use of consistent
parameters and coefficients.  The verification analysis
shall consider at least one and no more than three steady
state water quality data sets.  Verification shall consider
data sets obtained under different temperature, waste
loading, or flow conditions within Green Bay.

Additional output shall be obtained to determine the
sensitivity of the system to variation in the model coe-
fficients.
                      Application of Model
The contractor shall apply the dissolved oxygen and- the
coliform models to predict the effects of man's activities
and changes in natural inputs.  The GLEPS program from the
Great Lakes Basin Commission's studies will provide data on
changes in waste loads resulting from population growth,
economic development, and changes in land use patterns.
These data on input waste loads will be employed in the
modeling effort to predict the anticipated level of dissolved
oxygen and coliform bacteria in the study area.  Specifically,
the model will develop projections for at least three levels
of growth and future activities.


                          Final Report


The contractor will prepare a final project report which
will describe in detail the modeling effort,, the data base
employed for model verification, the loads"employed, and the
results of the verification and sensitivity analysis.  The
final report will present the specific projected results for
the three levels of,future development in Southern Lake
Michigan.
                                39

-------
                          SECTION IX

                       FOOD CHAIN MODELING


                      Scope of Application
The study area will consist of Lake Ontario as an
illustration of Lakewide scale food chain modeling
and Lake St. Clair as an illustration of regional
modeling of this problem.  In both cases, the emphasis
is on a broad scale model to provide a basis for tracking
potential toxic substances through the food web in the
aquatic environment.  The analysis should consider, as
a minimum, two heavy metals, two organic pesticides, and
two other potential toxicants.  At least seven compart-
ments in the food web should be considered as subse-
quently identified.  In addition, the spatial segmenta-
tion should be on the order of one hundred to four
hundred spatial compartments.
                        Study Input Data
The contractor will review and gather data on the systems
geomorphology.  Specifically, the contractor will gather
information on the geometry of Lakes Ontario and St. Clair,
including the depths, volumes, and location of thermoclines.

The contractor will, in addition, develop a hydrologic
balance for the regions, gathering information as required
on inflows, outflows, precipitation over the lake, eva-
poration, and variations in lake volumes.  -An advective-
dispersive model is to be employed.  Therefore, the
contractor must also evaluate data on lake velocities
and dispersion coefficients.  The hydrologic balance must
be developed for periods when observed water quality data
for the variables being modeled are available.

The contractor is required to examine the observed data base
and gather information on the concentration of the two heavy
metals, two pesticides, and two other potential toxicants
in each of the food web compartments to be specified on the
modeling framework.  In addition, the contractor will be
                               40

-------
required to develop estimates of the biomass in each of
the food web compartments so specified.  Data sets and
information on concentrations of toxicants and biomass
estimates may be obtained by piecing together information
from a number of different years of observations.


                       Modeling Framework


The contractor will be required to employ a linear inter-
active compartment program, as previously developed under
other tasks in this study.  Modifications to the computer
software may be necessary to provide flexibility to allow
for easy rearrangement of compartment transport and
dispersion phenomena as a function of the ecological
variable.  Any computer program modifications should be
thoroughly tested for integrety of the software and to
demonstrate veracity.  Program modifications should be
fully documented.

The ecological compartments to be considered should consist
of at least the following:

     1.  water column
     2.  sediments

     3.  plankton  (10-100 microns)

     4.  carnivorous zooplankton  (100-1,000 microns)

     5.  carnivorous zooplankton  (1,000-10,000 microns)

     6.  representative small fish  (1-10 centimeters)
     7.  representative large fish  (10 centimeters-1 meter)
                                           s
   (Note:  Sizes given are provided as only'a general guideline)


The basic modeling framework for each compartment should
include the mass of toxicant/mass of compartment and the
total biomass measured in convenient units  (for example,
carbon) of each of the compartments.  Also, data on point
and distributed sources of each of the toxicants are
required as input to the modeling effort and may be obtained
from measurements or estimates derived from the  literature.
For Lake Ontario, the results of the IFYGL program should be
reviewed in detail and compiled for use.
                                41

-------
Finally, for establishment of the modeling framework,
specific consultation will have to be sought to determine
the appropriate food web so as to delineate the interac-
tions between the various compartments in the system.

The final step required in the analysis will be a litera-
ture survey and review to identify threshold levels of
each of the potential toxicants and the effects such levels
will have on the ecological compartment.
                      Verification Analysis
Verification analysis shall consist, in part, of compar-
isons of computed model output and observed water quality
and biological data.  Specifically, a single tracer in the
food web should be selected and the model applied to Lake
Ontario to determine the magnitude of the linear transfer
coefficients.  With the same tracer, a verification analysis
should be conducted with available data from Lake St. Clair.
This will provide one check on the veracity of the linear
transfer coefficients.  Hydrodynamic transport regimes would
be externally supplied information from existing computer
programs and models.  This verification procedure will then
be repeated for each of the potential toxicants to be studied.
Additional model output shall be obtained to determine the
sensitivity of the system to variations in the model
coefficients and linear transfer parameters.
                    Application of the Model
The contractor shall apply the food chain model to predict
the effect of man's activities on the distribution of toxi-
cants in the two lakes.  Specifically, examination of
increases and decreases in point and distributed sources
of the toxicants should be evaluated.  The GLEPS program
from the Great Lakes Basin Commission studies will provide
data on changes in waste loads resulting from population
growth, economic development, and changes in land use
patterns or legal regulations.  In addition, the contractor
will examine the effect of potential changes in growth rates
of the various ecological compartments on the distribution
and concentration of toxicants.  Specifically, the food chain
modeling will be employed to develop projections for at least
three levels of future activities in the Great Lakes Basin.
                                42

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
  EPA-600/3-77-126C
                             2.
                                                           3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
 Great Lakes Environmental  Planning Using Limnological
 Systems Analysis:  Model Specifications
             5. REPORT DATE
              November 1977 issuing  date
             6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
 Hydroscience, Inc.,  363 Old  Hook Road,  Westwood, New
 Jersey  07675	
                                                           8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 prepared for the Great  Lakes  Basin Commission
                  P.O. Box 999
                  3475 Plymouth Road
                  Ann Arbor, Michigan  48106
                                                           10. PROGRAM ELEMENT NO.
             11. CONTRACT/GRANT NO.
                 DACW-35-71-C0030
12. SPONSORING AGENCY NAME AND ADDRESS
 Environmental Research  Laboratory - Duluth, MN  55804
 Office of Research and  Development
 U.S.  Environmental Protection Agency
 Duluth MN  55804
             13. TYPE OF REPORT AND PERIOD COVERED
                 Final	
             14. SPONSORING AGENCY CODE
                 EPA-600/03
15. SUPPLEMENTARY NOTES
16. ABSTRACT
 The report documents  the  deliberate decision making process  used  by  the Great Lakes
 Basin Commission  in concluding that rational modeling methodologies  could be used
 to evaluate the effect  of different planning alternatives  on the  Great Lakes and
 that planning for specific problems affecting the Great Lakes  system can be technically
 and economically  supported through mathematical modeling and systems analysis.  It
 assesses the technical  and economical feasibility of developing mathematical models
 to assist in making selections from among alternative management  strategies and
 structural solutions  proposed for solving water resource problems of the Great Lakes.
 The study reviews, evaluates  and categorizes present and future water resources
 problems, presently available data, problem-oriented mathematical models and the
 state of models and model synthesis for large lakes.  A demonstration modeling frame-
 work for planning is  developed and applied to western Lake Erie and  the Great Lakes
 system.  The report evaluates four widely ranging alternatives for future modeling
 efforts in the Great  Lakes and recommends the modeling level most feasible to answer
 planning questions on scales  ranging from the Great Lakes  to regional areas.  Also
 discussed is a proposed Commission study which will apply  limnological systems
 analysis to the planning  process.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS
                             COSATI Field/Group
 Limnology, Systems, Mathematical Models,
 Water Resources, Planning,  Hydrology,
 Ecology
 systems analysis, Great
 Lakes,  ecosystems, long
 term planning,
 environmental effects,
 large lakes
08 H
13 B
18. DISTRIBUTION STATEMENT

 Release unlimited
19. SECURITY CLASS (This Report}
   Unclassified
                                                                         21. NO. OF PAGES
                                              20. SECURITY CLASS (This page)

                                                 Unclassified
                              -5Q_
                                                                         22. PR: E
 EPA Form 2220-1 (Rev. 4-77)
                       PREVIOUS EDITION iS OBSOLETE
                                            43

-------