6143
                                   DRAFT FINAL REPORT




                                           TO




                          GREAT LAKES NATIONAL PROGRAM OFFICE




                      UNITED STATE ENVIRONMENTAL PROTECTION AGENCY












                      MODELING THE BEHAVIOR AND FATE OF NUTRIENTS




                       AND TRACE CONTAMINANTS IN THE UPPER GREAT




                               LAKES CONNECTING CHANNELS
                                     NOVEMBER 1987






                         INTERAGENCY AGREEMENT DW  13931213-01-0




                                        BETWEEN






                     GREAT LAKES ENVIRONMENTAL RESEARCH LABORATORY




                    NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION




                                  ANN ARBOR, MICHIGAN






                                          AND






                          GREAT LAKES NATIONAL PROGRAM OFFICE




                     UNITED STATES ENVIRONMENTAL PROTECTION AGENCY




                                   CHICAGO,  ILLINOIS

-------
                                      U.S. DEPARTMENT OF COMMERCE
                                      National Oceanic and Atmo&^t.ariu Administration
                                      ENVIRONMENTAL RESEARCH LABORATORIES

                                       Great Lakes Environmental Research Laboratory
                                       2205 Commonwealth Blvd.
                                       Ann Arbor, Michigan  48105-1593
                                       December 8, 1987
R/E/GL
Mr.  Anthony Kizlauskas,  Project Officer
Great Lakes National Program Office
U.S.  EPA
230 S.  Dearborn
Chicago, IL  60604

Dear Tony,

As part of our interagency agreement (IAG No. DW13931213-01-0), I am pleased
to furnish you with a draft final report on research conducted during  the
period December 1, 1984 - November 30, 1987.

We welcome your comments and suggestions.

                                       Sincerely,
                                       Thomas D. Fontaine
                                       Head, Environmental  Systems  Studies

Enclosure

cc:  Beeton

-------
              DRAFT FINAL REPORT




                      TO




     GREAT LAKES NATIONAL PROGRAM OFFICE




 UNITED STATE ENVIRONMENTAL PROTECTION AGENCY
 MODELING  THE BEHAVIOR AND FATE OF NUTRIENTS




  AND  TRACE  CONTAMINANTS  IN THE UPPER GREAT




           LAKES  CONNECTING CHANNELS
                 NOVEMBER 1987







    INTERAGENCY AGREEMENT DW  13931213-01-0




                    BETWEEN







 GREAT LAKES ENVIRONMENTAL RESEARCH LABORATORY




NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION




              ANN ARBOR,  MICHIGAN







                      AND







      GREAT LAKES NATIONAL PROGRAM OFFICE




 UNITED STATES ENVIRONMENTAL PROTECTION AGENCY




               CHICAGO, ILLINOIS

-------
                            TABLE OF CONTENTS

                                                                        PAGE


EXECUTIVE SUMMARY	   1

INTRODUCTION	   6

REPORTS

Unsteady Flow Model of Entire St. Clair River  	   8
        J. A. Derecki, L. L. Makuch, and J. R. Brook

St. Clair and Detroit River Current Measurements 	   92
        J. A. Derecki, K. A. Darr, and R. N. Kelley

Development of a Shallow Water Numerical Wave Model for Lake St. Clair  .  135
        D. J. Schwab  and P. C. Liu

Modeling Particle Transport in Lake St. Clair	145
        D. J. Schwab  and A. H. elites

Total  Phosphorus Budget  for Lake  St. Clair:   1975  - 1980	161
        G. A. Lang, J. A. Morton, and T. D.  Fontaine,  III

 Phosphorus Release  from  Sediments and Mussels  in Lake  St.  Clair,
   with Notes  on  Mussel Abundance  and Biomass 	  190
         T.  F. Nalepa, W.  S. Gardner, and J.  M. Malczyk

 Sediment  Transport  in Lake  St.  Clair	213
         N.  Hawley  and B.  Lesht

 Accumulation of  Fallout  Cesium-137  and Chlorinated Organic Contaminants
   in Recent Sediments of Lake St. Clair	289
         J.  A. Robbins and B.  G.  Oliver

 Toxicokinetics of Selected Xenobiotics in Hexagenia limbata:   Laboratory
   Studies and Simulation Model 	  357
         P.  F. Landrum and R.  Poore

 Modeling the Fate and Transport of Contaminants in Lake St. Clair   . . .  386
         G.  A. Lang and T. D.  Fontaine, III

-------
                            EXECUTIVE SUMMARY
The following points summarize the findings of our research:


Unsteady Flow Model of Entire St. Clair River

-  This is the only flow model for the entire St. Clair River, including its
   extensive delta.

-  Model provides for flow separation around islands in the upper river and
   through the main delta channels in the lower river.

-  Three model versions provide  information on river stages (profile),
   discharge, or velocities along preselected portions or the entire river.
   Daily or monthly tabulations  of data with corresponding means can be
   furnished to users.

 -  The model furnishes hydrological data needed  to compute contaminant mass
   balances.
 St. Clair and Detroit River  Current Measurements

 -  The measurements represent  a  unique  time  series of long-term, continuous
   velocity  measurements  in  the  Great Lakes  connecting  channels  in  the  lake.

 -  The measurements permit evaluation of  winter  ice  effects and  weed  effects
   during most  of  the year on  the flows of the St. Clair  and  Detroit  Rivers.

 -  The measurements permit evaluation of  different types  of current meters.
   Single-point, contact  sensors (electromagnetic meters)  present problems
   because  of weed clogging  effects;  remote  sensors  (acoustic Doppler-shift
   meters)  are  ideally  suited  for use  in  the river but  are expensive.   The
   acoustic Doppler current  profiler  measures velocity  distribution in  the
   overhead vertical  water  column, which  could not be duplicated with a
   string  of point-measuring meters because of ice and  navigation problems.

 -  The  measurements provide  data for  adjusting winter flows  in the  St.  Clair
    and  Detroit  Rivers.
 Shallow Water Wave Model for Lake St. Clair

 -  A shallow water version of our deepwater wave model tends to
    underestimate the highest waves at all stations in the lake.

 -  The deep water version of the model provides quite acceptable estimates
    of waveheight, even for the largest waves at the shallowest stations and
    is therefore quite acceptable in Lake St. Clair.

-------
  The wave model  can be  used to drive  sediment  resuspension  in contaminant
  fate models.
Particle Transport Model for Lake St.  Glair

-  Although the  average hydraulic residence time for Lake  St.  Clair is about
  nine days,  residence times of water from individual tributaries range
  from 4  to over 30 days in the absence of wind.

-  East winds greatly decrease the expected residence time of water from the
  Thames  River, while northwest winds increase the residence time of water
  entering from the St. Clair Cutoff and St.  Clair Flats.

- Based on circulation patterns generated by 6 months of  real wind
  conditions,  water entering the lake from the Thames River or one of the 3
   lower  St. Clair River channels has a large probability  of mixing with
  waters  of the eastern two-thirds of the lake.

-  Flows  from the upper St. Clair River channels and the Clinton River
   follow  well-defined routes that are restricted to the western third of
   the lake.

   Circulation model predictions compare favorably with current measurements
   made by drifters and current meters in Lake St. Clair in 1985, although
   the model appears to be underestimating the current speeds.


Lake St.  Clair  Phosphorus Budget  1975 - 1980

 -  Lake Huron was  the major source of phosphorus to Lake St. Clair,
   accounting for  approximately  52% of the total annual load.  Hydrologic
   area loads (diffuse  and  indirect point  sources) contributed 43% of the
   total load.   The remaining 5%  came from the atmosphere, shoreline
   erosion, and direct  point sources.

 -  The Thames and  Sydenham  areas  of Ontario contributed 75% of the total
   hydrologic area load (92% of  the total  Canadian hydrologic area load).
   The Clinton  and Black areas  of the U.S. contributed 15% of the total
   hydrologic area load (83% of  the total  U.S. hydrologic area load).
   Eighty-five  percent of  the total hydrologic  area  load was  from diffuse
   sources.

 -  Averaged over the  six-year study period, estimated external loads were
   not significantly  different  from estimated  outflow losses.  Therefore,
   there  does not appear to be  a significant net  internal source  or  sink of
   phosphorus  in Lake  St.  Clair during  1975-80.


 Sediment  Phosphorus Dynamics  and Mussel  Populations

   Soluble reactive phosphorus  release  from Lake  St.  Clair  sediments
   averaged 19  ugP/m^/day.   This represents  about 1% of the  total
   bioavailable phosphorus load that  annually  enters the  lake from other

-------
  sources.  Mean maximum release was 47 ugP/m^/day, or about 2% of the
  annual bioavailable load from other sources.

  Mean mussel density was 2/m^ and mean biomass was 4.3 gDW/m^.  Of the 20
  species collected, Lampsilis radiata siliquodea and Leptodea fragilaris
  were the dominant forms, accounting for 45%  and 13% of  the total
  population.  The former species was dominated by older  individuals, which
  may indicate the population is declining.

  The mean excretion rate of phosphorus by mussels was 1.4 ugP/gDW/h.  On a
  lake-wide basis, this amounts to 5% of  the bioavailable load from other
  sources.
Sediment Transport  in Lake  St.  Clair

   Sediment  resuspension  in Lake  St.  Clair is  due  primarily to  wave  action.

-  The  initiation of sediment resuspension can be  predicted using  a  simple
   model with wave  orbital  velocity as the forcing function.

-  Critical  values  of wave  orbital velocity for resuspension range from 0.1
   cm/s to 0.9 cm/s when  calculated one meter  above the bed.

-  Variations in the critical orbital velocity may be related to substrate
   characteristics, but this cannot be proven  with the data available.

-  The  orbital velocities are calculated from  the  wave field produced by the
   GLERL wave model.  This  model requires over-lake wind velocity
   measurements as input.


Cesium-137 and Chlorinated Organics in Lake St. Clair Sediments

   Eight percent of atomic fallout Cs-137 that entered Lake St. Clair mainly
   in the  mid 1960s is still present in sediments 20 years later (1985).

   Comparison of total Cs-137 loading with 1985 Cs-137 storage indicates a
   sediment residence time of about 5 years.  This is consistent with
   previously studied loss rates of mercury with chlorinated organics from
   the bottom.  Changes  in total storage of the radionuclide between 1976
   and 1985, however, implies a longer residence time of about 15 years.

   Probably because of mixing and burial mechanisms, the residence time of
   this tracer and other contaminants appears to increase over time.

   Resuspension of sediments is continuing to supply radiocesium to the
   water where it  is  exported from the lake.

 -  Chlorinated organic compounds, like Cs-137, are associated with fine
   sediments  and preferentially deposit  in the deepest water.

-------
  The sediments presently contain nearly 1 metric ton of HCB and PCBs and
  0.2 metric tons of OCS, far higher contaminant masses than those  found in
  the St. Clair River.

  Based on the behavior  of  Cs-137,  crude estimates of total particle-
  associated contaminant  loads  to Lake  St. Clair are: HCB, 15 MT; QCB,  2.8
  MT; HCBD, 3.3 MT; OCS,  4.3 MT; PCBs,  20 MT;  total  DDT, 5.4 MT.
Hexagenia  and  Contaminants

-  Hexaeenia are an important food source for fish in the  connecting
   waterways of the Great Lakes.   Because of Hexagenia's position in the
   foodweb,  their losses due to toxic effects of contaminants,  or their
   bioaccumulation of contaminants and subsequent transfer to fish,  are
   important topics to study.

-  Contaminant uptake and depuration rates of Lake St. Clair Hexagenia
   limbata vary seasonally.  A simulation model suggests that temperature
   mediated changes in the depuration rate constant is the term responsible
   for most of the seasonal variation.

 -  The contaminant accumulation rate constant of Hexagenia limbata is
   similar to that of other Great Lakes invertebrates, but its depuration
   rate constant is much larger.

 -  Hexagenia limbata obtain a greater percentage of their contaminant body
   burden from  the sediment  than do other Great Lakes invertebrates.
 Generic Contaminant Fate and Transport Modeling

 -  A multi-segment model to simulate the distribution and transport of
   contaminants  in Lake St. Clair was developed based on the EPA's Chemical
   Transport  and Fate Model TOXIWASP.  The model segmented Lake St. Clair
   into 126 well-mixed segments: 42 water segments, 42, active sediment
   layer  segments, and 42  deep  sediment layer segments.

 -  Simulations of the fate and  behavior of chloride, cesium-137, and  the
   organic contaminants octachlorostyrene (OCS) and polychlorinated
   biphenyls  (PCB) were carried out.

 -  The ability of the model  to  accurately predict  the movement and
    concentrations of a conservative  tracer was substantiated using chloride
    data gathered during four,  1974  cruises.

 -   The ability of the model  to  accurately predict  the distribution and
    concentrations of highly  partitioned contaminants was  tested with  Cesium-
    137 data.   Through calibration of the  spatially-varying  organic carbon
    content of the  sediments,  the model matched the magnitude  and
    distribution  of  observed  1976 cesium-137  in surficial  (0-2  cm)  sediments.
    The calibrated  sediment organic carbon content  values  ranged from  0.14%
    to 5%  (areal  mean  =  1.27%)  and are well within  the  range and spatial
    distribution of  observed  values.   Simulations were  continued out  to  1985

-------
for model verification purposes.  The predicted distribution of sediment-
bound Cs-137 in the active layer compared well with the observed 1985
values .

Using an estimated load (1.9 Ibs/d) of DCS to Lake St. Clair, the model
was run until simulated DCS levels in the active layer (0-10 cm) agreed
with observed 1983 values.  This occurred at 4500 days and implies that
DCS loadings began in the latter part of 1970.  This result is consistent
with speculation that OCS was introduced to the lower Great Lakes
beginning in the 1970s.  The model predicted 1983 mean and maximum OCS
sediment concentrations of 3.8 and 23.1 pg/kg, respectively.  These
compare with observed values of 2.7 and 26.2 jjg/kg.  The model predicted
1983 active layer bio-bound OCS levels of 0-96 A»g/kg dry weight, with a
mean concentration of 20 pg/kg.  The observed range was 2-154 /ig/kg dry
weight (mean = 43 MgAg) measured in whole clam tissue collected in 1983.
Using observed 1970 PCB levels as initial conditions and an estimated PCB
load (5.1 Ib/d) , the model accurately predicted  the 1974 PCB  sediment
(0-2 cm) concentrations in Anchor Bay and open- lake sediments.  However,
the model tended to overpredict  the PCB values along the eastern  and
western segments of the main  lake, which may  indicate additional  or
increased PCB sources in these areas.  The data  showed a decline  in mean
lake -wide sediment concentration of total PCB from 19 /ig/kg  (max. = 40
/ig/kg) in 1970 to 10 pg/kg (max. - 28 pg/kg)  in  1974.  The model
simulated a similar 4 year decline in mean sediment concentration from
19.8 to 8.7 pgAg (max. from  39  to 26 pg/kg) .  The model predicted a 1974
mean active layer bio-bound PCB  concentration of 97 ^g/kg dry wt. (range
-  31-212 pg/kg) .  This compares  with observed mean values of  90.6 and
44.2 /ig/kg dry wt. (Aroclors  1254 and 1260,  respectively) measured in
whole clam tissue in 1983.

-------
                               INTRODUCTION

[he Upper Great Lakes Connecting Channels  Study  (UGLCCS)  is a multi-agency,
nulti-national study of the  St. Marys River,  the St. Clair River,  Lake St.
Dlair, and the Detroit River.   The  goals of the  study  include:

1.  Determining the present  environmental  status of  the  study area;

2.  Identifying and quantifying sources of ecosystem degradation in the
   study area;

3.  Assessing  the adequacy of existing or  planned  control programs;

4.  Developing long-term  monitoring programs  for assessing  the  effectiveness
   of control programs;

5.  Facilitating  the development of remedial  action plans by  the the
   Province  of Ontario  and  the State of  Michigan.

Towards  accomplishing  these  goals,  the  Great  Lakes Environmental Research
Laboratory  (GLERL)  of  the National Oceanic and Atmospheric  Administration (NOAA)
designed modeling,  field, and laboratory  studies of the connecting channels
study area  and processes  therein.   Through the Activities Integration Committee
or other less formal avenues, GLERL's studies were carefully coordinated with
proposed or ongoing studies  of other agencies in order to maximize scientific
insights.   A cross  reference showing the  correspondence between GLERL's studies
and UGLCCS  activity numbers  is provided in Table 1.  In some cases the GLERL
studies  were associated with more than one of the UGLCCS activities.

As part  of  the NOAA -  EPA interagency agreement, GLERL provided the Great Lakes
National Program Office  (GLNPO) with written quarterly reports describing
progress towards meeting the goals of the study, and oral presentations
summarizing each year's work.  Scientists at GLERL are presently  submitting the
results of their work to professional scientific journals.   This  draft final
report documents research conducted during the entire interagency agreement
period,  December 1, 1984 through November 30, 1987

-------
?able 1.  Correspondence of GLERL Activities  with UGLCCS Activities.

5LERL Activity                                              UGLCCS Activity No.


Jnsteady Flow Model of  Entire  St. Clair River 	    C.5

;t. Clair and Detroit River Current Measurements 	    C.5

)evelopment of  a  Shallow Water Numerical Wave Model
:or Lake St. Clair  	    C.2

Modeling Particle Transport  in Lake St. Clair 	    C.2

[otal  Phosphorus  Budget for Lake St. Clair:  1975 - 1980  	    C.I

Phosphorus Release  from Sediments and Mussels in Lake St. Clair
vith Notes on Mussel  Abundance and Biomass 	    H. 20

Sediment Transport  in Lake St. Clair  	    G. 3

Accumulation  of Fallout Cesium-137 and Chlorinated Organic
Contaminants  in Recent Sediments of Lake St. Clair  	    G.1.G.2

Toxicokinetics  of Organic Xenobiotics  in the Hexagenia  limbata:
laboratory Studies  and Simulation Model  	    H. 16

Modeling the  Fate and Transport  of Contaminants  in Lake St.  Clair .    C.1,C.2,C.3

-------
             UNSTEADY  FLOW  MODEL OF ENTIRE ST.  CLAIR RIVER
         Jan A.  Derecki,  Laura L.  Makuch,  and Jeffrey R.  Brook
                                 ABSTRACT









    This report describes  the  development  and calibration of an unsteady




flow model for the  entire St. Clair River,  from Lake  Huron to Lake St.




Clair,  to simulate  hourly and daily flow rates.    Unlike previous St. Clair




River hydraulic models that are limited to  the upper  single-stem river




channel, the  present model  versions (hourly or daily) provide flow




separation around  Stag and  Fawn Islands in the upper  and middle river, and




through the main delta channels (North, Middle, South, and Cutoff) in the




lower river.    The model provides three options for simulation of the river




stages (profile),  discharge, or velocities, respectively.   This information




is needed  in order to predict the fate and transport of pollutants in the




entire river channel, including  the island and delta obstructions to  its




flow.   The model can be run for the entire river or any preselected  river




reach bounded by water  level gages.
                                INTRODUCTION
      A series  of  hydrodynamic models have  been developed and used




 extensively  at the  Great Lakes Environmental  Research Laboratory (GLERL)  to

-------
Consequently,  interest in and applications of model-simulated flows are




expanding from total flow rates to flow distribution and localized effects.









     The principal goal of this Upper Great Lakes Connecting Channels  Study




(UGLCCS) activity was to develop the hydrodynamic models for selected  upper




connecting channels of the Great Lakes.   The St. Clair River was selected




because it forms the upper portion of the outlet through the St. Clair River




- Lake St. Clair - Detroit River system from the upper Great Lakes




(Superior, Michigan, and Huron).   This unsteady flow model of  the entire




St. Clair River, from Lake Huron to Lake  St. Clair, provides flow separation




around Stag and Fawn Islands  in the upper and middle river, and through  the




main delta channels of the North, Middle, and South Channels and the  St.




Clair Cutoff  in the lower river (Figure 1).   The forcing  functions  in the




model are the river stages recorded at the water level gages enclosing




preselected river  reaches  (entire  river or  longitudinal  segments).    These




water levels  from  the extreme gage  locations  form the model's boundary




 conditions.
                             MODEL  DEVELOPMENT









      The present unsteady flow model for the entire St.  Clair River is an




 extension and modification of the existing GLERL upper river model versions.




 All these models are driven by water level data taken from appropriate water




 level gages along the river.   The present model also uses the St. Clair




 Shores gage in Lake St. Clair to indicate water levels at the mouth of the
                                       10

-------
river.    To permit near real-time model applications only the official

National  Oceanic and Atmospheric Administration (NOAA) water level gage data

available at GLERL are used in the model.



Unsteady Flow Equations



     The unsteady  flow model  is  based on complete one-dimensional partial

 differential equations of  continuity  and momentum.   The  momentum equation

 includes the effects  of  motion but  neglects  the  effects of  wind stress and

 ice.   Except  for  short  periods  associated with  storms, the wind stress

 effects  were found to be generally  insignificant on the St. Clair River

 flows  in a previous study (Derecki  and Kelley, 1981) and  the wind data are

 normally not  available for real-time applications.   The  effects of

 transient ice  flows and resulting ice jams in the lower river are

 significant and may be substantial  during winter and early spring in the St.

 Clair River.    However,  no tested method for  including these effects is

 presently available.



      Expressed in terms of flow Q  and stage Z above a fixed datum,  the

  equations of continuity and  motion are  as follows:
  3t   T  3X
  1 3Q .  2QT 3Z + (g .  Q2T)  3Z + gn^Q/Q/ _ _ 0                       (2)
  A 3t   A2  3t        A3   aX   2.208 A2RV3
                                        11

-------
where   X - discharge in the positive flow direction

       t - time

       A - channel cross-sectional area

       T - top width of the channel at the water surface

       g - acceleration due to gravity

       R - hydraulic radius

       n - Manning's roughness coefficient

       3 - partial  derivative function

       // - absolute value.


 Channel definition  is  shown  in Figure 2.    Equations  (1)  and (2) were placed

 in finite  difference form at  point M  in  an  X-t grid (Figure 3)  to yield

 respectively,


 Zu' + Zd'  -  Zu -  Zd .  6 (Qd'  - Qu') + (1-9) (Od -  Out  _ Q              (3)
         2 At                       T  AX


 Qu' + Qd'  -  Qu -  Qd .  QT (Zu'  +  Zd' - Zu -  Zd) +

        2 A At                   A2 At


      (g - 02T) .  6 f(Zd' - Zu')  + (1-6)(Zd - Zu)1  +

          A3                  AX
        gn2 0/Q/     _  0                                                (4)
            _2 4/3
     2.208  A R
  where a prime indicates location and overbars indicate mean, such  that
  9 = A£l                                                                (5)
      At
                                        12

-------
Q- 0.5  19  (Qu'  +  Qd')  + (1-9) (Qu + Qd)]                              (6)







A - 0.5  [8  (Auf  +  Ad')  + (1-6) (Au + Ad)]                              (7)









     Solution of equations  (3) and  (4) by  the  implicit method forms the




basis of the numerical unsteady  flow model.    A stable solution for these




equations is provided by the  weighting  coefficient 9,  which was selected




empirically (Quinn and Wylie,  1972)  to  be  0.75.   Application of the




equations at the  river's cross-sections selected to define the actual river




 channel produces  a set of  nonlinear equations  that are solved simultaneously




 with linear approximations by the Newton-Raphson numerical iteration




 procedure.   In the  initial St.  Clair River model version an idealized river




 channel, based  on averaged river cross-sections for selected reaches, was




 used.   The use of  idealized river channel simplifies simulation of




 discharge  but prevents  valid velocity determination.   Description of the




 initial St.  Clair River model, including calibration, sensitivity analysis,




 program listings, and output samples, are given by Quinn  and Hagman  (1977).




 The  initial model has been revised by Derecki  and Kelley  (1981) to replace




 the  idealized river channel with the actual configurations  of  the river and




 to include wind stress effects.









 Mathematical Solution









      A schematic diagram  for the entire  St. Clair River  model,  including  its




  delta,  is shown  in  Figure 4.    The model  can  be run for  any river  reach




  containing at  least three NOAA  water level gages by specifying gage




  locations  for  the beginning and ending boundaries of the reach;  the  mid-gage
                                        13

-------
is used to check the accuracy of the computed river profile by comparing




deviations between computed and measured water levels.   In the previous St.




Clair  River  models the mathematical solution of the model equations was




provided  by  using banded matrix, which is most efficient for solving single -




channel configurations.









     However, this matrix is impractical for solving  flow separations  and




was replaced with sparce matrix  in  the present model.  The Yale Sparce




Matrix package available in  the  GLERL computer library is used  in the  model.




All the St.  Clair River model versions are  the hydraulic transient models,




which differ from standard profile  or backwater  computations; the hydraulic




 transient models  include time dependent  terms of mass continuity  and




 momentum, which  allows the simulation of wave propagation as well as




 profiles along  the  river.









      Initial work on the new delta-model development included extending the




 existing single-stem upper river model  through  the middle  river (St.  Clair




 to Algonac), with modification to provide  flows  around  the  Stag and Fawn




 Islands.









      Computations for continuity and momentum around an island start at the




  downstream channel junction or node, proceed along one  side of the island to




  a breakpoint section just below the upstream node, then return to the other




  side of the island and proceed  to  its breakpoint; the breakpoints are then




  combined at the upstream node and  the computations continue upstream  in the




  single-stem channel  (Figure 5) .    Because of mass continuity at the nodal
                                        14

-------
      , representing  channel  separation or confluence, the water  level is




;he  same for the joint  and separate channels, and the flow in the joint




:hannel is the  sum  of flows in separate channels.   This provides additional




continuity equations  for the  water surface and flows at the nodes, which are




listed below.
WSn - WSsl - WSs2                                                       (8)






Qn - Qsl + Qs2                                                          (9)









     The island-model version was extended  to  Lake  St.  Clair by treating the




upper delta as an island and the lower delta as  a composite  of two islands




with separate channels.   The composite delta  islands  are  terminated with




short imaginary channels in the lake.   This treatment of  delta flow




distribution provides separation of  flows  through the  North  Channel and




South Channel in the upper delta, and consequent respective  separation




through the lower North and Middle Channels, and the lower South and St.




Clair Cutoff Channels  in the lower delta,  covering  all the main delta




channels.   To help  the program converge  on a  solution more quickly, the




 initial flow values  around the  islands  and the delta are provided in the




 model.    These initial  flows are  the fixed percentages of the normal total




 flow, based on existing measurements.    Nearly all  gages used in the model




 were moved at various  times  and  some with more recent moves may have




 different numbers  specified  in  the  model.    Occasionally,  a gage in the same




 location may have  experienced  a vertical movement,  due to some corrective




 measure performed  on the  gage;  such corrections for the vertical movement
                                       15

-------
are also  specified in the model.   These corrections are based on results




determined  in  previous studies and are included in previous model versions,




where appropriate.   Because all physical and hydraulic input data are  in




English system,  the basic model computations are performed in the English




units and the  final result-output converted to the SI system of units,  if




desired.









     To initialize the computations  the model  is operated until a steady




state is achieved, prior to simulation of actual data.   This is




accomplished by successive  iterations of the continuity and momentum




equations in their discrete form  (finite difference) until an acceptable




tolerance is attained.









      Based on previous models,  the  number of  this  "steady  state"  iterations




 is preset at 12.   Occasionally,  the water  level gages break down providing




 erroneous or no water level records.   However,  these  records are necessary




 for the  gages forming the  model's boundary  conditions.   To permit




 initiation of computations the missing  data are  estimated  from  long-term




 means (or previous values  within the run),  which most  likely would  not be




 sufficiently  accurate.    Initiation of  computations in such cases may




 require unreasonably large number of iterations  to achieve "fictitious"




 steady state  (with erroneous  results),  and  the preset  number  of 12




 iterations eliminates such possibility.    To make  the  user aware of possible




 inaccuracies  (along with causes) , all partial or missing/estimated water




 level inputs  are  flagged in the model outputs.
                                       16

-------
    Model computations are performed for  each river reach between


successive sections used  to define  the river channel.    For the entire


river, with the delta,  this involves 180 cross-sections of the river


channel.   Input  data  for these computational reaches are obtained by


averaging records of  the  successive bounding sections.   Except for the


starting and  ending reaches,  each reach contains four unknowns, which  are


the upstream  and downstream water surfaces and flows.   The starting and


ending model  reaches contain three unknowns because the upstream  and


downstream sections in these reaches,  respectively, correspond to the  water


 level gages with known water surfaces (model's boundary conditions).    The


 model equations  are set-up in  the  sparce  matrix  as  alternating rows of


 continuity (odd  rows)  and momentum (even  rows) equations,  as  shown in Figure


 6.   The matrix  non-zero values are indicated by X's and are  the partial


 derivatives  of  the equations  for indicated rows  with respect  to the


 variables in indicated columns.   The partial derivatives of  the continuity


 and momentum equations with  respect  to water surface and flow are listed


 below.
  
-------
                         _2
3M   -    0 T    +  (g -  Q_I)  -  §_                                       (14)
5Zu'      _2              _3     AX
         A At            A
3Zd'       _2              _3    AX
          A At           A
3M    -     1     -  6 T (Zu' + Zd' - Zu  - Zd)
3Qu'        _                  _2
          2 A At             2 A At
          e Q T re (Zu' - zd') +  a  - euzu  -zd)i  +
                           _3
                           A AX
               2 _
          9 g n  0                                                      (16)
                _2 4/3
          2.208 A R
 3M    -  3M	                                                           (17)
 3Qd'      3Qu'
                             MODEL  CALIBRATION



 Model Scope



      As mentioned in the preceding discussion, the model has two versions

  (with separate programs) for the simulation of hourly or daily river

  profiles and resulting flows.   The model can be operated for the entire

  river, with separation of flows around the upper river islands and  through

  the main delta channels in the lower river, or for any river reach  bounded
                                        18

-------
by NOAA water  level gages,  which form the model's boundary conditions  (a




minimum of three gages are employed, with the mid-gage used  to  provide a




check on  simulation accuracy).    Because all river profile and  flow




information for the entire river will not fit on a computer  page,  three




separate  options are provided for the river stages,  discharge,  and




velocities, respectively.









Hydraulic  Parameters









      The  hydraulic parameters needed to operate the  model are the river




 stations,  the  top  channel  widths,  the datum reference elevations, and  the




 base areas below the  datum for  each section used to  define the river




 channel.   Because of the  large number of sections  (180), printout of  this




 information  is normally suppressed in the model output but is contained in




 the program  and can be easily  reinstated.   Other hydraulic parameters




 needed to run the model are the water surface elevations for the water level




 gages and the roughness coefficients for the river  reaches bounded by




 successive water level gages (8).   All other data  needed in the




 computations  (total channel area, hydraulic radius,  length  of  river reaches,




  etc.) are determined from the above data.









  Model Calibration









       Calibration  of  the model  consisted of adjusting the roughness




  coefficients  of the  river channel,  which is  the  unknown in the flow equation




  during periods of flow measurement.    The  channel  roughness coefficients
                                        19

-------
rere determined for each river reach  bounded by successive water level

jages, with separate coefficients  for the North and South delta channels,

for a total of 8.  The roughness coefficients were derived from 14 sets of

flow measurements on the St.  Clair River conducted by the Corps of Engineers

during 1959-77.   The equation used to compute the roughness coefficients is

the Manning equation, which is
             2/3
n - 1.486 A R    .  (Zu -  Zd +  0 AA )                                    (18)
          Q            L           3
                              g L A
     The relationships between computed roughness coefficients  for  the  8

 successive river reaches and either upstream or downstream river  stages are

 shown in Figures 7-14.   These relationships normally represent the best-fit

 lines derived by regression (least squares) for graphs  indicating slope or

 the arithmetic means  for graphs  in which plotted data did not  indicate  any

 slope.



      Thus,  the roughness coefficient  graphs for  the upstream reaches (FG-DP

 and DP-MBR)  indicate  positive  slope;  those for most of  the  single-channel

 river (MBR-DD, DD-MV, MV-SC, and SC-AL)  indicate no slope or change in

 channel roughness with  water level  elevation;  both  delta reaches  are between

 AL-SCS but are designated  AL-ND  and AL-SD  to  indicate  north and south delta

 channels,  respectively,  and show negative  slopes.    In the  lower  delta four

 separate channels are actually used but there were  insufficient data to

 derive separate  roughness  coefficients.    The downstream river channel, from

 the St. Clair City  through the delta, was  affected by  regimen changes
                                       20

-------
between 1959-63,  when extensive dredging was conducted for navigation




improvements.    For these downstream reaches separate roughness  coefficients




were derived for each regime, representing pre-project conditions  (through




1963) and current conditions (starting in 1964).   These  and other St.  Clair




River regimen changes are extensively analyzed and documented in various




studies (Derecki, 1985).   The computed roughness coefficients actually




represent channel roughness  and combined effect  of possible  errors, such as




 those  in flow measurements and determination of  channel parameters, and the




 computed coefficients for some reaches were modified somewhat when such




 change was  strongly  indicated  during  the model calibration process.   Thus,




 calibrated  roughness  does not  always  represent  the best-fit line for plotted




 data (Figures 9  and  11).   The calibrated  roughness  coefficients for the 8




 river reaches are  summarized in Table 1.









      The Ft.  Gratiot and St. Clair water  level  gages were moved in 1970,




 with apparent uncompensated hydraulic effects.    These effects were




 determined from a comparison study as a 0.055  m (0.18 ft) reduction for the




 Ft. Gratiot gage levels and a 0.027 m (0.09 ft)  increase for  the  St. Clair




 gage  levels (Quinn,  1976).    The Ft.  Gratiot gage was modified again in




 1981,  following blockage of its intake by silt,  with apparent uncompensated




 hydraulic  effect which was determined as an increase in  its  elevation  of




 0.037 m  (0.12 ft) from the preceding period (Derecki, 1982).    Thus,




 effective  Ft. Gratiot uncompensated hydraulic effect was decreased in  1981




  to a reduction  of 0.018 m (0.06 ft)  in the gage levels.   These vertical




  gage-record  corrections are included in the model to provide unbiased




  continuation of the  water levels  at  the gages.
                                        21

-------
Computer  Programs









     The  St.  Clair  River model for the entire river, with flow separation




around islands and  through main delta channels, uses water level data  from




GLERL computer disk pack files (VAX) , as did preceding models.   Two




generalized versions of the model for simulating hourly and daily  flow




rates, respectively, were prepared and stored in the computer files.    These




model versions operate on hourly or  daily computational time scales and




provide summary tables for daily or  monthly data,  respectively.    Each model




version has three options for the output of simulated  river stages at  the




 gage  locations and the total discharge or average  velocities at  selected




 points,  as well as separate values  around islands  and  through  the  delta.




 Separate  output options were provided since all this  information would not




 fit on a single computer page.   A  check of model  accuracy  is provided in a




 form of  water level deviations between computed and measured water levels at




 the water level gage  sites.   Basic model computations are  listed  in the




 program  in English  units;  the final results  are printed in  either  English or




 SI units, as specified in the output option.    The hourly and daily model




 versions are listed in the  Appendix as Figures A-15 and A-16,  respectively.




 Examples of model  outputs for the water  levels, discharge,  and velocity




 options are shown  in the Appendix Tables A-l through A-3 for the hourly




 model, and Tables  A-4 through A-6 for the daily model, respectively.
                                       22

-------
                               CONCLUSIONS









    This model was developed to correct or eliminate  the  shortcomings  of




•revious model versions.   It simulates the St. Clair  River profile  either




:or  the entire length of the river or for selected  segments,  and provides an




accuracy check for the computed profile at the water-level-gage locations.




rhe  model simulates flows (discharge or velocity)  in all  the  more important




river  channels, providing flow  separation around  the islands  in the  upper




and middle river, and through the main delta  channels  in  the  lower river.




As such,  it should become a  valuable tool for both hydraulic  and other water




resource  studies  in the  upper Great Lakes basin.
                             ACKNOWLEDGEMENT









     The authors gratefully acknowledge work performed by D. L. Schultz in




 the initial phases of model development.
                                       23

-------
                            LITERATURE CITED









•erecki, J.A., 1982.   Effect of the 1981 Fort Gratiot  gage  modifications on




    the hydraulic regime of the St. Clair River.    GLERL Open File Report,




    NOAA, Great Lakes Environmental Research Laboratory,  Ann Arbor, MI,




    3pp.









Derecki, J.A., 1985.   Effect of channel  changes  in the St.  Clair River




     during  the present  century.    Journal of Great Lakes Research.




     11(3):201-207.









Derecki,  J.A., and R.N.  Kelley,  1981.    Improved St. Clair River dynamic




     flow models  and comparison analysis. NOAA Tech. Memo.  ERL GLERL-34,




     NOAA Great  Lakes Environmental Research Laboratory, Ann Arbor, MI,  36





     PP-









Quinn, F.H., 1976.   Effect of Fort Gratiot  and St. Clair gage relocations




     on the apparent hydraulic regime of the St. Clair River.   GLERL Open




     File Report, NOAA Great Lakes Environmental Research Laboratory, Ann




     Arbor, MI, 7pp.









 Quinn, F.H., and  J.C. Hagman, J.C., 1977.   Detroit and St. Clair  River




      transient models.   NOAA Tech. Memo. ERL GLERL-14, NOAA  Great Lakes




      Environmental  Research Laboratory,  Ann Arbor,  MI, 45 pp.
                                       24

-------
Juinn, F.H.,  and E.B.  Wylie,  1972.   Transient analysis of the Detroit River




    by the implicit method.   Water Resources Research. 8(6):1461-1469.

-------
TABLE  1.  Roughness coefficients  for  the  St. Clair River  reaches.
 Reach       Roughness  coefficients (n)
 FR-DP        n -  0.0033947  (FG)  -  1.92253




 DP-MBR       n -  0.0002708  (DP)  -  0.12683




MBR-DD        n -  0.0221




 DD-MV        n -  0.0250




 MV-SC        a. Current regime (starting 1964):     n - 0.0240




              b. Pre-project regime (through 1963):  n - 0.0260




 SC-AL        a. Current regime:      n - 0.0230




              b. Pre-project regime: n - 0.0235




 AL-ND        a. Current regime:      n - -0.0017647 (SCS) + 1.04729




              b. Pre-project regime: n - -0.0053707 (SCS) + 3.11427




 AL-SD        a. Current regime:      n - -0.0011146 (SCS) + 0.66250




              b. Pre-project regime: n - -0.0032907 (SCS) + 1.90968
                                       26

-------
                          LIST OF  FIGURES
1. St. Clair River with location of water level  gages.




2. Channel definition sketch.




3. X-t grid for the implicit method.




4. Schematic model diagram.




5. Representation of an island.




6. Sparce matrix.




7. Roughness coefficient for FG-DP  reach.




8. Roughness coefficient for DP-MBR reach.




9. Roughness coefficient for MBR-DD reach.




10. Roughness coefficient for DD-MV  reach.




11. Roughness coefficient for MV-SC  reach.




12. Roughness coefficient for SC-AL  reach.




13.  Roughness coefficient for AL-ND  reach.




14.  Roughness coefficient for AL-SD  reach.
                                    27

-------
               Michigan
SCALE IN MILES
    *>

KILOMETERS
  20
10
       20
30
                                      Ft. Gratioti
                                    Dunn Paper<
                             Mouth of Black River*
                                     Dry
                                          Ontario
        St. Clair Shores

         Grosse Point
                                             LAKE HURON
                                23

-------
^*- Water Mass at t+At
     datum
            AX

-------

t+At


•





Q'u
Z'u
At
I
1 Qu
Zu



1
1
1
If
t 1
1 —
1
1 	 Ax 	 *•


Q'd
Z'd

Qd
Zd











 u            d




	X	
              30

-------
                          ST. OAIR SHORES
                                              LAKE
                                            ST. OAIR
                          ttARYStlUE (155)

                          MT DOCS (IS7

                          rlOUTH OF BLACK RIVER ( l»0)




                          DUNN PAPER (176)
                       ieo )  FORT 6RATIOT
                       ~*-S   (UU HURON)
ST CLAIR RIVER SCHEMATIC REPRESENTATION
                         31

-------
                              32
         j^i;i;;;^;i^1SLAND'1i;;;iii;;^ 10 *—
FLOW   »^;!iii;;j:ii!;i;;;;i;;;:i;i;i:!;Hi:iy^   FLOW

-------
c
•
c
•
c
1-2|
1-21
2-31
2-21
2-4 |
2-4 |
Cl
WS2
                 WS3    C3
x
X
X
X
X
X
X
X
                  X
                  X
                  X
                  X
                   X
                   X
                   X
                   X
                       X
                       X
                       X
                       X
                        33

-------
682n
581-
580-
579-
678H
577-
 576
 575-
 574
    0.02
                     N= 0.0033947 (FG)-192253
0.03         0.04         0.05
ROUGHNESS (MANNING'S N)
                                                    0.06

-------
682-,
 581-
      N= O.OXD2708 (DP)-0.12683
580-
 579 -I
 678-
 577-
 576-
 575 J
 574
1	1	1	1	1	1	1	1	1	1	r
                             -i	1	1	1	1	1	r
     0.027     0.028      0.029      0.030      0.031      0.032
                ROUGHNESS (MAILING'S N)
                        35

-------
   581
   580
   579-
   578-
    577-
LU  676
    575
    574-
    573
              N = 0.0221
o


o

o
        0.020
                                                  o
                                                  o
0.021      0.022      0.023      0.024

   ROUGHNESS (MANNNG'S  N)
                     0.025
                            36
                                                            f

-------
    681
   580
    579-
od
    578
3  677-
<
O



ULJ  576
    575-
     574
:
     573
                           o
                                   N = 0.0250
        0.023     0.024      0.025      0.026      0.027      0.028

                    ROUGHNESS (MANMNG'S  N)
                          37
                                                          10

-------
681
580H
679-
5 578-
a :
H
Uj 576-
D
575-

574-
R-7O -



O





N = 0.0240
(Current)
0 0

0 0
X
o
X
0
o
	 1 1 	 1 1 1 	 1 	 1 1 1
N = 0.0260
(pre-Project)



X



r i 	 1 	 1 i i i 	 1 i 	 1 i
    0.023     0.024      0.025      0.026      0.027      0.028
                ROUGHNESS (MANNWG'S N)
                         38

-------
578n
               N - -0.001H46 (SCS) + 0.66250
               (Current	o)
 574-
 673-
 572-
 571-
       N = -0.0032907(SCS) + 190968
       (Pre-Prqject	x)
 570
0.018   0.020    0.022     0.024    0.026    0.028
            ROUGHNESS (MANNING'S  N)
                                                      0.030

-------
                               APPENDIX A.
HOURLY AND DAILY UNSTEADY FLOW MODELS:
1. Model Programs (Appendix Figures A-15 and A-16)
2. Model Outputs (Appendix Tables A-l through A-6)
                                   42

-------
                 Program  [HYDRO.JDSTCLR]HDELTA.FOR

        This  is  the  St  Clair River  Transient  Model - Hourly Version.

        It is set to run in BATCH MODE...
        To run the program...

1.  Set desired parameters  in file  [HYDRO.JDSTCLR]HDELTA.PAR

        Line  1 - Starting and ending day,  month and yr.
                 DA  MO YR DA MO YR  (12,IX,12.IX,12,IX,12,IX.12,IX,12)

        Line  2 - Staring and ending points of model. (II,IX,II)
                 1  - Fort Gratiot
                 2  - Dunn Paper
                 3  - Mouth of Black River
                 4  - Dry Dock
                 5  - Marysville
                 6  - St Clair
                 7  - Algonac
                 8  - Lake St Clair

        Line  3 - Output Option  (II)
                 1   - Water  levels  and deviations.
                 2   - Total  discharge and discharge  around islands and, if
                     included,  discharge in  the delta channels.
                 3  - Velocity near the starting, ending and midpoint of
                     the simulated river and velocities around islands and,
                     if included, velocities in the delta channels.

        Line A  - Units  Option  (II)
                 0  - Metric units
                 1  - English units

 2.  Make sure that file  [HYDRO.JDSTCLR1HDELTA.DAT  is available.

 3.  Type:  SUBMIT HDELTA/NOTIFY

 4.  When your request is completed  the output will  appear in file:
                 [HYDRO.J DSTCLR]ZHDELTA.OUT
                 Note:  this file is  132 characters wide.
                                   43

-------
       PROGRAM ST_DELTA_HOURLY
c
C	This is an hourly version of PROGRAM ST_DELTA_DAILY.
C	Programmers   QUINN   LLM   JRB
C
C       THIS VERSION USES ROUGHNESS COEFFICIENTS CALCULATED FROM
C  . .   . '59-'77 VIA GUESSN.FOR, AND COMPARES ALL INTERMEDIATE GAGES.
C  ... IT WAS ADAPTED FROM XX2.FOR, AND THEREFORE ZYX AND SCDQMOD  TO
C       ALLOW FOR ISLANDS; ie SPARGE MATRIX AND NON-CONSEC. STATIONING.
C       FOR005) HRISLE.DAT:  PHYSICAL  DATA (STA,ABAS,DATU,&AT)
C       FOR006) THE OUTPUT FILE HRDEL.OUT
C  . .  . FOR007) SYS$OUTPUT
C  . .  . FOR010) TT
       IMPLICIT DOUBLE  PRECISION  (A-H.O-Z)
       REAL NPERC.MPERC.PERC
       LOGICAL NEWROW.MONFLAG
       INTEGER RR(358)ICC(358),ESP,PATH,FLAG
       CHARACTER*9  NAMMON(12) ,NAME(8)*20,NMM(8)*3,NOTE(8 . 24)*1 ,MARK(181)
      >*2
       COMMON /WAT001/IHOUR(24,31).MEAN(31),MEM,MAXV(31),MAXD(31) , I FLAG,
      > MINH(31).MIND(31),MAXM(4),MINM(4).IC,IGEAGE.MONAA,IYRR,IDUM(142)
       DIMENSION AA(180),ABAS(180),DATU(180),AT(180),X(180).STA(180)>
      > WS(50>180),Q(55,180),YVECT(358),XMTRX(358,358),T(180),AN(180),
      > A(180) ,U(180) ,R(180) ,QA(180) , SUM(44) ,AVE(44) ,ADJ(8) , IGAGE(8) ,
      > OLD(8) ,LOC(8) ,DEV(8) ,WSSAV(50,8) ,Al(8) ,B1(8) .NODE(IO) ,NBR(5) ,
      >IA(359),AVECT(1378),JA(1378),ICC(358),YV(358),RSP(5722),ISP(5722)
      >,VEL(180)
       EQUIVALENCE (ISP.RSP)
       DATA NAMMON/'   JANUARY','  FEBRUARY','     MARCH','     APRIL' ,
      >            •       MAY','      JUNE','      JULY','    AUGUST',
      >            ' SEPTEMBER','   OCTOBER','  NOVEMBER' ,'  DECEMBER' /
       DATA NAME/'    FT. GRATIOT    ','     DUNN PAPER
      >          'MOUTH OF BLACK RIVER' . '      DRY DOCK
      >          '     MARYSVILLE      ' , '      ST CLAIR
      >          '      ALGONAC        ','LAKE ST. CLAIR (SCS)'/
       DATA NMM/'  FG',' DP'.'MBR',' DD',' MV',' SC',' AL'.'SCS'/
       DATA LOC/180,178,160,157,155,129,77,I/
       DATA IGAGE/14098 ,14096 .14090.14087 ,14084,14080,14070 ,14052/
       DATA OLD/580.92 ,580.43,580.09,579.53,579.01,578.16,576 . 60, 576. 23/
 C    Al IS  SLOPE AND Bl IS INTERCEPT OF MANNINGS N'S FOR ALL REACHES
 c     this  is for years 1959-1977
       DATA Al/0.0033947,0.0002708,0. ,0. ,0. ,0. ,-0.0011146, -0.0017647/
        DATA  B1/-1.92253,-0.12683,0.0221,0.0250,0.0240,0.0230,
       >0.66250,1.04729/
 C
 C   .  .  ISLAND/DELTA  SPECIFIC VARIABLE ASSIGNMENTS
        DATA NODE/3,6.25,39,51,76,97,108,135,154/
        DATA NBR/15,36,45,102,144/
        EPERC1-.253
        WPERC1-1.-EPERC1
        EPERC2-.376
                                    44

-------
      WPERC2-1.-EPERC2
      NPERC-.35
      MPERC-.21
      CNPERC-.56
      FNPERO.56
      SPERC-.21
      COPERC-.23
      CSPERC-.44
      FSPERC-.44
      NSP-5722
      NISLANDS-5
      ICOUNT-0
      LRATIO-2
      IFLAG-0
C
C .  .  .  PHYSICAL DATA ACCESSED
      DO 10 I-1,LOC(1)
   10 READ(5,1020) STA(I),ABAS(I),DATU(I),AT(I)
C
C .  .  .  PROMPT FOR AND READ BEGINNING AND ENDING DATES
C      WRITE(7,2000)
      READ (10.1000) MONA,IDAYA,IYRA,MONB,IDAYB,IYRB
      IF(MONA.LE.O) THEN
         MONA-12
         IYRA-IYRA-1
      END IF
      IYRA-IYRA+1900
      IYRB-IYRB+1900
C
C .  .  .  PROMPT FOR LOCATION OF UPPER AND LOWER LIMITS TO BE RUN
C    9 WRITE(7,2020) (I,NAME(I), 1-1,8)
 9    READ(IO.IOIO) IUP.IDN
      IIDN-IDN-1
      IIUP-IUP+1
      IF(IUP.GE.IIDN) GO TO 9
C
C .  .  .  PROMPT FOR OPTION NUMBER
C  616    WRITE(7,8001)
 616     READ(10,8002)NOPT
         IF(NOPT.GT.3.0R.NOPT.LT.l) GO TO 616
C
      READ(10,8002) MUNITS
C
C .  .  .  DEFINE AND ADJUST PARAMETERS, BASED UPON LIMITS  FROM ABOVE
      NRM-LOC(IUP)-LOC(IDN)
      NMR-NRM+1
      IF(LOC(IDN).EQ.l) GO TO 27
      DO 24 I-l.NMR
      STA(I)-STA(I-1+LOC(IDN))
      ABAS (I)-ABAS(I-1+LOC(IDN))
      DATU(I)-DATU(I -l-i-LOC(IDN) )
   24 AT(I)-AT(I-1+LOC(IDN))
                                 45                                    A-ts.z

-------
     NODE(1)-NODE(1)-LOC(IDN)+1
     NODE(2)-NODE(2) -LOC(IDN)+1
     NODE(3)-NODE(3)-LOC(IDN)+1
     NODE(4)-NODE(4) -LOC(IDN)+1
     NODE(5)-NODE(5)-LOC(IDN)+1
     NODE(6)-NODE(6)-LOC(IDN)+1
     NODE(7)-NODE(7)-LOC(IDN)+1
     NODE(8)-NODE(8)-LOC(IDN)-H
     NODE(9)-NODE(9)-LOC(IDN)+1
     NODE(10)-NODE(10)-LOC(IDN)+1
     NBR(1)-NBR(1)-LOC(IDN)+1
     NBR(2)-NBR(2)-LOC(IDN)+1
     NBR(3)-NBR(3)-LOC(IDN)+1
     NBR(4)-NBR(4)-LOC(IDN)+1
     NBR(5)-NBR(5)-LOC(IDN)+1
      DO 25 I-IUP.IDN
   25 LOC(I)-LOC(I)-LOC(IDN)+1
C
C       CALCULATE DISTANCES BETWEEN SECTIONS
   27 DO 30 I-l.NRM
      X(I)-STA(I+1)-STA(I)
   30 IF(I.EQ.NBR(1).OR.I.EQ.NBR(2).OR.I.EQ.NBR(3).OR.I.EQ.
     >NBR(4).OR.I.EQ.NBR(5))  X(I)-0.0
C
C       The following lines which, have been COMMENTED with  "cO",  will
C  .  .  . WRITE the BASIC PHYSICAL DATA.   Should  the user desire  to  see
C       this  data, the  "cO's" would have  to be  eliminated and the
C       program re-compiled/linked etc.  TO RE-COMPILE submit HCDELTA.
C
cO     WRITE (6,3000)
cO     WRITE (6,3010)
cO     DO 40 I-l.NMR
 cO      MARK(I)-'   '
 cO        IF(I.EQ.NODE(3).OR.I.EQ.NODE(5).OR.I.EQ.NODE(6)
 cO     >.OR.I.EQ.NODE(7).OR.I.EQ.NODE(8).OR.I.EQ.NODE(9).OR.
 cO     >I.EQ.NODE(10))  MARK(I)-'<'
 cO        IF(I.GT.NODE(2).AND.I.LE.NBR(1))MARK(I)-'N '
 cO        IF(I.GT.NBR(1).AND.I.LT.NODE(3))MARK(I)-'M '
 C0          IF(I.GT.NODE(3).AND.I.LE.NBR(2))MARK(I)-'UN'
 cO          IF(I.GT.NODE(4).AND.I.LE.NBR(3))MARK(I)-'S  '
 cO          IF(I.GT.NBR(3).AND.I.LT.NODE(5))MARK(I)-'CO'
 cO          IF(I.GT.NODE(5).AND.I.LT.NODE(6))MARK(I)-'US'
 c0          IF(I.GT.NODE(7).AND.I.LE.NBR(4).OR.I.GT.NODE(9).AND.I
 cO     >.LE.NBR(5))MARK(I)-'W '
 c0          IF(I.GT.NBR(4).AND.I.LT.NODE(8).OR.I.GT.NBR(5).AND.I
 cO     >.LT.NODE(10))MARK(I)-'E  '
 cO      IF(I.LE.NODE(2).OR.I.GT.NBR(2).AND.I.LE.NODE(4))  GO  TO  40
 cO      WRITE(6,3020) STA(I),ABAS(I),DATU(I),AT(I),MARK(I)
 cO      DO 40  IJ-IUP.IDN
 cO   40 IF(I.EQ.LOC(IJ))  WRITE(6,3030) NAME(IJ)
 C
 C      THESE  NEXT 2 MANNINGS  N PRINTOUTS ARE HARDWIRED.   IF THE
                                    46

-------
C     NUMBER OF STATIONS CHANGES, CHANGE THESE!
C
cO      IF(IDN.EQ.S) VRITE(6,3040) Al(8),NMM(8),B1(8),STA(7),STA(36)
cO      IFUDN.EQ.8) WRITE(6.3040) Al(7) ,NMM(8) ,B1(7) ,STA(40) ,STA(77)
cO      DO 45 I-IIDN,IUP,-1
cO   45 IF(I.LT.7)WRITE(6,3040) Al(I),NMM(I),B1(I),STA(LOC(I+1) ) ,
cO     >STA(LOC(I))
C
C  .  .  . INITIALIZE AND ASSIGN  ADDITIONAL VARIABLES
       DO  50  1-1,4
    50  ADJ(I)-0.
       NVAR-NRM*2
       ANC-1.
       DT-ANC*3600.
       DO  56  1-1,42
    56  SUM(I)-0.
       TH-.75
       TH1-.25
       MM-0
       M-13
       istart-13
       iend-36
       KDK-IDAYA
       kkz-11
       MON-MONA
       IYR-IYRA
       MONFLAG-.FALSE.
       JRB-(IUP+IDN)/2
 C
 C
 C     Routine to read all of  the water  level data froa the  disk and
 C     store it in another tempory file.  This way the disk  is  not tied
 C     up for long periods of  time when  running  the program.
 C
 C  .  .  .  READ WATER LEVELS FROM DISC.
 C
   52  CALL NODAYS( IYR.MON.1.NDM.NDY,JD)
       DO 55 JJ-IUP.IDN
        IW-1
        IC-IGAGE(JJ)/10000
        IGAG-IGAGE( JJ )-IC*10000
        CALL GAGEIO(  IW.IC.IGAG ,MON, IYR.IB, IT, IDA, IDB.IDC, IER)
        IF(IER.NE.O)  THEN
           WRITE(6,3110)  NAMMON(MON),IYR,IER
           CALL EXIT
        END IF
        DO 54 J-l.NDM
        DO 53 1-1,24
         WRITE(9,1050) IHOUR(I.J)
    53  CONTINUE
        WRITE(9,1050)  HEAN(J)
    54  CONTINUE


-------
  55  CONTINUE
C
      IF(IYR-IYRB) 58,57,65
  57  IF(MON-MONB) 58,65,65
  58  CONTINUE
      IF(MON-12) 60,59,59
  59  IYR-IYR+1
  60  CONTINUE
C       UPDATE MONTH AND YEAR  AND RECHECK IF MORE DATA SHOULD BE USED
      MON-MON+1
      IF(MON-13)  62,61,61
  61  MON-1
  62  IF(IYR-IYRB)  64,63,65
  63  IF(MON-MONB)  64,64,65
  64  CONTINUE
 C
      GO TO 52
  65  CALL DISMOUNTPACK('WATER_LEVELS')
      MON-MONA
       IYR-IYRA
       REWIND 9
 C       COME HERE EACH DAY AND PRINT TITLES AND HEADINGS FOR
 C                             EACH OPTION
  70   continue
       if(Nopt.cq.1.and.iup.eq.1)  iiup-iiup+1
 C*********************OPTION  FOR WATER LEVELS********************
       IF(NOPT.EQ.l) THEN
        WRITE(6,3000)
        WRITE(6,3050) NAMMON(MON).kdk.IYR,NAME(IUP).NAME(IDN)
        WRITE(6,3060) ANC.NRM
           WRITE(6,3070) NMM(IDN).NMM(IUP),
      >    NMM(IUP),(NMM(I),I-IIDN,IIUP,-1)
           WRITE(6,3071)  (NMM(I),I-IIDN.IIUP.-1)
        IF(NOPT.EQ.LAND. IUP. EQ.DIIUP-IIUP-1
  C
  C  ******************OPTION FOR FLOWS WITHOUT DELTA*****************
  C
        ELSE IF(NOPT.EQ.2 .AND. IDN.NE.8) THEN
          WR1TE(6,3000)
          WRITE(6,8051)NAMMON(MON) ,kdk,IYR,NAME(IUP) .NAME(IDN)
          WRITE(6,3060)ANC,NRM
          WRITE(6,7020)
          WRITE(6,7030)NMM(IDN),NMM(JRB).NMM(IUP).NMM(IIDN),NMM(JRB),
       >  NMM(IUP) .NMM(JRB)
          WRITE(6,7040)
  C
  C*******************OPTION FOR  FLOWS WITH DELTA********************
  C
          ELSE IF(NOPT.EQ.2  .AND.  IDN.EQ.8) THEN
           WRITE(6,3000)
           WRITE(6,8051)NAMMON(MON) ,kdk,IYR,NAME(IUP) ,NAME(IDN)
           WRITE(6,3060)ANC,NRM
                                     48

-------
        WRITE(6,8020)
        WRITE(6,8030)NMM(IDN) ,NMM(IIDN-1) ,NMM(IUP) ,NMM(IUP) NMM(IIDN)
     >  NMM(IIDN-l)
        WRITE(6,8040)
C
C ****************QPTION TOR. VELOCITIES WITHOUT DELTA***************
C
       ELSE IF(NOPT.EQ.3 .AND. IDN.NE.8)  THEN
        WRITE(6,3000)
        WRITE(6,9051)NAMMON(MON) ,kdk, lYR.NAME(IUP) .NAME(IDN)
        WRJTE(6,3060)ANC,NRM
        WRITE(6.9025)
        WRITE(6, 7030)NMM(IDN) ,NMM(JRB) ,NMM(IUP) .NMM(IIDN) .NMM(JRB)
     >  NMM(IUP),NMM(JRB)
        WRITE(6,9041)
C
C*****************OPTION FOR VELOCITIES WITH DELTA******************
C
        ELSE IF(NOPT.EQ.3 .AND. IDN.EQ.8)  THEN
        WRITE(6.3000)
        WRITE(6,9051)NAMMON(MON) ,kdk,lYR.NAME(IUP) .NAME(IDN)
        WRITE(6,3060)ANC,NRM
        WRITE(6,9020)
        WRITE(6.9030)NMM(IDN) .NMM(IIDN-l) .NMM(IUP) .NMM(IUP) .NMM(IIDN)
     >  NMM(IIDN-l)
        WRITE(6,9040)
        END IF
  71     Continue
C
C  .  .  . ADJUST GAGE-SPECIFIC VARIABLES  BASED UPON  YEARS STUDIED
       IF(IYR.LT.1970)  IGAGE(1)-14099
       IF(IYR.LT.1971)  IGAGE(6)-14080
       IF(IYR.GT.1970)  ADJ(l)—.18
       IF(IYR.GT.1981)  ADJ(l)—.06
       IF(IYR.LT.1971)  ADJ(6)--.09
 C
 C
 C .  .  . READ WATER  LEVELS  FROM UNIT 9.  .  .
 C
       CALL NODAYS(  IYR.MON,l.NDM.NDY,JD)
 C
       DO  110 JJ-IUP.IDN
 C
       DO  77  J-l.NDM
        DO 75 1-1,24
  75     READ(9,1050,ENI>-78)  IHOUR(I.J)
  77   READ(9,1050,END-78)  MEAN(J)
 C
 C     THE BLOCK OF CODE FOR READING THE WATER  LEVELS DISC
 C     IS  DIFFERENT FROM THE DAILY VERSION OF THE  PROGRAM.
 C
  78   CONTINUE

-------
     DO 110 J-lstart,l«nd
      IF(KXZ.EQ.ll)  THEN
       I^J-12
      ELSE
       I-J-1
      END IF
C
C .  .  . FLAG MISSING DATA AND ASSIGN GAGE LEVELS TO WSSAV
      IF(  IHOUR(I.KDK).GT.O) THEN
       HOTE(JJ.KDK)-1 '
      ELSE
       NOTE(JJ,KDK)-'E'
      END IF
      if(lhour(i,kdk).le.O)  then
       wssav(j.jj)-old(jj)
      else
       wssav(j,Jj)-(ihour(i,kdk)+ib)/100.+adj(;jj)
      end if
      old(Jj)-wssav(j,JJ)
      IF(JJ.EQ.IUP.OR.JJ.EQ.IDN)  WS(J ,LOC(JJ))-WSSAV(J ,JJ)
  110 CONTINUE
      IF(MONFLAG) GO TO  200
 C
 C  .  .  . SET WS  FOR  12 PREVIOUS TIME STEPS,  TO ACHIEVE 'STEADY STATE'
      DO  120 1-1,12
       DO 120 J-1,8
        VSSAV(I.J)-VSSAV(13,J)
   120    IF(J.EQ.IUP .OR. J.EQ.IDN) US(I,LOC(J))-WSSAV(13,J)
 C
 C  .  .  .  ZERO MATRIX, SET CHANNEL PARAMETERS,  & SET INITIAL CONDITIONS.
      DO 130  I  - l.NVAR
       YVECT(I)-0.
       DO 130 J - l.NVAR
   130   XMTRX(J,I)-0.
       XSUM-STA(LOC(IUP) ) -STA(LOC(IDN) )
       SLOPE- (USSAV(1 , IUP) -WSSAV(1 , IDN) ) /XSUM
       AA(1)-ABAS(1)*AT(1)*(US(1,1)-DATU(1))
       DO 150 I-l.NRM
        IF(I.NE.NRM) THEN
         WS(1,I+1)-WS(1,I)+SLOPE*X(I)
         IF(I.EQ.NBRd)) WS(1.I+1)-WS(1,NODE(2))
         IF(I.EQ.NBR(2)) WS(1,I-H)-WS(1.SODE(1))
          IF(I.EQ.NBR(3)) WS(1,1+1)-WS(1 ,NODE(4))
          IF(I.EQ.NBR(4)) WS(1, U1)-WS(1 ,NODE(7))
          IF(I.EQ.NBR(5)) WS(1 , 1+1)-WS(1 ,NODE(9) )
         END  IF
         AA(I-i-l)- ABAS(I-»-l)+AT(I-H)*(WS(l,I+l)-DATU(I+l))
    150  R(I)-A(I)/T(I)
               Al(IIDN)*VSSAV(M,IIDN)+Bl(IIDN)
                                        50

-------
      QSTART-1.486*A(l)*R(l)**(2./3.)*(VS(l,2)-WS(1,!>>**. 5/AN(l)
c
c
     .  SPLIT FLOW INITIALLY AROUND THE  ISLANDS  FOR FASTER CONVERGENCE
     DO 190 I-l.NMR
      PERC-1.
      IF(I.GT.NODEU) .AND. I. LE.NODE(2))  PERC-FNPERC
                            I.LE.NBR(l))  PERC-NPERC
                           I.LT.NODE(3))
                            I.LE.NBR(2))
                           I.LE.NODE(4))
                           . I.LE.NBR(3))
                           I.LT.NODE(S))
                            I.LT.NODE(6))
   190
   200
   210
   230
 C
 C
 c
IF(I.
IF(I
.AND.
AND.
.AND.
AND.
.AND.
AND.
.AND.
.AND.
AND.
.AND.
AND.
                                           PERC-KPERC
                                           PERC-CNPERC
                                           PERC-FSPERC
                                           PERC-SPERC
                                           PERC-COPERC
                                            PERC-CSPERC
                              I.LE.NBR(4)) PERC-WPERC1
                             I.LT.NODE(8)) PERC-EPERC1
                              I.LE.NBR(S)) PERC-VPERC2
                             I.LT,NODE(10)) PERC-EPERC2
 IF
-------
    IU-1D+1
    DO 255 J-1,5
     JJ-NBR(J)
     KK-NODE(2*J)
     LL-NODE(2*J-1)
     IFdDN.NE.S.AND.ID.EQ.LL.OR.IDN.NE.S.AND.IU.EQ.KX.OR.
  >    IDN.EQ.8.AND.I.GT.106.AND.ID.EQ.LL.OR.IDN.EQ.8.AND.I.GT.106
  >    .AND.IU.EQ.KX.OR.IDN.EQ.8.AND.ID.EQ.6.0R.IDN.EQ.8.AND.ID.EQ.
  >    24.0R.IDN.EQ.8.AND.ID.EQ.39.0R.IDN.EQ.8.AND.ID.EQ.50.0R.
  >    IDN.EQ.8.AND.ID.EQ.3) THEN
       YVECTd)—(WS(N,IU)+WS(M,IU)-WS(N,ID)-WS(M,ID))/2.
       XMTRXd, I-D —0.5
       XMTRX(I,I-H)-+0.5
       GO TO 260
      ELSE IFUD.EQ.JJ.AND.IDN.NE.8.0R.ID.EQ.JJ.AND.IDN.EQ.8
   >   .AND.J.CE.4) THEN
       YVECTd)--(VS(N.KX)+WS(M,KK)-WS(N,JJ)-WS(M,JJ))/2.
       XMTRX(I,I-1)--0.5
       XMTRX(I,2*KK-2)-+0.5
       GO TO 260
       ELSE IF(IDN.EQ.8.AND.1D.EQ.NBR(1)) THEN
       YVECT(I)—(VS(N,ID)+WS(M,ID)-WS(N,NODE(3))-WS(M,NODE(3)))/2.
       XMTRX(I,28) - +.5
       XMTRX(I,48) - -.5
       GO  TO  260
       ELSE IF(IDN.EQ.8.AND.ID.EQ.NBR(2)) THEN
       YVECT(I)--(WS(N,ID)+WS(M,ID)-WS(N,NODE(6))-WS(K,NODE(6)))/2.
       XMTRX(I,70)-  +.5
       XMTRX(I,150)- -.5
        GO TO 260
       ELSE IF(IDN.EQ.8.AND.ID.EQ.NBR(3)) THEN
        YVECT(I)—(WS(N,ID)+WS(M,ID)-WS(N,NODE(5))-WS(M,NODE(5)))/2
        XMTRX(I,88)- +.5
        XMTRX(I.IOO)- -.5
        GO TO 260
       END IF
 255   CONTINUE
       YVECT(I)— ((WS(N,ID)+WS(N.IU)-WS(M,ID)-WS(M,IU))/(2.*DT) +
    >  (TH*(Q(N,ID)  -Q(N,IU))+TH1*(Q(M,ID)-Q(M,IU)))/(T(ID)*X(ID)))
       XMTRXd, I)-TH/(T(ID)*X(ID))
       XMTRXd , 1+2)—XMTRX(I, I)
       IF(I.EQ.l) THEN
        XMTRXd, 2 )-l./(2-*DT)
       ELSE
        XMTRXd, I-1)-1./(2.*DT)
        XMTRXd, 1+1)-!./(2. *DT)
         IF(I.EQ.NRD) XMTRX(I,I+l)-XMTRX(I,I-(-2)
         IF(I.EQ.NRD) XMTRXd,H-2)-0.
       END  IF
  260 CONTINUE
C
C .  .   . MOMENTUM EQUATIONS
                                    52

-------
 DO 280 I-2.NVAR.2
  ID-I/2
  IU-ID+1
  DO 265 J-1,5
   JJ-NBR(J)
   KK-NODE(2*J)
   LL-NODE(2*J-1)
   IF(IDN.NE.8.AND.ID.EQ.LL.OR.IDN.EQ.8.AND J GE 4 AND
>  ID.EQ.LL) THEN                           '  '   '

    YVECT(I)-(Q(N,JJ+1)^5(M,JJ+1)4<3(N>IU)4<5(M,IU)-Q(N,ID).Q(M.ID))
    XMTRXd.I-l) — 0.5
    XMTRX(I,I+1)-+0.5
    XMTRX(I,2*JJ+1)-+0.5
    GO TO 280
   ELSE IF(IDN.NE.8.AND.ID.EQ.JJ.OR.IDN.EQ.8.AND J GE 4 AND
>  ID.EQ.JJ) THEN
    YVECT(I)-(VS(N,IU)+WS(M,IU)-WS(N,LL)-VS(M LL))/2
    XMTRX(I,2*LL-2)--0.5
    XMTRX(I.I)-+0.5
    GO TO 280
   ELSE  IF(IDN.NE.8.AND.IU.EQ.KK.OR.IDN.EQ.8.AND J GE 4 AND
>  IU.EQ.KK) THEN
    YVECT(I)-(Q(N,KK)4.5
    GO TO 280
    ELSE IF(IDN.EQ.8.AND.ID.EQ.NODE(1)) THEN
    YVECT(I)-  -(Q(N,NODE(1))-K3(M,NODE(1))-Q(N,NODE(1)-H)
 >   -Q(M,NODE(l)+l)-Q(N,NBR(2)-H)-Q(M,NBR(2)+l))/2.
    XMTRX(I,I-1)-  +.5
    XMTRX(I,I+1)-  -.5
    XMTRX(I,73)- -.5
    GO TO 280
    ELSE IF(IDN.EQ.8.AND.ID.EQ.NODE(2)) THEN
    YVECT(I)-(Q(N,NODE(2))-K5(M,NODE(2))-Q(N,NODE(2)+1)
 >   -Q(M,NODE(2)+l)-Q(N,NBR(l)-H).Q(M,NBR(l)-H))/2.
    XMTRX(I,I-1)- +.5
     XMTRX(I.I-H)- -.5
     XMTRX(I,31)- -.5
     GO TO  280
    ELSE IF(IDN.EQ.8.AND.ID.EQ.NODE(4)) THEN
     YVECT(I)-(Q(N,NODE(4))+Q(M.NODE(4))-Q(N,NODE(4)-H)
 >   -Q(M,NODE(4)+l)-Q(N.NBR(3)+l)-Q(M,NBR(3)+l))/2.
     XMTRX(I.I-l)- +.5
     XMTRX(I,I+1)- -.5
     XMTRX(I,91)- -.5
     GO TO  280
    ELSE IF(IDN.EQ.8.AND.ID.EQ.NBR(1)) THEN
     YVECT(I)—(WS(N,NBR(1)+1)+WS(M.NBR(1)+1)-WS(N,NODE(2))
 >   -WS(M,NODE(2)))/2.
                               53

-------
    XMTRX(I.I)- +.5
    XMTRX(I.IO)-  -.5
    GO TO 280
    ELSE IF(IDN.EQ.8.AND.IU.EQ.NODE(3)) THEN
    YVECT(I)— (Q(N,NODE(3»+Q(M,NODE(3))-Q(N,NODE(3)-1)
 >   -Q(M.NODE(3)-l)-Q(N,NBR(l))-Q(M,NBR(l)))/2.
    XMTRX(I,I+D- +.5
     XMTRX(I.I-l)- -.5
     XMTRX(I,29)- -.5
     GO TO 280
    ELSE IF(IDN.EQ.8.AND.ID.EQ.NBR(2)) THEN
     YVECT(I)— (WS(N.NODE(1))+WS(M,NODE(1))-WS(N,NBR(2)+1)
 >   -WS(M,NBR(2)+l))/2.
     XMTRX(I,72)- -.5
     XMTRX(I,4)- +.5
     GO TO 280
    ELSE  IF(IDN.EQ.8.AND.ID.EQ.NBR(3)) THEN
     YVECT(I)— (WS(N,NODE(4))+WS(M,NODE(4))-WS(N,NBR(3)+1)
  >    -WS(M,NBR(3)+l))/2.
     XMTRX(I,76)- +.5
     XMTRXU.90)-  -.5
      GO TO 280
     ELSE IF(IDN.EQ.8.AND.IU.EQ.NODE(5))  THEN
      YVECT(I)—    -Q(M,NODE(5)-l)-Q(N,NBR(3))-Q(M.NBR(3)))/2.
      XMTRX(I.lOl)- +.5
      XHTRX(I,99)- -.5
      XMTRX(I,89)- -.5
      GO TO 280
     ELSE IF(IDN.EQ.8.AND.1U.EQ.NODE(6)) THEN
      YVECT(I)—(Q(N,NODE(6))-K)(M,NODE(6))-Q(N,NODE(6)-1)
  >   -Q(M,NODE(6)-l)-Q(N,NBR(2))-Q(M,NBR(2)))/2.
       XMTRX(I,15D-  *-5
       XMTRX(I,71)-  -.5
       XMTRX(I,U9)-  -.5
       GO TO  280
      END IF
265   CONTINUE
     Z41—QA(ID)*T(ID)*(WS(N,IU)+WS(N,ID)-WS(M,IU)-WS(M,ID))/
   >  (2.*DT*A(ID)**2.)*2.
     Z61-(Q(N,ID)-KKN,IU)-Q(M.ID)-Q(M,IU))/(2.*DT*A(ID))
     Zll-32 17*AN(ID)**2.*QA(ID)*U(ID)/(2.21*A(ID)**2.*R(ID)**(A./3.))
     Z21-32.17*(TH*(WS(N,ID)-WS(N,IU))+TH1*(WS(M,ID)-WS(M,IU)))/X(ID)
     Z31—(QA(ID)**2.*(AA(ID)-AA(IU))/(A(ID)**3.*X(ID)»
     YVECT(I)— (Z1UZ21+Z31+Z4UZ61)
     XMTRX(I.I)—(32.2-QA(ID)**2.*T(ID)/(A(ID)**3.))/X(ID)*TH-QA(ID)
    >    *T(ID)/(DT*A(ID)**2.)
     IF(I.LT.A) GO TO 270
     XMTRX(I,I-2)-(32.2-QA(ID)**2.*T(ID)/(A(ID)**3.))A(ID)*TH-QA(ID)
    >    *T(ID)/(DT*A(ID)**2.)
 270 ZZZ1-ABS(QA(ID))
     Pl-32.2*AN(ID)**2.*ZZZl*TH/(2.2082*A(ID)**2.*R(ID)**(4./3-))
                                                                       - AT.//
                                54

-------
        IF (YV(I).LT.-20.) YV(I)—20.
        II-I/2-H
        WS(N,II)-WS(N,II)+YV(I)
  310 CONTINUE
C
C .  .  .  ASSIGN MAXIMUM DELTA VALUE, AND TEST  FOR EXCESSIVE ITERATIONS
      DO 319 I-2.NNR.2
  319  IF(ABS(YV(I)).GT.YVMAX) YVMAX-YV(I)
      DO 320 I-2.NNR.2
        IF(ABS(YV(I)).GT..002) THEN
          IFOTER.GT.20)  THEN
            WRITE(6,1133)  ITER.YVMAX.I/2
 1133       FORMATC  ',12,'  ITERATIONS.   YVMAX -',E10.3,'AT I-',12,/)
            GO  TO  500
          END IF
          ITER-ITER+1
          GO TO 210
         END IF
   320 CONTINUE
 C
 C  .  .  .  DETERMINE THE DEVIATION OF CALCULATED FROM MEASURED LEVELS
      JB-N+1
      DO 330  IJ-IIUP.IIDN
       IF(WSSAV(N.IJ).LT.50.) WSSAV(N,IJ)-WS(N,LOC(IJ))-.0001
   330 DEV(IJ)-WS(N,LOC(IJ))-WSSAV(N,IJ)
       MM-MM+1
       NM-MM-kkz
       IF(NM.LE.O) GO TO 340
       IF(NOPT.EQ.3) THEN
        DO 6262 I-l.NRM+1
  6262  VEL(I)-Q(N,I)/AA(I)
       END IF

 C ,  .  . METRIC OPTION FOR OUTPUT  . .
        IF(MUNITS.EQ.O)THEN
         DO 6700 MC - IUP.IDN
  6700   WSSAV(N.MC)  - WSSAV(N,MC)/3.28083
          DO 6720 MC - IIUP.IIDN
          WS(N,LOC(MC)) - WS(N,LOC(MC))/3.28083
  6720    DEV(MC) - DEV(MC)/3.28083
 C
          Q(N,LOC(IUP)) -  Q(N,LOC(IUP))*0.02832
          Q(N,LOC(IIDN)) - Q(N,LOC(IIDN))*0.02832
          Q(N,NBR(5)+5) -  Q(N,NBR(5)-i-5)*0.02832
          Q(N.NBR(5)-4) -  Q(N,NBR(5)-4)*0.02832
          Q(N,NODE(10)+1)  - Q(N,NODE(10)+1)*0.02832
          Q(N,NBR(4)+3) -  Q(N,NBR(4)+3)*0.02832
          Q(N,NBR(4)-2) -  Q(N,NBR(4)-2)*0.02832
          Q(N,NODE(8)+1) - Q(N,NODE(8)+1)*0.02832
          Q(N.NBR(l)-3) -  Q(N,NBR(1)-3)*0.02832
          Q(N,NBR(l)+6) -  Q(N,NBR(1)46)*0.02832
          Q(N,NBR(3)-3)  -  Q(N,NBR(3)-3)*0.02832


-------
        Q(N,NBR(3)+3) - Q(N,NBR(3)+3)*0.02832
        Q(N,LOC(JRB)) - Q(N,LOC(JRB))*0.02832
C
        VEL(LOCUUP)) - VEL(LOC(IUP))/3.28083
        VEL(LOC(IIDN)) -  VEL(LOC(IIDN))/3.28083
        VEL(NBR(5)+5) - VEL(NBR(5)+5)/3.28083
        VEL(NBR(5)-4) - VEL(NBR(5)-4)/3.28083
        VEL(NBR(4)+3) - VEL(NBR(4)+3)/3.28083
        VEL(NBR(4)-2) - VEL(NBR(4)-2)/3.28083
        VEL(NBR(l)-3) - VEL(NBR(l)-3)/3.28083
        VEL(NBR(l)+6) - VEL(NBR(l)+6)/3.28083
        VEL(NBR(3)-3) - VEL(NBR(3)-3)/3.28083
        VEL(NBR(3)+3) - VEL(NBR(3)+3)/3.28083
        VEL(LOC(JRB)) - VEL(LOC(JRB))/3.28083
       END IF
 C
 C
 C  .  .  .  PRINT OUTPUT.
         if(NOPT. EQ.1.and.lup.eq.1) iiup-iiup+1
 C*********************QPTION FOR WATER LEVELS********************
 C
       IF(NOPT.EQ.l)  THEN
        WRITE(6,3120) nm,WSSAV(N,IDN) ,NOTE(IDN,KDK) ,
      >  (WS(N,LOC(I)).I-IIDN,IIUP,-1)
        WRITE(6,3121) WSSAV(N,IUP),NOTE(IUP,KDK),
      >  Q(N.LOC(IUP)),
      >  (WSSAV(N,I).NOTE(I,KDK),DEV(I), I-IIDN, HUP,-1)
 C
 C ****************QPTION FOR FLOW WITHOUT DELTA******************
 C
       ELSE IF(NOPT.EQ.2.AND.IDN.NE.8) THEN
       WRITE(6,7050)NM,WSSAV(N,IDN),NOTE(IDN,KDK),WS(N,LOC(JRB)),
      >WSSAV(N,IUP),NOTE(IUP,KDK),Q(N,LOC(IIDN)),Q(N,LOC(JRB)),
      >Q(N,LOC(IUP)),Q(N,NBR(5)-t-5)IQ(N1NBR(5)-4),Q(N,NBR(4)+3),
      X3(N,NBR(4)-2),WSSAV(N,JRB),NOTE(JRB>KDK),DEV(JRB)
 C
 C *******************OPTION FOR FLOW WITH DELTA*****************
 C
       ELSE IF(NOPT.EQ.2  .AND. IDN.EQ.8) THEN
        WRITE(6,8050)NM,WSSAV(NfIDN),NOTE(IDN,KDK)>WS(N,LOC(IIDN-l)),
      >WSSAV(N,IUP).NOTE(IUP.KDK),Q(N.LOC(IUP)),Q(N.LOC(HDN)).
      >Q(N,NBR(5)-«-5),Q(N,NBR(5)-4),Q(N>NBR(4)+3),Q(N,NBR(4)-2),
      >Q(N,NBR(l)-3) ,Q(N,NBR(l)+6) ,Q(N,NBR(3)-3) ,Q(N,NBR(3)+3) .
      >WSSAV(N,IIDN-1),NOTE(IIDN-1.KDK),DEV(IIDN-1)
 C
 C******************OPTION FOR VELOCITIES WITH DELTA*************
 C
        ELSE  IF(NOPT.EQ.3  .AND. IDN.EQ.8) THEN
 C      WRITE(*,*)'NM- '.NM
        WRITE(6,9050) NM,WSSAV(N,IDN) ,NOTE(IDN,KDK),WS(N,LOC(IIDN-1))
       >WSSAV(N,IUP) ,NOTE(IUP,KDK) ,VEL(LOC(IUP)) ,VEL(LOC(IIDN)) ,
       >VEL(NBR(5)+5) ,VEL(NBR(5) -4) ,VEL(NBR(4)+3) ,VEL(NBR(4) -2) ,
                                    57

-------
     >VEL(NBR(1)0),VEL(NBR(1)^),VEL(NBR(3).3),VEL(NBR(3)-K3)
     >WSSAVWSSAV(N,IUP),NOTE(IUP.KDK),VEL(LOC(IIDN)),VEL(LOC(JRB))
     >VEL(LOC(IUP)),VEL(NBR(5)+5) ,VEL(NBR(5)-4) ,VEL(NBR(4)+3) '
     >VEL(NBR(4)-2),WSSAV(N,JRB),NOTE(JRB,KDK),DEV(JRB)
W
      END IF
C
         if(Nopt.eq.l.and.iup.eq.l)  iiup-iiup-l
 331   continue
C
C  .  .  .  SHOW  HOUR ON TERMINAL SCREEN AND COMPUTE MEAN VALUES
      WRITE(7,2030)MON,KDK,m-1900,NM
      SUM(1)-SUM(1)+WSSAV(N,IUP)
      DO 334  1-2,7
   334 SUM(I)-SUM(I)+WS(N,LOC(I))
      SUM(8)-SUM(8)+WSSAV(N,IDN)
      DO 335  1-1,6
   335 SUM(I+8)-SUM(I+8)+WSSAV(N,Ul)
      DO 336  1-1,6
   336 SUM(H-14)-SUM(I+14)+DEV(I+1)
C
C
C  .  .  .  THESE ARE THE CALCULATIONS  FOR FLOW AVERAGES
      SUM(21)-SUM(21)-KKN,LOC(IUP))
      SUM(22)-SUM(22)+Q(N,NBR(5)+5)
      SUM(23)-SUM(23)-KKN,NBR(5) -4)
      SUM(24)-SUM(24)-K)(N,NODE(10)+1)
      SUM(25)-SUM(25)-KXN,NBR(4)+3)
      SUM(26)-SUM(26)-K}(N,NBR(4).2)
      SUM(27)-SUM(27)-H5(N,NODE(8)+1)
      SUM(28)-SUM(28)-K5(N,LOC(IIDN))
      SUM(29)-SUM(29)-KJ(N,NBR(l)-3)
      SUM(30)-SUM(30)-KKN,NBR(l)+6)
      SUM(31)-SUM(31)-Kj(N,NBR(3)-3)
      SUM(32)-SUM(32)-K3(N,NBR(3)+3)
      SUM(43)-SUM(43)-K3(N,LOC(JRB))
      SUM(33)-SUM(33)-«-VEL(LOC(IUP))
      SUM(34)-SUM(34)+VEL(LOC(IIDN))
       SUM(35)-SUM(35)+VEL(NBR(5)-»-5)
       SUM(36)-SUM(36)+VEL(NBR(5)-4)
       SUM(37)-SUM(37)+VEL(NBR(4)+3)
       SUM(38)-SUM(38)-KVEL(NBR(4)-2)
       SUM(39)-SUM(39)-»-VEL(NBR(l)-3)
       SUM(40)-SUM(40)^-VEL(NBR(l)+6)
       SUM(41)-SUM(41)-fVEL(NBR(3)-3)
       SUM(42)-SUM(42)+VEL(NBR(3)+3)
58
                                                                       A 'IS.

-------
c
c
 6730
 6740
     SUM(44)-SUM(44)+VEL(LOC(JRB) )

     .  CONVERT  FROM METRIC BACK TO ENGLISH UNITS
     IF(MUNITS.EQ.O)THEN
       DO 6730  MC - IU?,IDN
       VSSAV(N.MC) - WSSAV(N,MC)*3.28083
       DO 6740  MC - IIUP.IIDN
        WS(N,LOC(MC)) - WS(N.LOG(MC))*3.28083
       DEV(MC)  - DEV(MC)*3.28083

       Q(N,LOC(IUP)) - Q(N,LOC(IUP))/0.02832
       Q(N,LOC(IIDN)) - Q(N,LOC(IIDN))/0.02832
       Q(N.NBR(5)+5) - Q(N,NBR(5)+5)/0.02832
       Q(N,NBR(5)-4) - Q(N,NBR(5)-4)/0.02832
       Q(N.NODE(10)+1) - Q(N,NODE(10)-H)/0.02832
       Q(N,NBR(4)+3) - Q(N.NBR(4)+3)/0.02832
       Q(N,NBR(4)-2) - Q(N,NBR(4)-2)/0.02832
       Q(N,NODE(8)+1) - Q(N,NODE(8)+1)/0.02832
 C
 C
         Q(N,NBR(l)-3)
         Q(N,NBR(l)+6)
         Q(N,NBR(3)-3)
         Q(N,NBR(3)+3)
         Q(N,LOC(JRB))
                       Q(N,NBR(l)-3)/0.02832
                       Q(N,NBR(l)+6)/0.02832
                       Q(N,NBR(3)-3)/0.02832
                       Q(N,NBR(3)+3)/0.02832
                       Q(N.LOC(JRB))/0.02832
         VEL(LOC(IUP))  - VEL(LOC(IUP))*.28083
         VEL(LOC(IIDN)) - VEL(LOC(IIDN))*3.28083
         VEL(NBR(5)+5)  - VEL(NBR(5)+5)*3.28083
         VEL(NBR(5)-4)
         VEL(NBR(4)^-3)
         VEL(NBR(4)-2)
         VEL(NBR(l)-3)
         VEL(NBR(l)+6)
         VEL(NBR(3)-3)
         VEL(NBR(3)+3)
         VEL(LOC(JRB))
       END IF
                        VEL(NBR(5)-4)*3.28083
                        VEL(NBR(4)+3)*3.28083
                        VEL(NBR(4)-2)*3.28083
                        VEL(NBR(l)-3)*3.28083
                        VEL(NBR(l)+6)*3.28083
                        VEL(NBR(3)-3)*3.28083
                        VEL(NBR(3)+3)*3.28083
                        VEL(LOC(JRB))*3.28083
   340 DO 350 I-l.NMR
   350 Q(JB,I)-2.*Q(N,I)-Q(M,I)
       DO 360 1-2,NRM
   360 WS(JB,I)-2.*WS(N,I)-WS(M,I)
       write(7,2030)mon,KDK,iyr-1900,NM
       M-M+1

       IF(M-kb)2333,2333,370
  2333 continue

   .  . . HOURLY RETURN LOOP
       IF(nM-24.Lt.O) GO TO 200
   370 DO 380 1-1,44
   380 AVE(I)-SUM(I)/NM
C
C
                                 59

-------
c
C .  .  .  PRINT DAILY MEAN VALUES
        if (Nopt.eq.1.and.iup.eq.1) iiup-llup+1
C
C*********************OPTION FOR WATER LEVELS*********************
C
      IF(NOPT.EQ.l) THEN
        WRITE(6,3130) AVE(8) , (AVE(I) ,1-1IDH, IIUP, -1)
        WRITE(6,3131) AVE(1),AVE(21),
     >    (AVE(1+7),AVE<1+13),  I-IIDN, IIUP,-1)
        if(Nopt.eq.1.and.iup.eq.1)  iiup-iiup-1
C
C ******************OPTION FOR FLOW WITHOUT DELTA*****************
C
      ELSE  IF(NOPT.EQ.2.AND.IDN.NE.8)  THEN
        WRITE(6,7060)AVE(8),AVE(3),AVE(1),AVE(28).AVE(43),AVE(21),
      >AVE(22) ,AVE(23) ,AVE(25) ,AVE(26) ,AVE(10) ,AVE(JRB+13)
 C
 C********************OPTION FOR FLOW WITH DELTA*******************
 C
       ELSE IF(NOPT.EQ.2 .AND.  IDN.EQ.8) THEN
        WRITE(6 ,8060)AVE(8) .AVE(IIDN-l) ,AVE(1) ,AVE(21) ,AVE(28) ,AVE(22) ,
      >AVE(23) ,AVE(25) ,AVE(26) ,AVE(29) ,AVE(30) ,AVE(31) ,AVE(32) ,
      >AVE(IIDN+6),AVE(IIDN+12)
 C
 C******************OPTION  FOR VELOCITIES WITH DELTA***************
 C
        ELSE  IF(NOPT.EQ.3  .AND.  IDN.EQ.8) THEN
        WRITE(6,9060) AVE(8).AVE(IIDN-l),AVE(1) ,AVE(33) ,AVE(34),AVE(35) ,
      >AVE(36) ,AVE(37) ,AVE(38) ,AVE(39) ,AVE(40) ,AVE(41) ,AVE(42) ,
      >AVE(IIDN+6) ,AVE(IIDN+12)
 C
 C  *****************OPTION FOR VELOCITIES  WITHOUT DELTA************
 C
        ELSE IF(NOPT.EQ.3  .AND.  IDN.NE.8)  THEN
        WRITE(6,9065)AVE(8) ,AVE(10) ,AVE(1) ,AVE(34) ,AVE(44) ,AVE(33) ,
       >AVE(35) ,AVE(36) ,AVE(37) ,AVE(38) ,AVE(3) ,AVE(JRB+13)
  C
        END IF
  C
   381  continue
        MONFLAG-.TRUE.
        IF(IYR-IYRB) 388,386,500
    386 IF(MON-MONB) 388,387.500
    387 IF(KDK-IDAYB) 388,500,500
    388 CONTINUE
        if(kdk-ndm)390,333,333
    333 IF(MON-12) 390,389,389
    389 IYR-IYR+1
        mon-0
    390 CONTINUE
        DO  392 1-1,42
                                    60

-------
  392 SUM(I)-0.
     DO 400  I-l.NMR
  400 Q(2,I)- 'BASIC  DATA'//, 23X, 'STATION', 5X, 'ABASE', 5X, 'DATUM', 5X, 'WIDTH',/)
  3020  FORMAT(20X,F10.0,F10.0,F9.2,F8.0,1X,A2)
  3030  FORMATC '+',62X,A20)
  3040  FORMAT (/,12X,'  MANNING n -'.F11.8,' * WS@',A3,' +'
      >  F10 7,'  FOR STATIONS' ,F7.0,'  THRU '.F7.0)
   3050  FORMAT(///45X, ' ST. CLAIR RIVER HOURLY TRANSIENT MODEL' ,//,52X,A9,
       >  Ix i2,',',i5,/,10X,A20,'  to  '.A20.40X, 'WATER LEVELS  VERSION',//)
   3060 FORMAT (36X.F5.1, IX, 'HOUR TIME  INCREMENTS' ,11X, 13, LX, 'REACHES' ,//)
   3070 FORMAT(6X,A3,4X,'| ........ COMPUTED LEVELS ......... | ' ,2X,A3,5X,A3,
       > 4X,'| ........ MEASURED LEVELS AND COMPUTED DEVIATIONS  (C-M) ......
                  'HR' ,3X,'MEAS.' ,5(4X,A3))
                                        61

-------
 3071 FORMAT C + M50,' MEAS.  COMP. Q' .IX, 4(IX, ' | ----', A3 ,'--- 1' ) ,1X,
    >M. ___ •  A3, '---I',/)
 3110 FORMAT(ix,A9,lX,I4,lX/ ERROR OF TYPE ',12)
 3120 FORMAT(1X,I2,F8.2,A1,5(F7.2))
 3121 FORMATC + ',T48,F7.2,A1,F8.0,5(F7.2,A1,1X,F4.2))
 3130 FORMAT (/, IX, 'AVE ' ,F6 .2,1X,5(F7.2))
 3131 FORMATC + ',T48,F7.2,F8.0,1X,5(F7.2,2X,F4.2))
 6190  FORMAT ('  FLAG-', 15,'  ESP-', 15,'  PATH-' ,15)
C7000  FORMAT (/' ENTER OPTION NUMBER'/' 1. OUTPUT SHOWS WATER LEVELS
C    > AND DEVIATIONS'/'  2.  OUTPUT SHOWS FLOWS AROUND ISLANDS')
 7010 FORMAT(Il)
C8001 FORMATC Enter option for  delta  output:'//' 1. Output shows  Water
C    >Levels  and  Deviations'/1 2.  Output shows  flows around delta  and  i
C    >slands'/' 3. Output shows  velocities  around delta and islands'/)
 8002 FORMAT(Il)
 8051 FORMAT (///,45X/ ST. CLAIR RIVER HOURLY TRANSIENT MODEL' ,//,52X,A9 ,
     >lx,i2,',',i5,//,10X,A20/  to '.A20.40X, 'RIVER DISCHARGE  VERSION1,/
         '
  8020  FORMAT (//,5X/ | ---- RIVER PROFILE ..... | ' ,2X, ' j --TOTAL FLOW--1',3X,
     >'| .......... ISLAND FLOWS .......... |'.3X,'| ........ DELTA  FLOWS .....
     > ---- |'2X,'|---DEV ---- |',//)
  8030  FORMATC  HR' ,4X,2(A3 , 5X) ,A3,6X,A3,6X,A3,6X, 'STAG E',3X,'STAG W'
     > .4X/FAWN E',3X,'FAWN W' ,3X, 'N.CH. ' , 3X, 'M.CH. ' ,3X, 'S .CH. ' , 3X,
     >'CUTOFF',3X,A3,4X,'DEV')
  8040  FORMAT(6X, 'MEAS. ' ,4X, 'COMP. '  ,3X, 'MEAS. ' ,4X, 'FLOW' , 5X, 'FLOW ,8X, 'Q
     >' ,8X, 'Q' ,8X, 'Q'  ,8X, 'Q' ,8X, 'Q' ,7X, 'Q' ,7X. 'Q' ,8X,'Q' ,5X, 'MEAS. • ,2X,
     >'C-M',/)
  8050  FORMAT(1X,I2,F8.2,A1,1X,F7.2,F8.2,A1,1X,F8.0,U,F8.0,2X,F8.0,2X,
      >F7.0,1X,F8.0.2X,F8.0,1X,F7.0,1X,F7.0,1X,F7.0,1X,F7.0,
      >1X,F7.2,A1,F5.2)
  8060 FORMAT(/,1X,'AVE',F7.2.2X,F7.2,F8.2,2X,F8.011X,F8.0,2X,F8.0,2X>
      >F7. 0,1X^8.0,2X^8.0,1X^7.0,1X^7.0,1X^7.0,11^7.0,1X^7. 2, IX,
      >F5.2)
  9020  FORMAT (//,5X,' | ---- RIVER PROFILE ---- | ' , 3X, ' | --TOT.  VEL.--|',4X,
      >'| ...... MID ISLAND VELOCITIES ..... | ' ,3X, ' | ---MID DELTA VELOCITIES -
             ,   ,
   9030  FORMATC HR' ,4X,2(A3 ,5X) ,A3,6X,A3 ,6X,A3,6X, 'STAG E',3X,'STAG W
      > ,4X,'FAWN  E',3X,'FAWN W' ,3X, 'N.CH. ' ,3X, 'M.CH. ' ,3X, 'S.CH. ' ,2X,
      >'CUTOFF',3X,A3,5X,'DEV')
   9040  FORMAT(6X, 'MEAS. ' ,4X, 'COMP. ' ,3X, 'MEAS. ' ,4X, 'VEL. ' ,5X, 'VEL. ' ,8X,
      >'V ,8X, 'V ,8X, 'V , 8X, 'V ,8X, 'V ,7X, 'V ,7X, 'V ,7X, 'V ,5X, 'MEAS. ' .
          ,,
   9050 FORMAT(1X>I2,F8.21A1,1X,F7.2,F8.2,A1,1X,F6.2,3X,F6.2,4X,F6.2>4X,
       >F5.2.4X.F6.2,3X,F6.2f3X,F5.2,3X.F5.2.3X.F5.2,3X.F5.2.2X,F7.2.Al.
       >1X,F5.2)
   9051 FORMAT(///, 4 5X,' ST. CLAIR RIVER HOURLY TRANSIENT MODEL' ,//, 52X.A9 ,
       >lx,i2,',',i5,//,10X,A20,' TO ' .A20.40X, 'RIVER VELOCITIES VERSION'

   9060>FORHAT(/,f AVE' ,F7. 2.2X,F7.2,F8.2,2X,F6.2,3X,F6.2 .4X.F6.2.4X,
       >F5.2,4X.F6.2,3X,F6.2,3X,F5.2,3X,F5.2,3X,F5.2,3X,F5.2,2X,F7.2,2X,
       >F5.2)
   7020 FORMAT (//,17X,' | ..... RIVER  PROFILE ..... |',4X,'|---  TOTAL DISCHA
                                     62

-------
    >RGE ---I  |	ISLAND FLOWS	|  |	DEV	!',/)
7030 FORMAT(8X,'DAY',7X,A3,8X,A3,8X,A3,8X,A3,6X,A3,6X,A3,3X,
    >'STAG E',2X,'STAC V',2X,'FAWN E',2X,'FAWN V,4X,A3,7X,'DEV')
7040 FORMAT(ISX.'MEAS.1,6X,'COMP.',6X,'MEAS. • ,6X,2('FLOW',5X),'FLOW',3X
    >,4('FLOW'14X),lX,'KEAS.'f4X,'(C-M)',/)
7050 FORMAT(9X,I2,6X,F6.2,A1,4X,F6.2,5X,F6.2,A1,1X,3(2X,F7.0),
    >4(1X,F7.0)>2X,F6.2,A1,2X,F6.2)
7060 FORMAT(/,8X,'AVE',6X,F6.2,5X,F6.2,5X,F6.2.2X,3(2X,F7.0),4(1X,F7.0)
    >.2X,F6.2,4X,F5.2)
9025 FORMAT (//,17X, '|	RIVER   PROFILE	|',5X,'|--  TOTAL VELOC IT
    >IES  --I  I--  MID ISLAND VELOCITIES ---|   |	DEV	1'/)
9041 FORMAT (ISX.'MEAS.1 .6X/COMP.' ,6X,'MEAS.' , 2X, 3(5X,'VEL. ' ) ,2X,
    >'VEL.',3(4X,'VEL.'),5X,'MEAS.',4X,'(C-M)f,/)
9055 FORMAT(9X,I2,6X,F6.2,A1,3X,F6.2,6X,F6.2,A1I3(5X,F4.2),3X,F4.2,
    >3(4X,F4.2),4X,F6.2)A1,2X,F6.2)
9065 FORMAT(/,8X,'AVE'.6X(F6.2,4X,F6.2,6X,F6.2,IX,3(5X,F4.2)I3XIF4.2,
    >3(4X,F4.2),4X.F6.2,4X,F5.2)
C
C
      END
                                   63

-------
                  Program [HYDRO.JDSTCLR]DDELTA.FOR

        This is the St Clair River Transient Model - Daily Version.

        It is set to run in BATCH MODE...
        To run the program...

1.  Set desired parameters in file [HYDRO.JDSTCLR]DDELTA.PAR

        Line 1 - Starting and ending month and yr.
                 MO YR MO YR  (12 , IX, 12 ,IX, 12 , IX, 12)

        Line 2 - Staring and ending points of model. (II,IX,II)
                 1 - Fort Gratiot
                 2 - Dunn Paper
                 3 - Mouth of Black River
                 4 - Dry Dock
                 5 - Marysville
                 6 - St Clair
                 7 - Algonac
                 8 - Lake St Clair

        Line 3 - Output Option (II)
                 1 - Water levels and deviations.
                 2 - Total discharge and discharge around  islands and, if
                     included, discharge in the delta channels.
                 3 - Velocity near the starting,  ending and midpoint of
                     the simulated river and velocities around islands and,
                     if included, velocities in the delta  channels.

        Line 4 - Units Option (II)
                 0 - Metric units
                 1 - English units

2.  Make sure that file [HYDRO.JDSTCLRJDDELTA.DAT  is available.

3.  Type: SUBMIT DDELTA/NOTIFY

4.  When your request is completed the output will  appear in file:
                 [ HYDRO. JDSTCLR ] ZDDELTA. OUT
                 Note: this file is 132 characters wide.
                                 64

-------
     >IA(359),AVECT(1378),JA(1378),ICC(358),YV(358),RSP<5722),ISP(5722)
     >,VEL(180)
      EQUIVALENCE  (ISP.RSP)
      DATA NAMMON/'   JANUARY' ,'  FEBRUARY' , '    MARCH','     APRIL',
     >            '       MAY' , '      JUNE' , '     JULY' ,'    AUGUST' ',
     >            'SEPTEMBER','   OCTOBER',' NOVEMBER','  DECEMBER'/
      DATA NAME/'     FT.  GRATIOT    ','      DUNN PAPER
     >          'MOUTH OF BLACK RIVER' . '       DRY DOCK       ' ,
     >          '      MARYSVILLE     ','       ST CLAIR       '\
     >          '       ALGONAC       ','LAKE ST. CLAIR (SCS)'/
      DATA NMM/'  FG',' DP','MBR','  DD','  MV',' SC', '  AL'.'SCS'/
      DATA LOC/180,178,160,157,155,129,77,1/
      DATA IGAGE/14098,14096,14090,14087,14084,14080,14070,14052/
      DATA OLD/580.92,580.43.580.09,579.53,579.01,578.16,576.60,576.23/
C     Al IS SLOPE AND Bl  IS INTERCEPT OF MANNINGS N'S FOR ALL REACHES
      DATA Al/0.0033947,0.0002708,0.,0.,0.,0.,-0.0011146,-0.0017647/
      DATA B1/-1.92253,-0.12683,.0221,.0250,.0240,.0230,
     >0.66250,1.04729/
C
C .   . .  ISLAND/DELTA SPECIFIC VARIABLE ASSIGNMENTS
C
      DATA NODE/3,6,25,39,51,76.97.108,135,154/
      DATA NBR/15,36,45,102,144/
C
      EPERC1-.253
      WPERC1-1.-EPERC1
      EPERC2-.376
      WPERC2-1.-EPERC2
      NPERC-.35
      MPERC-.21
      CNPERC-.56
      FNPERC-.56
      SPERC-.21
      COPERC-.23
      CSPERC-.44
      FSPERC-.44
      NSP-5722
      ICOUNT-0
      LRATIO-2
      IFLAG-0
C
C .   . . PHYSICAL DATA ACCESSED
C
      DO  10 I-l.LOC(l)
   10 READ(5.1020) STA(I),ABAS(I),DATU(I),AT(I)
C
C .   . . PROMPT FOR AND READ BEGINNING AND ENDING DATES
C
C       WRITE(7,2000)
      READ (10,1000) MONA.IYRA.MONB.IYRB
      IF(MONA.LE.O) THEN
          MONA-12
                                 66

-------
         IYRA-IYRA-1
      END IF
      IYRA-IYRA+1900
      IYRB-IYRB+1900
C
C .  .  .  PROMPT FOR LOCATION OF UPPER AND LOWER LIMITS TO  BE  RUN
C
C    9 WRITE(7,2020)  (I,NAME(I), 1-1,8)
9     READ(IO.IOIO) IUP.IDN
      IIDN-IDN-1
      IIUP-IUP+1
      IF(IUP.GE.IIDN) GO TO 9
C
C .  . . PROMPT FOR OPTION NUMBER
C
C  616    WRITE(7,8001)
  616    READ(10,8002)NOPT
         IF(NOPT.GT.3.0R.NOPT.LT.l) GO TO  616
C
        READ(10,8005) MUNITS
C
C .  . . DEFINE AND ADJUST PARAMETERS, BASED  UPON  LIMITS FROM ABOVE
      NRM-LOC(IUP)-LOC(IDN)
      NMR-NRM+1
      IF(LOC(IDN).EQ.l) GO TO 27
      DO 24  I-l.NMR
      STA(I)-STA(I-1+LOC(IDN))
      ABAS (I) -ABAS (I - 1+LOC (IDN ) )
      DATU (I) -DATU (I - 1+LOC (IDN) )
   24 AT(I)-AT(I-1+LOC(IDN))
      NODE(1)-NODE(1) -LOC(IDN)+1
      NODE(2)-NODE(2) -LOC(IDN)+1
      NODE(3)-NODE(3) -LOC(IDN)+1
      NODE(4)-NODE(4) -LOC(IDN)+1
      NODE(5)-NODE(5) -LOC(IDN)+1
      NODE(6)-NODE(6) -LOC(IDN)+1
      NODE(7)-NODE(7) -LOC(IDN)+1
      NODE(8)-NODE(8) -LOC(IDN)+1
      NODE(9)-NODE(9) -LOC(IDN)+1
      NODE (10)-NODE (10) -LOC(IDN)+1
      NBR(1)-NBR(1)-LOC(IDN)+1
      NBR(2)-NBR(2)-LOC(IDN)+1
      NBR ( 3 ) -NBR( 3 ) - LOG (IDN)+1
      NBR(4)-NBR(4)-LOC(IDN)+1
      NBR(5)-NBR(5)-LOC(IDN)+1
      DO  25  I-IUP.IDN
    25 LOC(I)-LOC(I)-LOC(IDN)+1
 C
 C  .  .  .  CALCULATE DISTANCES BETWEEN SECTIONS
    27 DO 30  I-l.NRM
      X(I)-STA(I+1)-STA(I)
    30  IF(I.EQ.NBR(1).OR.I.EQ.NBR(2) .OR. I .EQ.NBR(3) .OR. I .EQ.
                                  67

-------
     >NBR(4).OR.I.EQ.NBR(5» X(I)-0.0
C
C .  .  .  The following lines which, have been COMMENTED with "cO", will
C .  .  .  WRITE the BASIC PHYSICAL DATA.  Should the user desire to see
C .  .  .  this data, the "cO's" would have to be eliminated and the
C .  .  .  program re-compiled/1inked etc. TO RE-COMPILE submit DCDELTA
C
cO      WRITE (6,3000)
cO      WRITE (6,3010)
cO      DO 40 I-l.NMR
cO       MARK(I)-'  '
cO        IF(I.EQ.NODE(3).OR.I.EQ.NODE(S).OR.I.EQ.NODE(6)
cO     >.OR.I.EQ.NODE(7).OR.I.EQ.NODE(8).OR.I.EQ.NODE(9).OR.
cO     >I.EQ.NODE(10)) MARK(I)-'<'
cO        IF(I.GT.NODE(2).AND.I.LE.NBR(1))MARK(I)-'N  '
cO        IF(I.GT.NBR(1).AND.I.LT.NODE(3))MARK(I)-'M  '
cO          IF(I.GT.NODE(3).AND. I .LE.NBR(2) )MARK(I)-'UN'
cO          IF(I.GT.NODE(4).AND.I.LE.NBR(3))ttARK(I)-'S  '
cO          IF(I.GT.NBR(3).AND.I.LT.NODE(5))MARK(I)-'CO'
cO          IF(I.GT.NODE(5).AND.I.LT.NODE(6))MARK(I)-'US'
cO          IF(I.GT.NODE(7).AND.I.LE.NBR(4).OR.I.GT.NODE(9).AND.I
cO     >.LE.NBR(5))MARK(I)-'W  •
cO          IF(I.GT.NBR(4).AND.I.LT.NODE(8).OR.I.GT.NBR(5).AND.I
cO     >.LT.NODE(10))MARK(I)-'E  '
cO       IF(I.LE.NODE(2).OR.I.GT.NBR(2).AND.I.LE.NODE(4)) GO TO 40
cO       WRITE(6,3020) STA(I),ABAS(I),DATU(I),AT(I),MARK(I)
cO       DO 40 IJ-IUP.IDN
cO   40  IF(I.EQ.LOC(IJ))  WRITE(6,3030) NAME(IJ)
C
C     THESE NEXT  2 MANNINGS N  PRINTOUTS ARE HARDWIRED.  IF THE
C     NUMBER OF STATIONS  CHANGES,  CHANGE THESE!
C
cO       IF(IDN.EQ.S) WRITE(6,3040) Al(8),NMM(8),B1(8) ,STA(7),STA(36)
cO       IF(IDN.EQ.S) WRITE(6,3040) Al(7),NMM(8),B1(7) ,STA(40),STA(77)
cO       DO 45 I-IIDN.IUP.-l
cO   45  IF(I.LT.7)WRITE(6,3040) Al(I),NMM(I),B1(I),STA(LOC(I+1)),
cO     >STA(LOC(I))
C
C  .  .  .  INITIALIZE AND ASSIGN  ADDITIONAL VARIABLES
      DO 50 1-1,4
    50 ADJ(I)-0.
      NVAR-NRM*2
      ANC-24.
      DT-ANC*3600.
      DO 51 1-1,42
    51 SUM(I)-0.
      TH-.75
      TH1-.25
      MM-0
      M=13
       istart-13
       iend-43
                                  68

-------
      kkz-11
      MON-MONA
      IYR-IYRA
      MONFLAG-.FALSE.
      JRB-(IUP+IDN)/2
C
C     Routine to read all of the water level data from the disk and
C     store it in another tempory file.  This way the disk is not tied
C     up for long periods of time when running the program.
C
C .  .  .  READ WATER LEVELS FROM DISC.
C
 52   CALL NODAYS( IYR.MON,1.NDM.NDY, JD)
      DO 55 JJ-IUP.IDN
      IU-1
      IC-IGAGE(JJ)/10000
      IGAG-IGAGECJJ)-IC*10000
      CALL GAGEIO( IW.IC.IGAG  ,MON,IYR,IB.IT.IDA.IDB.IDC,IER)
      IF(IER.NE.O) THEN
         WR1TE(6.3110) NAMMON(MON).IYR,IER
         CALL EXIT
      END  IF
      DO 54 J-l.NDM
      DO 53  1-1,24
        WRITE(9,56) IHOUR(I.J)
  53  CONTINUE
      WRITE(9,56) MEAN(J)
  54  CONTINUE
  55  CONTINUE
   56    FORMAT(I6)
 C
       IF(IYR-IYRB)  58,57,65
   57   IF(MON-MONB)  58,65,65
   58   CONTINUE
       IF(MON-12) 60,59,59
   59   IYR-IYR+1
   60   CONTINUE
 C .   .  . UPDATE MONTH AND YEAR AND RECHECK IF MORE DATA  SHOULD BE USED
       MON-MON+1
       IF(MON-13) 62,61,61
   61   MON-1
   62  IF(IYR-IYRB)  64,63,65
   63  IF(MON-MONB)  64,64,65
   64  CONTINUE
 C
       GO TO 52
   6 5  CALL DISMOUNTPACK( ' WATER_LEVELS' )
       MON-MONA
        IYR-IYRA
       REWIND 9
 C
 C       COME HERE EACH MONTH AND PRINT TITLES AND HEADINGS FOR
                                  69

-------
C       EACH OPTION
 70   continue
      if(Nopt.eq.1.and.iup.eq.1)  iiup-iiup+1
C
C ********************QPTION FOR  WATER  LEVELS********************
C
      IF(NOPT.EQ.l) THEN
       WRITE(6,3000)
       WRITE(6,3050) NAMMON(MON) , IYR,NAME(IUP) .NAME(IDN)
       WRITE(6,3060) ANC.NRM
         WRITE(6,3070) NMM(IDN),NMM(IUP),
     >   NMM(IUP),(NMM(I),I-IIDN,IIUP,-1)
         WRITE(6,3071) (NMM(I),I-IIDN.IIUP,-1)
      IF(NOPT.EQ. LAND. IUP. EQ. 1)IIUP-IIUP-1
C
C ******************QPTION FOR  FLOWS  WITHOUT  DELTA*****************
C
      ELSE  IF(NOPT.EQ.2  .AND. IDN.NE.8) THEN
        WRITE(6,3000)
        WRITE(6, 8051 )NAMMON(MON) ,lYR.NAME(IUP) .NAME(IDN)
        WRITE(6,3060)ANC,NRM
        WRITE(6,7020)
        WRITE(6,7030)NMM(IDN) .NMM(JRB)  .NMM(IUP) .NMM(IIDN) ,NMM(JRB) ,
     >  NMM(IUP),NMM(JRB)
        WRITE(6,7040)
C
C  ******************QPTION FOR  FLOWS WITH DELTA********************
C
      ELSE  IF(NOPT.EQ.2  .AND.  IDN.EQ.8) THEN
        WRITE(6,3000)
        WRITE(6,8051 )NAMMON(MON)  ,IYR,NAME(IUP) ,NAME(IDN)
        WRITE(6,3060) ANC.NRM
        WRITE(6.8020)
        WRITE(6,8030)NMM(IDN) .NMM(IIDN-l) ,NMM(IUP) .NMM(IUP) ,NMM(IIDN) ,
      >  NMM(IIDN-l)
        WRITE(6,8040)
 C
 C  ****************OPTION FOR VELOCITIES WITHOUT DELTA***************
 C
        ELSE IF(NOPT.EQ.3 .AND.  IDN.NE.8) THEN
         WRITE(6,3000)
         WRITE(6,9051)NAMMON(MON) , lYR.NAME(IUP) .NAME(IDN)
         WRITE(6,3060)ANC,NRM
         WRITE(6,9025)
         WRITE(6,7030)NMM(IDN) ,NMM(JRB) .NMM(IUP) ,NMM(IIDN) ,NMM(JRB) ,
      >  NMM(IUP),NMM(JRB)
         WRITE(6,9041)
 C
 C  ****************OPTION FOR VELOCITIES WITH DELTA******************
 C
        ELSE IF(NOPT.EQ.3 .AND.  IDN.EQ.8) THEN
         WRITE(6,3000)
                                 70

-------
        WRITE(6,9051)NAMMON(MON) ,IYR,HAME(IUP) .NAME(IDN)
        WRITE(6,3060)ANC,NRM
        WRITE(6,9020)
        WRITE(6,9030)NMM(IDN) .NMM(IIDN-l),NMM(IUP) ,NMM(IUP) .NMM(IIDN) ,
     >  NMM(IIDN-l)
        WRITE(6.9040)
       END IF
C
 71   continue
C       ADJUST GAGE-SPECIFIC VARIABLES  BASED UPON YEARS STUDIED
      IF(IYR.LT.1970)  IGAGE(1)-14099
      IF(IYR.LT.1971)  IGAGE(6)-14080
      IF(IYR.GT.1970)  ADJ(l)—.18
      IF(IYR.GT.1981)  ADJ(l)—.06
      IF(IYR.LT.1971)  ADJ(6)—.09
C
C        READ WATER LEVELS  FROM UNIT 9 ...
C
      CALL NODAYS( IYR.MON,1.NDM.NDY,JD)
C
      DO 110 JJ-IUP.IDN
 C
       DO 77 J-l.NDM
        DO 75 1-1,24
  75     READ(9,1050,END-78) IHOUR(I.J)
  77   READ(9,1055,END-78) MEAN(J)
 C
  78   CONTINUE
       KK -  1
       DO 100 J-ISTART.IEND
       DO 80 1-1,24
 C
 C       FLAG MISSING  DATA AND  ASSIGN GAGE LEVELS TO WSSAV
        IF(IHOUR(I,KK).GT.O)  THEN
         NOTE(JJ.KK)-'  '
        ELSE
         NOTE(JJ.KK)-'*'
          GO TO 90
        END IF
     80  CONTINUE
     90  WSSAV(J.JJ)-0.0
        IF(MEAN(KK).LE.O) THEN
          WSSAV(J,JJ)-OLD(JJ)
          NOTE(JJ,KK)-'E'
        ELSE
          WSSAV(J , JJ)-(MEAN(KK) +IB)/100.0+ADJ (JJ)
        END  IF
        OLD(JJ) - WSSAV(J.JJ)
        IF(JJ.EQ.IUP.OR.JJ.EQ.IDN) WS(J ,LOC(JJ))-WSSAV(J,JJ)
    100 KK - KK+1
    110 CONTINUE
         IF(MONFLAG)  GO TO  200
                                     71

-------
c
C .  .  .  SET WS FOR 12 PREVIOUS  TIME STEPS, TO ACHIEVE 'STEADY STATE'.
      DO 120 1-1,12
       DO 120 J-1,8
        WSSAV(I.J)-WSSAV(13,J)
  120   IF(J.EQ.IUP  -OR.  J.EQ.IDN)  WS(I,LOC(J) )-WSSAV(13, J)
C
C . . . ZERO MATRIX,  SET CHANNEL PARAMETERS, & SET INITIAL CONDITIONS.
      DO  130 I - 1,NVAR
       YVECT(I)-0.
       DO 130 J  - l.NVAR
  130   XMTRX(J,I)-0.
      XSUM-STA(LOC( IUP) ) - STA(LOC{ IDN) )
      SLOPE-
-------
  200  CONTINUE
      N-M+1
      ITER-1
  210  YVMAX-0.
      DO 230 I-l.NRM
       QA(I)-TH/2.^Q(N,I)4   IDN.EQ.8.AND.I.GT.106.AND.ID.EQ.LL.OR.IDN.EQ.8.AND.I.GT.106
     >   .AND.IU.EQ.KK.OR.IDN.EQ.8.AND.ID.EQ.6.0R.IDN.EQ.8.AND.ID.EQ.
     >   24.0R.IDN.EQ.8.AND.ID.EQ.39.0R.IDN.EQ.8.AND.ID.EQ.50.0R.
     >   IDN.EQ.8.AND.ID.EQ.3) THEN
         YVECT(I)— (WS(N,IU)+WS(M,IU)-WS(N,ID)-WS(M,ID))/2.
         XMTRX(I,I-1)— 0.5
         XMTRX(I,I+1)-+0.5
         GO TO 260
        ELSE IF(ID.EQ.JJ.AND.IDN.NE.8.0R.ID.EQ.JJ.AND.IDN.EQ.8
     >   .AND.J.GE.4) THEN
         YVECT(I)— (WS(N,KK)+WS(M,KK) -WS(N.JJ) -WS(M,JJ))/2 .
         XMTRX(I,I-1)— 0.5
         XMTRX( I , 2*KK- 2 ) -+0 . 5
         GO TO 260
        ELSE IF(IDN.EQ.8.AND.ID.EQ.NBR(1)) THEN
         YVECT(I)--(VS(N,ID)+WS(M,ID)-WS(N,NODE(3))-WS(M,NODE(3)))/2.
         XMTRX(I,28) - +.5
         XMTRX(I,48) - -.5
         GO TO 260
         ELSE  IF(IDN.EQ.8.AND.ID.EQ.NBR(2)) THEN
         YVECT(I)--(WS(N,ID)+WS(M,ID)-WS(N,NODE(6))-WS(M,NODE(6)))/2.
         XMTRX(I,70)- +.5
                                 73

-------
        XMTRX(I,150>-  -.5
        GO TO 260
       ELSE IF(IDN.EQ.8.AND.ID.EQ.NBR(3)) THEN
        YVECT(I)— (WS(N, ID)+WS(M. ID) -WS(N.NODE(5)) -WS(M,NODE(5)) )/2 .
        XMTRX(I,88)- +.5
        XMTRXd.lOO)-  -.5
        GO TO 260
       END IF
  255   CONTINUE
       YVECT(I)— ((WS(N,ID)+WS(N,IU)-WS(M,ID)-WS(M,IU))/(2.*DT) +
    >   (TH*(Q(N,ID)  -Q(N,IU))+TH1*(Q(M,ID)-Q(M,IU)))/(T(ID)*X(ID)))
       XMTRX(I,I)-TH/(T(ID)*X(ID))
       XMTRX(I,I+2)--XMTRX(I,I)
       IF(I.EQ.l) THEN
        XMTRX(1,2)-1./(2.*DT)
       ELSE
        XMTRX(I,I-1)-1./(2.*DT)
        XMTRX(I, I+D-1./(2 . *DT)
         IF(I.EQ.NRD)  XMTRX(I,I+l)-XMTRX(I,I+2)
         IF(I.EQ.NRD)  XMTRX(I,I+2)-0.
       END IF
  260  CONTINUE
C
C .  .   . MOMENTUM EQUATIONS
      DO 280 I-2.NVAR.2
       ID-I/2
       IU-ID+1
       DO 265 J-1,5
        JJ-NBR(J)
        KK-NODE(2*J)
        LL-NODE(2*J-1)
        IF(IDN.NE.8.AND.ID.EQ.LL.OR.IDN.EQ.8.AND.J.GE.4.AND.
     >  ID.EQ.LL) THEN
         YVECT(I)—(Q(N,JJ-H)4<5(M,JJ+1)4<3(N>IU)-K}(M,IU)-Q(N,ID)-Q(M,ID))
     >     /2.
         XMTRX(I,I-1) —0.5
         XMTRX(I,I+D-+0.5
         XMTRX( 1, 2*J J+l)-+0. 5
         GO TO 280
        ELSE IF(IDN.NE.8.AND.ID.EQ.JJ.OR.IDN.EQ.8.AND.J.GE.4.AND.
     >  ID.EQ.JJ) THEN
         YVECT(I)—(WS(N,IU)+WS(M,IU)-WS(N>LL)-WS(M,LL))/2.
         XMTRX(I,2*LL-2)--0.5
         XMTRX(I,I)-+0.5
         GO TO 280
        ELSE IF(IDN.NE.8.AND.IU.EQ.KK.OR.IDN.EQ.8.AND.J.GE.4.AND.
     > IU.EQ.KK) THEN
         YVECT(I)— (Q(N,KK)-KHM,KK) -Q(N,ID)-Q(M,ID) -Q(N, JJ) -Q(M, JJ)
         XMTRX(I,2*JJ-1)--0.5
         XMTRX(I,I-D —0.5
         XMTRX(I,H-1)-+0.5
         GO TO  280


-------
ELSE IF(IDN.EQ.8.AND.ID.EQ.NODE(1)) THEN
 YVECT(I)- -(Q(N,NODE(1))-KKM,NODE(1))-Q(N,NODE(1)+1)
 -Q(M,NODE(l)+l)-Q(N,NBR(2)+l)-Q(M,NBR(2)+l))/2.
 XMTRX(I,I-D- +.5
 XMTRX(I,I+1)- -.5
 XMTRX(I,73)- -.5
 GO TO 280
ELSE IF(IDN.EQ.8.AND.ID.EQ.NODE(2)) THEN
 YVECT(I)--(Q(N,NODE(2))-KKM,NODE(2))-Q(N,NODE(2)+1)
 -Q(M,NODE(2)-H)-Q(N,NBR(l)+l)-Q(M,NBR(l)+l))/2.
 XMTRX(I.I-l)- +.5
 XMTRX(I,I+1)- -.5
 XMTRX(I,31)- -.5
 GO TO 280
ELSE IF(IDN.EQ.8.AND.ID.EQ.NODE(4)) THEN
 YVECT(I)—(Q(N,NODE(4))4Q(M,NODE(4))-Q(N.NODE(4)+1)
 -Q(M,NODE(4)+l)-Q(N,NBR(3)+l)-Q(M,NBR(3)+l))/2.
 XMTRX(I.I-l)- +.5
 XMTRX(I,I+1)- -.5
 XMTRX(I,91)- -.5
 GO TO 280
ELSE IF(IDN.EQ.8.AND.ID.EQ.NBR(1))  THEN
 YVECT(I)—(WS(N,NBR(1)+1)+WS(M,NBR(1)+1)-US(N,NODE(2))
  -WS(M,NODE(2)))/2.
 XMTRX(I.I)- +.5
 XMTRX(I,10)- -.5
 GO TO 280
 ELSE IF(IDN.EQ.8.AND.IU.EQ.NODE(3)) THEN
 YVECT(I)—(Q(N,NODE(3))-KJ(M,NODE(3))-Q(N,NODE(3)-1)
  -Q(M,NODE(3)-l)-Q(N,NBR(l))-Q(M,NBR(l)))/2.
 XMTRX(I.H-l)- +.5
  XMTRX(I.I-l)-  -.5
  XMTRX(I,29)-  -.5
  GO TO 280
 ELSE  IF(IDN.EQ.8.AND.ID.EQ.NBR(2»  THEN
  YVECT(I)--(WS(N,NODE(1))+WS(M,NODE(1))-WS(N,NBR(2)+1)
  -WS(M,NBR(2)+l))/2.
  XMTRX(I,72)-  -.5
  XMTRX(I,4)- +.5
  GO TO  280
 ELSE  IF(IDN.EQ.8.AND.ID.EQ.NBR(3))  THEN
  YVECT(I)—(WS(N,NODE(4))+WS(M,NODE(4))-WS(N,NBR(3)+1)
  -WS(M,NBR(3)+l))/2.
  XMTRX(I.76)- +.5
  XMTRX(I,90)-  -.5
  GO TO  280
 ELSE  IF(IDN.EQ.8.AND.IU.EQ.NODE(5)) THEN
  YVECT(I)—(Q(N,NODE(5))-H5(M,NODE(5))-Q(N,NODE(5)-1)
  -Q(M,NODE(5)-l)-Q(N,NBR(3))-Q(M,NBR(3)))/2.
  XMTRX(I.lOl)- +.5
  XMTRX(I,99)-  -.5
  XMTRX(I,89)-  -.5
                        75

-------
         GO TO 280
        ELSE IF(IDN.EQ.8.AND.IU.EQ.NODE(6)) THEN
         YVECT(I)— (Q(N,NODE(6))4Q(M,NODE(6))-Q(N,NODE(6)-1)
     >   -Q(M,NODE(6)-l)-Q(N,NBR(2))-Q(M,NBR(2)))/2.
         XMTRX(I,151)- +.5
         XMTRX(I,71)- -.5
         XMTRX(I,149)- -.5
         GO TO 280
        END IF
  265  CONTINUE
       241— QA(ID)*T(ID)*(WS(N,IU)+WS(N,ID)-WS(M,IU)-WS(M,ID))/
     >  (2.*DT*A(ID)**2.)*2.
       Z61-(Q(N1ID)-K}(N,IU)-Q(MIID)-Q(M,IU))/(2.*DT*A(ID))
       Zn-32.17*AN(ID)**2.*QA(ID)*U(ID)/(2.21*A(ID)**2.*R(ID)**(4 /3 )
       Z21-32 . 17*(TH*(WS(N, ID) -WS(N, IU) )+THl*(WS(M, ID) -WS(M, IU) ) ) A(ID)
       Z31— (QA(ID)**2.*(AA(ID)-AA(IU))/(A(ID)**3.*X(ID)))
       YVECT(I)--(Z11+Z21+Z31+Z41+Z61)
       XMTRX(I,I)— (32.2-QA(ID)**2.*T(ID)/(A(ID)**3.))/X(ID)*TH-QA(ID)
     >   *T(ID)/(DT*A(ID)**2.)
       IF(I.LT.4)  GO TO 270
       XMTRX(I,I-2)-(32.2-QA(ID)**2.*T(ID)/(A(ID)**3.))A(ID)*TH-QA(ID)
     >   *T(ID)/(DT*A(ID)**2.)
  270  ZZZ1-ABS(QA(ID))
       Pl-32.2*AN(ID)**2.*ZZZl*TH/(2.2082*A(ID)**2.*R(ID)**(4./3.))
     >  -QA(ID)/A(ID)**3.*(AA(ID)-AA(IU))*TH/X(ID)-TH*T(ID)*(WS(N ID)
     >  +WS(N,IU) -WS(M, ID) -WS(M, IU))/(2 .*DT*A(ID)**2. )
       XMTRX(I,I-1)-1./(2.*A(ID)*DT)+P1
       XMTRX(I , I+D-XMTRXU , I- 1)
       IF(I.EQ.NVAR) XMTRX(I,I)-XMTRX(I,I-H)
       IF(I.EQ.NVAR) XMTRX(I,I+1)-0.
  280 CONTINUE
C
C .   . .SET UP AVECT, JA, AND IA,  AND ACCESS SPARCE MATRIX SOLVER
      DO 290 I-l.NVAR
  290 IA(I)-0.
      DO 291 1-1,900
       AVECT(I)-0.
  291  JA(I)-0.
      K-0
      DO 292 I-l.NVAR
       NEWROW-.TRUE.
       DO 292 J-l.NVAR
        IF(XMTRX(I,J).NE.O.)  THEN
         K-K+1
         AVECT(K)-XMTRX(I , J)
         JA(K)-J
         IF(NEWROW)  THEN
          NEWROW-. FALSE.
                                76

-------
         END IF
        END IF
  292 CONTINUE
      IA(NVAR+1)-K+1
      NSP-100+8*NVAR+2+2*K
      PATH-2
      IF(ICOUNT.EQ.O) PATH-1
      ICOUNT-ICOUNT+1
      CALL CDRVCNVAR.RR.CC.ICC.IA.JA.AVECT.YVECT.YV.NSP ISP RSP ESP
     > PATH,FLAG)
      IF(FLAG.NE.O) THEN
       WRITE(6,6190) FLAG,ESP,PATH
       GO TO 500
      END IF
C
C .  . .  DETERMINE NEW FLOWS AND LEVELS
      NNR-NVAR+1
      DO 300 I-1.NNR.2
        II-I/2+1
        IF(II-NMR.EQ.O) Q(N,NMR)-Q(N,NMR)+YV(NVAR)
  300 CONTINUE
      NNR-NVAR-2
      DO 310 I-2.NNR.2
        IF (YV(I).LT.-20.) YV(I)—20.
        II-I/2+1
        WS(N,II)-WS(N,II)+YV(I)
  310 CONTINUE
C
C .  . .  ASSIGN MAXIMUM DELTA VALUE, AND TEST FOR EXCESSIVE ITERATIONS
      DO 319 I-2.NNR.2
  319  IF(ABS(YV(I)).GT.YVMAX) YVMAX-YV(I)
      DO 320 I-2.NNR.2
        IF(ABS(YV(I)).GT..002) THEN
          IF(ITER.GT.20) THEN
            WRITE(6,1133) ITER,YVMAX,I/2
 1133       FORMATC  ',12,' ITERATIONS.  YVMAX -' ,E10.3 'AT I-' 12 /)
            GO TO 500                                          '
          END IF
          ITER-ITER+1
          GO TO 210
        END IF
  320 CONTINUE
C
C .  . .  DETERMINE THE DEVIATION OF CALCULATED FROM  MEASURED LEVELS
      JB-N+1
      DO 330 IJ-IIUP.IIDN
      IF(WSSAV(N,IJ).LT.50.) WSSAV(N, IJ)-WS(N,LOC(IJ)) - .0001
  330 DEV(IJ)-WS(N,LOC(IJ))-WSSAV(N,IJ)
      MM-MM+1
      NM-MM-kkz
      IF(NM.LE.O) GO TO 340
                                77

-------
      IF(NOPT.EQ.3)  THEN
       DO 6262 I-l.NRM+1
 6262  VEL(I)-Q(N,I)/AA(I)
      END IF
C
C .  . .  METRIC OPTION FOR OUTPUT .  .
      IF(MUNITS.EQ.O)THEN
        DO 6700 MC - IUP.IDN
 6700   WSSAV(N.MC)  - WSSAV(N,MC)/3.28083
        DO 6720 MC- IIUP.IIDN
         WS(N,LOC(MC)) - WS(N.LOC(MC))/3.28083
 6720   DEV(MC) - DEV(MC)/3.28083
C
        Q(N,LOC(IUP)) - Q(N,LOC(IUP))*0.02832
        Q(N,LOC(IIDN)) - Q(N,LOC(IIDN))*0.02832
        Q(N,NBR(5)+5) - Q(N,NBR(5)+5)*0.02832
        Q(N,NBR(5)-4) - Q(N.NBR(5)-4)*0.02832
        Q(N,NODE(10)+1) - Q(N,NODE(10)+1)*0.02832
        Q(N,NBR(4)+3) - Q(N,NBR(4)+3)*0.02832
        Q(N,NBR(4)-2) - Q(N,NBR(4)-2)*0.02832
        Q(N,NODE(8)+1) - Q(N,NODE(8)+1)*0.02832
        Q(N,NBR(l)-3) - Q(N,NBR(1)-3)*0.02832
        Q(N,NBR(l)+6) - Q(N,NBR(1)+6)*0.02832
        Q(N,NBR(3)-3) - Q(N,NBR(3)-3)*0.02832
        Q(N,NBR(3)+3) - Q(N,NBR(3)+3)*0.02832
        Q(N,LOC(JRB)) - Q(N,LOC(JRB))*0.02832
C
        VEL(LOC(IUP)) - VEL(LOC(IUP))/3.28083
        VEL(LOC(IIDN)) - VEL(LOC(IIDN))/3.28083
        VEL(NBR(5)+5) - VEL(NBR(5)+5)/3.28083
        VEL(NBR(5)-4) - VEL(NBR(5)-4)/3.28083
        VEL(NBR(4)+3) - VEL(NBR(4)-t-3)/3 .28083
        VEL(NBR(4)-2) - VEL(NBR(4)-2)/3.28083
        VEL(NBR(l)-3) - VEL(NBR(l)-3)/3.28083
        VEL(NBR(l)+6) - VEL(NBR(l)+6)/3.28083
        VEL(NBR(3)-3) - VEL(NBR(3)-3)/3.28083
        VEL(NBR(3)+3) - VEL(NBR(3)+3)/3.28083
        VEL(LOC(JRB)) - VEL(LOC(JRB))/3.28083
      END IF
C
C  .  .  . PRINT OUTPUT.
         if(NOPT.EQ.l.and.iup.eq.l) iiup-iiup+1
C
C ********************QPTION FOR WATER  LEVELS********************
C
       IF(NOPT.EQ.l) THEN
        WRITE(6,3120) NM.WSSAV(N,IDN),NOTE(IDN,NM),
      >   (WS(N.LOC(I)),I-IIDN,IIUP,-1)
        WRITE(6,3121) WSSAV(N,IUP),NOTE(IUP,NM),
      >  Q(N,LOC(IUP)),
      >   (WSSAV(N,I),NOTE(I,NM),DEV(I),  I-IIDN, IIUP,-1)
                                                                         A-tUV

-------
C ****************OPTION FOR FLOW WITHOUT DELTA******************
C
      ELSE IF(NOPT.EQ.2.AND.IDN.NE.8) THEN
      WRITE(6,7050)NM,WSSAV(N,IDN),NOTE(IDN,NM),WS(N,LOC(JRB))
     >WSSAV(N,IUP)1NOTE(IUPINM),Q(N,LOC(IIDN)),Q(N,LOC(JRB))
     XKN.LOC(IUP)) ,Q(N,NBR(5)+5) ,Q(N,NBR(5) -4) ,Q(N.NBR(4)+3) ,
     >Q(N,NBR<4)-2),WSSAV(N,JRB),NOTE(JRB,NM).DEV(JRB)
C
C *******************QPTION FOR FLOW WITH DELTA*****************
C
      ELSE IF(NOPT.EQ.2  .AND. IDN.EQ.8) THEN
       WRITE(6,8050)NM,WSSAV(N.IDN),NOTE(IDN,NM),WS(N,LOC(IIDN-1))
     >WSSAV(N,IUP),NOTE(IUP,NM),Q(N,LOC(IUP)),Q(N,LOC(IIDN)),
     >Q(N,NBR(5)+5).Q(N,NBR(5)-4),Q(N,NBR(4)+3),Q(N,NBR(4)-2),
     >Q(N>NBR(l)-3)IQ(N,NBR(l)+6),Q(N1NBR(3)-3)1Q(N,NBR(3)+3),
     >WSSAV(N,IIDN-1),NOTE(IIDN-1,NM),DEV(IIDN-1)
C
C *****************QPTION  FOR VELOCITIES WITHOUT DELTA**********
C
      ELSE IF(NOPT.EQ.3  .AND. IDN.NE.8) THEN
      WRITE(6,9055)NM,WSSAV(N,IDN),NOTE(IDN,NM),WS(N,LOC(JRB)),
     >WSSAV(N,IUP) ,NOTE(IUP,NM) ,VEL(LOC(IIDN)) ,VEL(LOC(JRB) ) ,
     >VEL(LOC(IUP)),VEL(NBR(5)+5),VEL(NBR(5)-4).VEL(NBR(4)+3),
     >VEL(NBR(4)-2),WSSAV(N,JRB),NOTE(JRB,NM),DEV(JRB)
C
C *****************QPTION  FOR VELOCITIES WITH DELTA*************
C
      ELSE IF(NOPT.EQ.3  .AND. IDN.EQ.8) THEN
       WRITE(6,9050) NM,WSSAV(N,IDN),NOTE(IDN,NM),WS(N,LOC(IIDN-1))
     >WSSAV(N,IUP),NOTE(IUP,NM),VEL(LOC(IUP))1VEL(LOC(IIDN)),
     >VEL(NBR(5)+5) ,VEL(NBR(5) -4) ,VEL(NBR(4)+3) ,VEL(NBR(4) -2) ,
     >VEL(NBR(l)-3),VEL(NBR(l)+6),VEL(NBR(3)-3),VEL(NBR(3)+3),
     >WSSAV(N,IIDN-1),NOTE(IIDN-1,NM),DEV(IIDN-1)
       END IF
        if (Nopt.eq. Land, iup.eq.l) iiup-iiup-1
 333  continue
C
C .  . . COMPUTE MEAN VALUES.
      SUM(1)-SUM(1)+WSSAV(N,IUP)
      DO  334 1-2,7
  334 SUM(I)-SUM(I)+WS(N,LOC(I))
      SUM(8)-SUM(8)+WSSAV(N,IDN)
      DO  335 1-1,6
       SUM(I+8)-SUM(I+8)+WSSAV(N,I+l)
  335 SUM(H-14)-SUM(1+14)+DEV(H-l)
C
C  .  .  . THESE ARE THE  CALCULATIONS FOR FLOW AVERAGES
      SUM(21)-SUM(21)+Q(N,LOC(IUP))
      SUM(22)-SUM(22)+Q(N,NBR(5)+5)
      SUM(23)-SUM(23)+Q(N,NBR(5)-4)
      SUM(24)-SUM(24)+Q(N,NODE(10)+1)
      SUM(25)-SUM(25)4Q(N,NBR(4)+3)


-------
      SUM(26)-SUM(26)-KJ(N,NBR(4)-2)
      SUM(27)-SUM(27)-KKN,NODE(8)+1)
      SUM(28)-SUM(28)-K}(N,LOC(IIDN))
      SUM(29)-SUM(29)4<3(N.NBR(l)-3)
      SUM(30)-SUM(30)+Q(N.NBR(l)+6)
      SUM(31)-SUM(31)+Q(N,NBR(3)-3)
      SUM(32)-SUM(32)-KKN,NBR(3)+3)
      SUM(43)-SUM(43)-KKN,LOC(JRB))
      SUM(33)-SUM(33)+VEL(LOC(IUP))
      SUM(34)-SUM(34)+VEL(LOC(IIDN))
      SUM(35)-SUM(35)+VEL(NBR(5)+5)
      SUM(36)-SUM(36)+VEL(NBR(5)-4)
      SUM(37)-SUM(37)+VEL(NBR(4)+3)
      SUM(38)-SUM(38)+VEL(NBR(4)-2)
      SUM(39)-SUM(39)+VEL(NBR(l)-3)
      SUM(40)-SUM(40)+VEL(NBR(l)+6)
      SUM(41)-SUM(41)+VEL(NBR(3)-3)
      SUM(42)-SUM(42)+VEL(NBR(3)+3)
      SUM(44)-SUM(44)+VEL(LOC(JRB))
C
C .  .  CONVERT BACK TO ENGLISH UNITS .  .
      IF(MUNITS.EQ.O)THEN
        DO 6730 MC - IUP.IDN
 6730   WSSAV(N.MC) - WSSAV(N,MC)*3.28083
        DO 6740 MC - IIUP.IIDN
         WS(N,LOC(MC)) - WS(N,LOC(MC))*3.28083
 6740   DEV(MC) - DEV(MC)*3.28083
C
        Q(N,LOC(IUP)) - Q(N,LOC(IUP))/0.02832
        Q(N,LOC(IIDN)) - Q(N,LOC(IIDN))/0.02832
        Q(N,NBR(5)+5) - Q(N,NBR(5)+5)/0.02832
        Q(N,NBR(5)-4) - Q(N,NBR(5)-4)/0.02832
        Q(N,NODE(10)+1) - Q(N,NODE(10)+1)/0.02832
        Q(N,NBR(4)+3) - Q(N,NBR(4)+3)/0.02832
        Q(N,NBR(4)-2) - Q(N,NBR(4)-2)/0.02832
        Q(N,NODE(8)+1) - Q(N,NODE(8)+1)/0.02832
        Q(N,NBR(l)-3) - Q(N,NBR(l)-3)/0.02832
        Q(N,NBR(l)+6) - Q(N,NBR(l)+6)/0.02832
        Q(N,NBR(3)-3) - Q(N,NBR(3)-3)/0.02832
        Q(N,NBR(3)+3) - Q(N,NBR(3)+3)/0.02832
        Q(N,LOC(JRB)) - Q(N,LOC(JRB))/0.02832
C
        VEL(LOC(IUP)) - VEL(LOC(IUP))*3.28083
        VEL(LOC(IIDN)) - VEL(LOC(IIDN))*3.28083
        VEL(NBR(5)+5) - VEL(NBR(5)+5)*3.28083
        VEL(NBR(5)-4) - VEL(NBR(5)-4)*3.28083
        VEL(NBR(4)+3) - VEL(NBR(4)+3)*3.28083
        VEL(NBR(4)-2) - VEL(NBR(4)-2)*3.28083
        VEL(NBR(l)-3) - VEL(NBR(l)-3)*3.28083
        VEL(NBR(l)+6) - VEL(NBR(l)+6)*3.28083
        VEL(NBR(3)-3) - VEL(NBR(3)-3)*3.28083
        VEL(NBR(3)+3) - VEL(NBR(3)+3)*3.28083
                                  80

-------
        VEL(LOC(JRB» - VEL(LOC(JRB))*3.28083
      END IF
C
  340 DO 350 I-1,NMR
  350 Q(JB,I)-2.*Q(N,I)-Q(M.I)
      DO 360 1-2,NRM
  360 WS(JB,I)-2.*WS(N,I)-WS(M,I)
      write(7,2030)mon,nm,iyr-1900
      M-M+1
      IF(M-kb)2333,2333,370
 2333 continue
C
C .   . .  DAILY RETURN LOOP.
      IF(NM-NDM.LT.O) GO TO  200
  370 DO 380 1-1,44
  380 AVE(I)-SUM(I)/NM
C
C .   . .  PRINT MONTHLY MEAN VALUES
C
         if(Nopt.eq.1.and.iup.eq.1)  iiup-iiup+1
C
C ********************OPTION FOR WATER LEVELS*********************
C
      IF(NOPT.EQ.l)  THEN
        WRITE(6.3130) AVE(8),(AVE(I),I-IIDN,HUP,-1)
        WRITE(6,3131) AVE(1),AVE(21),
      >     (AVE(I+7),AVE(I+13),  I-IIDN,HUP,-1)
         if (Nopt. eq. Land, iup.eq. 1)  iiup-iiup-1
 C
 C ******************OPTION FOR FLOW WITHOUT DELTA*****************
 C
      ELSE IF(NOPT.EQ.2.AND.IDN.NE.8)  THEN
         WRITE(6, 7060)AVE(8)  ,AVE(3) ,AVE(1) ,AVE(28) ,AVE(43) ,AVE(21) ,
      >AVE(22) ,AVE(23) ,AVE(25) ,AVE(26) ,AVE(10) ,AVE(JRB+13)
 C
 C *******************OPTION FOR FLOW WITH DELTA*******************
 C
      ELSE IF(NOPT.EQ.2 .AND.  IDN.EQ.8) THEN
        WRITE(6, 8060)AVE(8) .AVE(IIDN-l) ,AVE(1) ,AVE(21) ,AVE(28) , AVE(22) ,
      >AVE(23) ,AVE(25) ,AVE(26) ,AVE(29) ,AVE(30) ,AVE(31) ,AVE(32) ,
      >AVE(IIDN-i-6) ,AVE(IIDN+12)
 C
 C *****************OPTION FOR VELOCITIES WITHOUT  DELTA************
 C
      ELSE IF(NOPT.EQ.3 .AND.  IDN.NE.8) THEN
        WRITE(6,9065)AVE(8) ,AVE(10) ,AVE(1) ,AVE(34) ,AVE(44) ,AVE(33) .
      >AVE(35) , AVE(36) ,AVE(37) ,AVE(38) ,AVE(3) ,AVE(JRB+13)
 C
 C *****************OPTION FOR VELOCITIES WITH DELTA***************
 C
       ELSE IF(NOPT.EQ.3 .AND.  IDN.EQ.8) THEN
        WRITE(6,9060) AVE(8) ,AVE(IIDN-1) ,AVE(1) ,AVE(33) ,AVE(34) ,AVE(35)


-------
     >AVE(36) ,AVE(37) ,AVE(38) ,AVE(39) ,AVE(40) ,AVE(41) ,AVE(42) ,
     >AVE(HDN+6) ,AVE(IIDNf 12)
      END IF
C

 381   continue
      MONFLAG-.TRUE.
      IF(IYR-IYRB)  388,386,500
  386 IF(MON-MONB)  388,500,500
  388 CONTINUE
      IF(MON-12)  390,389,389
  389 IYR-IYR+1
  390 CONTINUE
      DO 392 1-1 ,42
  392 SUM(I)-0.
      DO 400 I-l.NMR
  400 Q(2,I)-Q(JB,I)
      DO 410 1-2, NRM
  410 WS(2,I)-WS(JB,I)
 C
 C  .  .  . UPDATE MONTH AND YEAR  AND RECHECK IF MORE DATA SHOULD BE USED
      MON-MON+1
       IF(MON-13) 414,412,412
  412 MON-1
  414  IF(IYR-IYRB) 418,416,500
  416  IF(MON-MONB) 418,418,500
   418  CONTINUE
       kkz>0
       kb-36
       lstart*2
       iend-32
       MM»0
       M-l
 C
 C . . . MONTHLY  RETURN LOOP AND END PROGRAM LOCATION
       GO TO 70
   500 CALL EXIT
 C
 C . . . FORMAT  STATEMENTS.
  1000 FORMAT(I2,3(1X,I2))
  1010 FORMAT(I1,1X,I1>
  1020 FORMAT(F8.0,F8.0,F8.2,F8.0)
  1050 FORMAT(I6)
  1055 FORMAT(I6)
 C2000 FORMAT(/'   ENTER BEGINNING AND ENDING DATES'/' MM/YY-MM/YY' )
 C2020 FORMAT(/'  ENTER STARTING AND ENDING STATIONS: ' /8( 12 ,') ' ,A20/))
  2030 FORMATC  ' ,12 ,2( ' /'  ,13))
  3000 FORMAT(lHl)
  3010 FORMAT(///,26X,'ST. CLAIR  RIVER TRANSIENT MODEL' ,/ ,36X,
      >  'BASIC DATA' //,23X,' STATION' ,5X, 'ABASE' ,5X, 'DATUM', 5X, 'WIDTH1 ,/)
                                                                         A-ICJ*
                                      82

-------
3020 FORMATC20X,F10.0,F10.0,F9.2,F8.0,IX,A2)
3030 FORMATC'+',62X,A20)
3040 FORMATC/,12X,' MANNING n -',F11.8,' *  WS€',A3,' + '
    > ,F10.7,' FOR STATIONS',F7.0,' THRU ',F7.0)
3050 FORMATC///45X,'ST.CLAIR RIVER  TRANSIENT  MODEL',//,52X,A9 ,15,//,
    >  10X.A20,' to ',A20,40X,'WATER LEVELS VERSION',//)
3060 FORMAT(36X,F5.1,IX,'HOUR TIME  INCREMENTS',11X,13,IX,'REACHES',//)
3070 FORMAT(6X,A3,4X,'1	COMPUTED LEVELS	V ,2X,A3,5X,A3,4
    >X,'1	MEASURED LEVELS AND COMPUTED DEVIATIONS (C-M)	
    >1'/6X,'MEAS.',5(4X,A3))
3071 FORMATC'+',T50,'MEAS.   COMP.Q1 ,1X,4(1X,'1	',A3,'	1'),1X,
    > '1	',A3,'	1',/)
3110 FORMAT(1X,A9,1X,I4,1X,' ERROR  OF TYPE ',12)
3120 FORMAT(1X,I2,F8.2,A1,5(F7.2))
3121 FORMATC'+* ,T48,F7.2,A1,F8.0,5(F7.2,A1,1X,F4.2))
3130 FORMAT(/,1X,'AVE ' ,F6.2,1X,5(F7.2))
3131 FORMATC-I-' ,T48,F7.2 ,F8.0,1X,5(F7 .2.2X.F4.2))
6190 FORMATC   FLAG-',15,'   ESP-',15,'   PATH-',15)
C7000 FORMATC/'  ENTER OPTION NUMBER'/'  1.  OUTPUT SHOWS WATER LEVELS
C    > AND  DEVIATIONS'/'  2.  OUTPUT  SHOWS FLOWS AROUND ISLANDS')
7010 FORMAT(II)
7020 FORMATC//,17X/1	RIVER  PROFILE 	1',4X,'1	 TOTAL DISCHA
    >RGE	1 1	ISLAND FLOWS	1  1	DEV 	11,/)
 7030 FORMATCSX.'DAY' ,7X,A3 ,8X,A3,8X,A3 ,8X,A3 ,6X,A3,6X,A3,3X,
    >'STAG E',2X,'STAG W',2X,'FAWN E',2X,!FAWN W',4X,A3,7X,'DEV )
 7040 FORMATC18X,'MEAS.' ,6X,'COMP.'  ,6X,'MEAS.',6X,2C'FLOW' ,5X),'FLOW' ,3X
    >,4C'FLOW',4X),1X,'MEAS.' ,4X,'CC-M)' ,/)
 7050  FORMATC9X,I2,6X,F6.2,A1,4X,F6.2,5X,F6.2,A1,1X,3(2X,F7.0),
    >4C1X,F7.0),2X,F6.2,A1,2X,F6.2)
 7060  FORMATC/,8X,'AVE' ,6X,F6.2,5X,F6.2,5X,F6.2,2X,3C2X,F7.0) ,4ClX,F7.0)
     >,2X,F6.2,4X,F5.2)
C8001  FORMATC' Enter option for delta output:'//' 1. Output shows Water
C    >Levels and Deviations'/'  2. Output shows flows around delta and i
C    >slands'/' 3. Output shows velocities  around delta and islands'/)
 8002  FORMATC ID
 8005 FORMATCII)
 8020 FORMATC//,5X,'1	RIVER PROFILE	1' ,2X,' 1—TOTAL FLOW	1',3X,
     >i^	ISLAND FLOWS	1',3X,'1	DELTA FLOWS	
     >	1' ,3X,'1	DEV	1' ,//)
 8030 FORMATC7X,2CA3,5X),A3,6X,A3,6X,A3,6X,'STAG E',3X,'STAG  W',4X,'FAWN
     > E',3X,'FAWN W',SX.'N.CH.'.SX.'M.CH.'.SX.'S.CH.'^X,'CUTOFF',4X,
     >A3,5X,'DEV')
 8040 FORMATC6X,'MEAS.',4X,'COMP.',3X,'MEAS.',4X,'FLOW1,5X,'FLOW',8X,'Q
     >' ,8X,'Q' ,8X,'Q' ,8X,'Q' ,8X,'Q' ,7X,'Q' ,7X,'Q' ,8X,'Q' ,6X,'MEAS.' ,3X,
     >*C-M' /)
 8050 FORMATC1X,I2,F8.2,A1,1X,F7.2,F8.2,A1,F8.0,1X,F8.0,2X,F8.0,2X,
     >F7.0,1X,F8.0,2X,F8.0,1X,F7.0,1X,F7.0,1X,F7.0,2X,F7.0,
     >2X,F7.2,A1,1X,F5.2)
 8051 FORMATC///,45X,'ST.CLAIR RIVER TRANSIENT MODEL',//,52X,A9 ,15,//,
     MOX.A20,'  to  ',A20,40X,'RIVER DISCHARGE VERSION',//)
 8060 FORMATC/,1X,'AVE',F7.2,2X,F7.2,F8.2,1X,F8.0,1X,F8.0,2X,F8.0,2X,
     >F7.0,1X,F8.0,2X,F8.0,1X,F7.0,1X,F7.0,1X,F7.0,2X,F7.0,2X,F7.2,2X,
                                                                            -"-ft

-------
    >F5.2)
9020 FORMAT(//,5X,'1 -- RIVER PROFILE - 1' ,3X,' 1— TOT. VEL.— 1',4X,
    >'1 -- MID ISLAND VELOCITIES - 11 ,3X,'1 - MID DELTA VELOCITIES-
    > - 1 ' ,2X, ' 1 -- DEV --- 1 ' ,//)
9025 FORMAT(//,17X,'1 --- RIVER  PROFILE - 1',5X,'1— TOTAL VELOCIT
    >IES  —1 1— MID ISLAND  VELOCITIES --- 1  1 -- DEV - I'/)
9030 FORMAT(7X,2(A3,5X),A3,6X,A3,6X,A3,6X,(STAG E',3X,'STAG W',4X,'FAWN
    > E',3X,'FAWN W'^X.'N.CH.'.SX.'M.CH.'.SX.'S.CH.'^X, 'CUTOFF', 3X,
    >A3,5X,'DEV)
9040 FORMATCex.'MEAS.' ,4X,'COMP.' ,3X,'MEAS.' ,4X, 'VEL. ' ,5X, 'VEL. ' ,8X,
     'V ,8X,'V ,8X,'V ,8X,'V ,8X,'V' ,7X,fV ,7X,'V ,7X,'V' .SX.'MEAS.1 ,
 9041  FORMAT (18X,1 ME AS.1 ,6X,'COMP.' ,6X,'MEAS.' ,2X,3(5X,'VEL.' ) ,2X,
    >'VEL.' ,3(4X,'VEL.'),5X,'MEAS.' ,4X,'(C-M)' ,/)
 9050  FORMAT(1X,I2,F8.2,A1,1X,F7.2,F8.2,A1,1X,F6.2,3X,F6.2,4X,F6.2,4X,
    >F5.2,4X,F6.2,3X,F6.2,3X,F5.2,3X,F5.2,3X,F5.2,3X,F5.2,2X,F7.2,A1,
    MX.F5.2)
 9051  FORMAT(///,45X,! ST. GLAIR RIVER TRANSIENT MODEL1 ,//,52X,A9 ,15 ,//,
    >10X,A20,'  to  ' ,A20,40X,' RIVER VELOCITY  VERSION',//)
 9055  FORMAT(9X,I2,6X,F6.2,A1,3X,F6.2,6X,F6.2,A1,3(5X,F4.2),3X,F4.2,
    >3(4X,F4.2),4X,F6.2,A1,2X,F6.2)
 9060  FORMAT(/,'  AVE' ,F7.2,2X,F7.2 ,F8.2,2X,F6.2 ,3X,F6.2,4X,F6.2,4X,
    >F5.2,4X,F6.2,3X,F6.2,3X,F5.2,3X,F5.2,3X,F5.2,3X,F5.2,2X,F7.2,2X,
    >F5.2)
 9065  FORMAT(/,8X,IAVE',6X,F6.2,4X,F6.2,6X,F6.2,1X,3(5X,F4.2),3X,F4.2,
    >3(4X,F4.2),4X,F6.2,4X,F5.2)
C
      END
C
                                                                         A-K.U,

-------
                                  TABLES
  1.  Roughness coefficients for the St.  Clair River reaches.




A-2.  Hourly model output - water levels  option.




A-3.  Hourly model output - river discharge option.




A-4.  Hourly model output - river velocities option.




A-5.  Daily model output - water levels option.




A-6.  Daily model output - river discharge option.




A-7.  Daily model output - river velocities option.
                               85

-------
                            Table A-l.   Hourly model  output - water levels  option.
             FT.  CRATIOT      to  LAKE  ST. CLAIR  (SCS)
                                     1.0 HOUR TIME INCREMENTS
                                            ST.CLAIR RIVER HOURLY TRANSIENT MODEL



                                                        MAY  8, 1987
                                    WATER  LEVELS  VERSION
               179 REACHES
cr c
HR
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
JV..J
MEAS.
176
176
175
175
175
176
176
176
175
176
176
175
176
176
178
176
176
176
176
176
176
176
176
176
.46
.47
.47
.47
.47
.44
.47
.46
.47
.47
.48
.48
.47
.48
.47
.48
.46
.46
.46
.46
.46
.46
.46
.46
i
1 •
176
176
176
175
175
175
176
176
175
176
175
176
175
176
175
176
176
176
176
175
175
176
176
176
_ _, rnuDiiTcri i cuci c i
AL
.62
.62
.63
.63
.63
.61
.62
.62
.62
.63
.64
.64
.63
.63
.63
.63
.63
.62
.62
.61
.61
.61
.61
.61
SC MV DO
176.01
176.02
176.02
176.03
176.03
176.02
176.02
176.02
176.02
176.04
176.06
176.04
176.04
176.04
176.03
176.03
176.03
176.03
176.03
176 .02
176.02
176.02
176.01
176.00
176.29
176.30
176.30
176.31
176.31
176.31
176.31
176.31
176.31
176.33
176.33
176.32
176.33
176.32
176.31
176.32
176.31
176.32
178.32
176.31
176.31
176.30
176.29
176.28
176.42
176.43
176.43
176.44
176.45
176.44
176.44
176.44
176.44
176.47
176.46
176.44
176.46
176.46
176.44
176.46
176.43
176.46
176.46
176.44
176.44
176.44
176.42
176.41
MBR
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
.53
.65
.64
.56
.66
.65
.66
.66
.55
.59
.68
.66
.59
.66
.66
.67
.66
.67
.67
.66
.56
.66
.63
.62
fC
r Ij
MEAS.
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
.76
.77
.76
.79
.78
.77
.78
.78
.77
.83
.79
.77
.82
.77
.78
.79
.76
.81
.79
.78
.79
.77
.74
.74
FG
COMP.Q
6168.
6214.
6162.
6276.
6217.
6186.
6270.
6246.
6219.
6450.
6245.
6162.
6369.
6139.
6240.
6266.
6143.
6344 .
6272.
6222.
6289.
6230.
6124.
6162.
1
MCACiiDcri i c\/ci c Akin rnuc
| 	 AL — -|
176.60
176.69
176.60
176.61
176.61
176.61
17S.61
175.61
175.61
176.60
176.62
176.61
176.62
176.62
176.62
175.61
176.62
176.60
176.61
176.62
176.61
175.63
175.62
176.61
0.01
0.03
0.02
0.02
0.02
0.00
0.01
0.01
0.01
0.03
0.02
0.03
0.01
0.02
0.01
0.02
0.01
0.01
0.01
0.00
0.00
-.02
-.01
-.01
jwnc.1/ UE-VtUs) I^TW *.
| 	 SC---I I— -
176.02
176.03
176.03
176.03
176.06
176.04
176.04
176.03
176.04
176.07
1 76 . 06
176.06
176.06
176.07
176.01
176.06
1 76 . 06
176.04
176.06
176.06
170.04
176.07
176.07
176.04
- .01
- .01
- .01
- .01
-.02
-.02
-.02
- .01
-.02
-.03
- .01
-.02
- .01
-.03
0.02
- .02
-.03
- .01
- .02
- .03
-.02
-.06
-.06
- .04
176.
176.
176.
176.
176.
176.
176.
176.
176.
176.
176.
176.
176.
176.
176.
176.
176.
176.
178.
176.
176.
176.
176.
176.
uiTcrt r\cuT*TTnwc /r_u\_ 	
, VffT' W < fc.1/ 1/Ur.A'llAl'i*** \ V
- MV 	 | | 	 DO---)
27
27
26
28
29
28
28
29
28
31
30
29
30
29
29
29
29
29
30
29
29
29
30
28
0.02
0.03
0.04
0.04
0.03
0.03
0.03
0.02
0.02
0.02
0.04
0.02
0.03
0.02
0.02
0.02
0.02
0.03
0.02
0.01
0.02
0.01
- .01
0.00
176
176
176
176
176
176
176
176
176
176
176
176
176
178
176
178
176
176
176
176
176
176
176
176
.41
.42
.39
.41
.42
.41
.41
.42
.41
.46
.43
.42
.43
.42
.41
.43
.38
.42
.43
.42
.42
.42
.43
.40
0.01
0.01
0.04
0.03
0.02
0.03
0.03
0.02
0.02
0.01
0.04
0.02
0.03
0.03
0.03
0.02
0.06
0.03
0.02
0.02
0.03
0.02
-.02
0.01
	 l
"> i
| 	 MBR---)
176.52
176.53
176.53
176.66
176.64
176.54
176.64
176.54
176.55
176.67
176.56
176.65
176.56
176.54
176.64
176.56
176.64
176.57
176.66
176.66
176.66
176.66
176.63
176.63
0.01
0.01
0.01
0.01
0.02
0.01
0.02
0.02
0.00
0.02
0.02
0.01
0.03
0.02
0.02
0.01
0.00
0.00
0.01
0.01
0.01
0.00
0.00
-.01
AVE 176.47  17S.62 176.03  176.31 176.44 176.56 176.78
6233.  175.61   0.01  176.05  -.02  176.29  0.02 176.42  0.02 176.56  0.01

-------
                                     Table  A-2.  Hourly model  output -  river discharge option.
                  FT.  CRATIOT
              ST.CLMR  RIVER HOURLY TRANSIENT MODEL

                          MAY  8, 1987

to LAKE ST.  CLAIR  (SCS)
                                                    RIVER DISCHARGE VERSION
                                          1.0  HOUR  TIME  INCREMENTS
                                         179 REACHES
         |	RIVER PROFILE	|   I--TOTAL FLOW--|
co
HR
SCS
MEAS.
1
2
3
4
6
6
7
8
9
10
11
12
13
14
16
16
17
18
19
20
21
22
23
24
AVE
176
176
176
176
176
175
176
176
176
175
176
175
176
176
176
176
176
176
176
176
176
175
176
175
176
.46
.47
.47
.47
.47
. 44
.47
.46
.47
.47
.48
.48
.47
.48
.47
.46
.46
.45
.46
.46
.46
.46
.45
.46
.47
SC
COMP.
176
176
176
176
176
176
178
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
.01
.02
.02
.03
.03
.02
.02
.02
.02
.04
.06
.04
.04
.04
.03
.03
.03
.03
.03
.02
.02
.02
.01
.00
.03
FG
MEAS.
176
176
176
176
176
176
176
176
176
176
178
176
176
176
176
176
170
176
176
176
176
176
176
176
176
.76
.77
.76
.79
.78
.77
.78
.78
.77
.83
.79
.77
.82
.77
.78
. 79
. 76
.81
.79
.78
.79
.77
.74
.74
.78
  FG
 FLOW

  6168.
  6214 .
  6162.
  6276.
  6217.
  6186.
  6270.
  6245.
  6219.
  6460.
  6246.
  6162.
  6369.
  6139.
  6240.
  0206.
  6143.
  6344.
  6272.
  6222.
  6289.
  6230.
  6124.
  6162.

  6233.
 AL
FLOW

 6156.
 6146.
 6168.
 6184 .
 6217.
 6299.
 6234.
 6227.
 6234.
 6240.
 6269.
 6266.
 6266.
 6266.
 0247.
 6227 .
 6740.
 6281 .
 6284.
 6303.
 6272.
 6269.
 6261.
 6173.

 6237.
, 	
STAG E
Q
2467.
2467.
246S.
2480.
2487.
2480.
2493.
2496.
2486.
2624.
2E19.
2480.
2602.
2490.
2481 .
2494.
2477.
2499.
2614.
2499.
2601.
2498.
2471.
2467.
STAC W
Q
3700.
3716.
3711.
3730.
3746.
3733.
3764.
3768.
3742.
3806.
3792.
3728.
3769.
3747.
3732.
3760.
3726.
3764 .
3787.
3760.
3766.
3760.
3716.
3698.
FAWN E
Q
777 .
777 .
778.
782.
786.
793.
789.
786.
786.
790.
793.
788.
789.
790.
786.
706.
766 .
792.
796.
796.
791.
790.
787.
777 .
	 ,
FAWN W
Q
6380.
6380.
6386.
6400.
6433.
6471.
6469.
6447.
6447.
6468.
6488.
6468.
6466.
6469.
6462.
E44S.
6446.
6471 .
6496.
6498.
6481 .
6474.
6462.
6399.
1 	
N.CH.
Q
1973.
1962 .
1977.
1982.
1996.
2043.
1973.
2001 .
1991 .
1992.
1998.
2011 .
2014.
2001.
2004.
1989.
?00«.
7021 .
2007.
2020.
1999.
2011.
1997.
1963.
M.CH.
Q
828.
822.
830.
832.
838.
869.
826.
840.
834.
836.
838.
845.
846.
839.
841.
834 .
842.
848.
841 .
860.
837 .
843.
836.
821.
r LUfij — — —
S.CH.
Q
1008.
997 .
1009.
1010.
1020,
1061.
996.
1031 .
1010.
1014.
1011 .
1024.
1026.
1014.
1019.
1004 .
1022.
1033.
1020.
1041 .
1016.
1032.
1018.
994.
CUTOFF
Q
2349.
2340.
2347 .
2362.
2369.
2413.
2362.
2391 .
2377.
2382.
2379.
2400.
2400.
2396.
2399.
2387.
239B.
2410.
2396.
2416.
2396.
2409.
2399.
2374.
1 	 i/v.
sc
MEAS.
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
.02
.03
.03
.03
.06
.04
.04
.03
.04
.07
.06
.06
.06
.07
.01
.06
.06
.04
.05
.06
.04
.07
.07
.04
i 	 i
OEV
C-M
-0.01
-0.01
-0.01
-0.01
-0.02
-0.02
-0.02
-0.01
-0.02
-0.03
-0.01
-0.02
-0.01
-0.03
.02
- .02
- .0*
- .01
- .02
- .03
- .02
- .06
-0.06
-0.04
                                                       2488.
                             3746.
787.
5449.
1997.
838.
1018.
2386.  176.05 -0.02

-------
                                Table  A-3.   Hourly model putout  - river  velocities option,
             FT. GRAT10T
                                        ST.CLAIR RIVER HOURLY TRANSIENT  MODEL

                                                     MAY   8,  1987

                          TO LAKE ST.  CLAIR (SCS)
                                                                                RIVER  VELOCITIES VERSION
                                     1.0 HOUR TIME INCREMENTS
                                                                    179  REACHES
    |	RIVER PROFILE	I
                          |--TOT.  VEL.--I
Hft
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
  SCS
 MEAS.

175.46
176.47
178.47
175.47
17S.47
176.44
175.47
176.46
176.47
176.47
176.48
176.48
176.47
175.48
176.47
178 .48
176.46
176.46
176.46
176.46
176.46
176.46
176.46
176.46
  SC
 COMP.

176.01
176.02
176.02
176.03
176.03
176.02
176.02
178.02
176.02
176.04
176.06
176.04
176.04
176.04
176.03
178.83
176.03
176.03
176.03
176.02
176.02
176.02
176.01
176.00
  FC
 MEAS.

176.76
176.77
176.76
176.79
176.78
176.77
176.78
176.78
176.77
176.83
176.79
176.77
176.82
176.77
176.78
178 .79
176.76
176.81
176.79
176.78
176.79
176.77
176.74
176.74
 FC
VEL.

1 .06
1 .06
1 .06
1 .07
1 .06
1 .05
1.06
1 .06
1.06
1 .09
1.06
1 .06
1.08
1 .04
1 .06
1 .06
1 .04
1.08
1 .06
1 .06
1.07
1 .06
1 .04
1 .06
 AL
VEL.

0.67
0.66
0.67
0.67
0.67
0.68
0.68
0.67
0.67
0.67
0.68
0.66
0.68
0.68
0.68
0.67
0.67
0.68
0.68
0.68
0.68
0.68
0.68
0.67
1 	 1
STAC E
V
0.98
0.98
0.98
0.99
0.99
0.99
0.99
0.99
0.99
1.00
1.00
0.98
0.99
0.99
0.99
0.99
0.98
0.99
1.00
0.99
0.99
0.99
0.98
0.98
fflk> launnu
STAG W

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
V
.86
.86
.86
.86
.86
.86
.87
.87
.86
.88
.87
.86
.87
.86
.86
.87
.86
.87
.87
.87
.87
.87
.86
.86
1 VCl.UV.4 1 11
FAWN E

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
V
.49
.49
.49
.49
.49
.60
.49
.49
.49
.49
.49
.49
.49
.49
.49
.49
.49
.49
.50
.60
.49
.49
.49
.49
-3 	 !
FAWN W
V
0.93
0.93
0.93
0.93
0.93
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.96
0.94
0.94
0.94
0.93
1 	 ""J.
N.CH.
V
0.78
0.78
0.78
0.79
0.79
0.81
0.78
0.80
0.79
0.79
0.79
0.80
0.80
0.79
0.79
0. 79
0.80
0.80
0.80
0.81
0.80
0.80
0. 79
0. 78
\J l/CV. 1 «
M.CH.
V
0.62
0.62
0.62
0.53
0.63
0.64
0.62
0.63
0.63
0.53
0.53
0.63
0.63
0.63
0.63
0.63
0.63
0.64
0.63
0.64
0.63
0.63
0.63
0.52
VC.UU*. A 1 J
S.CH.
V
0.36
0.36
0.35
0.35
0.35
0.37
0.36
0.36
0.36
0.36
0.36
0.36
0.36
0.36
0.36
0.35
0.36
0.36
0.36
0.36
0.35
0.36
0.36
0.36
.C J 	 f
CUTOFF
V
0.76
0.76
0.76
0. 77
0.77
0.79
0.77
0. 76
0.77
0.78
0.77
0.78
0.78
0.78
0.78
0.78
0. 78
0.79
0.78
0.79
0.78
0.79
0.78
0.77
r 	 •"•
SC
MEAS.
178.02
176.03
176.03
176.03
176.06
176.04
176.04
176.03
176.04
176.07
176.06
176.06
176.06
176.07
176.01
176.06
176.06
176.04
176.06
176.06
176.04
176.07
176.07
176.04
DEV
C-M
-0.01
-0.01
-0.01
-0.01
-0.02
-0.02
-0.02
-0.01
-0.02
-0.03
-0.01
-0.02
-0.01
-0.03
0.02
-0.02
-0.03
-0.01
-0.02
-0.03
-0.02
-0.05
-0.06
-0.04
AVE 175.47   176.03  176.78
                           1 .06
                           0.67
                             0.99
                                                      0.87
                                      0.49
                                                                             0.94
                                                                                     0.79
                                                     0.63
0.36
                                                                              0.78   176.06  -0.02

-------
                              Table A-4.   Daily model output -  water levels  option.
oo
VO
                   FT.  CRATIOT
              ST.CLAIR RIVER TRANSIENT MODEL


                         JUNE 1986


to LAKE  ST.  CLAIR  (SCS)
WATER LEVELS VERSION
                                         24.0 HOUR TIME  INCREMENTS
                                         179  REACHES

1
2
3
4
6
6
7
8
9
10
11
12
13
14
16
16
17
18
19
20
21
22
23
24
26
26
27
28
29
30
AVE
bLi
MEAS.
176.66
176.66
175.66
176.64
176.66
176.70
175.69
176.69
176.69
176.69
176.71
175.74
175.76
175.77
176.78
176.74
176.76
176.75
175.76
176.78
175.77
176.74
175.74
176.73
176.74
175.72
176.71
176.73
176.73
176.73
176.72
1 	
AL
176
176
176
176
176
176
175
176
176
176
176
175
175
176
176
175
176
175
176
176
176
175
176
176
176
176
175
176
176
176
176
.82
.82
.81
.80
.83
.86
.84
.84
.84
.83
.86
.87
.89
.91
.91
.88
.90
.89
.89
.92
.90
.88
.88
.88
.88
.86
.86
.88
.88
.88
.87
	 LUMri
SC
176.26
176.26
176.23
176.22
176.26
176.26
176.24
176.26
176.26
176.24
176.26
176.26
176.30
176.31
176.31
176.30
176.32
176.29
176.30
176.33
176.29
176.28
176.30
176.32
176.31
176.27
176.28
176.30
176.32
176.31
176.28
n c.u LtvtLa 	
MV DD
176.66
176.66
176.62
176.52
176.66
176.66
176.63
176.55
176.55
176.52
176.64
176.52
176.59
176.69
176.58
176.59
176.61
176.57
176.58
176.62
176.57
176.66
176.59
176.62
176.60
176.66
176.68
176.60
176.62
176.61
176.67
176.69
176.69
176.66
176.66
176.69
176.68
176.66
176.69
176.68
176.65
176.67
176.66
176.72
176.72
178.71
176.73
176.76
176.70
176.71
176.76
176.69
176.69
176.72
176.76
176.73
176.68
176.71
176.73
176.76
176.76
176.70
	 |
MBR
176
176
176
176
176
176
176
176
176
178
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
178
176
176
176
.81
.81
. 77
.78
.81
.80
.77
.81
.60
. 77
.79
.76
.84
.83
.82
.86
.87
.82
.83
.87
.81
.81
.84
.89
.85
.80
.83
.86
.69
.87
.82
rvj
MEAS.
177
177
177
177
177
177
177
177
177
177
177
176
177
177
177
177
177
177
177
177
177
177
177
177
177
177
177
177
177
177
177
.06
.06
.01
.02
.06
.03
.01
.05
.04
.00
.02
.98
.07
.06
.05
.09
.11
.06
.06
.11
.03
.04
.08
. 14
.10
.03
.08
.10
.14
.12
.06
rv»
COMP.Q
6690.
6690.
6639.
6604.
6647.
6662.
6486.
6638.
6582.
6480.
6488.
6319.
6690.
6507.
6454.
6664.
6666.
6527.
6662.
6658.
6414.
6500.
6633.
6816.
6671.
6497.
6663.
6704.
6812.
6740.
6693.
1 	 MCAaune
| 	 AL— | I--
176.78
176.77
176.79
176.79*
176.81
175.83
175.82
175.83
175.82
175.82
176.85
176.91
175.90
176.90
176.91
175.92
175.89
176.89
175.90
176.90
176.90
176.90
175.89
176.88
176.88
176.88
175.88
175.87
175.86
176.87
176.86
0.04
0.06
0.02
0.01
0.02
0.02
0.01
0.01
0.02
0.01
0.01
-.03
- .01
0.01
0.01
-.04
0.02
0.00
-.01
0.02
0.01
-.02
-.01
0.00
0.01
-.02
-.02
0.00
0.02
0.01
0.01
176
176
176
176
176
176
176
176
176
176
176
176
176
178
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
U LtVtLi «IW
-- SC---I )--
.27
.28
.26
.26
.28*
.28
.28
.29
.29
.27
.30
.32
.36
.35
.34
.35
.36
.33
.33
.36
.32
.32
.34
.36
.34
.32
.34
.35
.35
.34
.32
- .02
0.00
- .02
- .03
- .02
- .02
- .03
-.03
- .03
- .03
- .04
-.07
- .06
-.04
-.03
-.05
- .03
- .04
- .04
- .03
- .02
- .04
- .04
-.03
-.04
-.06
-.06
-.04
- .03
-.03
- .03
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
v.uiwru icf L>CVI« i iumo i>_-
-- MV — -| | 	 DD---I
.62
.61
.49
.49
.62
.62
.50
.63
.52
.50
.63
.53
.68
.57
.56
.59
.69
.56
.67
.60
.56
.66
.68
.60
.68
.56
.67
.69
.60
.69
.66
0.03
0.04
0.03
0.02
0.03
0.03
0.02
0.03
0.03
0.02
0.02
-.01
0.01
0.02
0.02
0.01
0.02
0.01
0.01
0.02
0.02
0.01
0.01
0.02
0.01
-.01
0.00
0.01
0.02
0.02
0.02
178
176
178
176
176
176
176
176
176
176
176
176
176
176
176
178
176
176
176
176
176
176
176
176
176
176
176
178
176
176
176
.66
.66
.63
.63
.67
.66
.64
.67
.67
.64
.66
.66
.71
.70
.69
.72
.72
.69
.71
.73
.68
.68
.71
.75
. 72
.68
.70
.72
.74
.73
.69
0.03
0.04
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.00
0.01
0.01
0.01
0.01
0.02
0.01
0.01
0.02
0.01
0.01
0.02
0.01
0.01
0.00
0.01
0.01
0.02
0.02
0.02
-iw; 	 |
| 	 MBR---I
176.79
176.78
176.76
176.76
176.79
176.78
176.76
176.80
176.79
176.76
176.78
176.78
176.84
176.83
176.82
176.84
176.86*
176.81
176.82
176.86
176.80
176.81
176.83
176.87
176.85
176.80
176.84
176.86
176.87
176.86
176.81
0.02
0.03
0.02
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.00
-.01
0.00
0.01
0.01
0.01
0.02
0.01
0.01
0.02
0.00
0.00
0.01
0.02
0.01
0.00
0.00
0.01
0.02
0.02
0.01

-------
                      Table  A-5.   Daily  model output - river  discharge  option,
         FT.  GRATIOT
              ST.CLAIR  RIVER  TRANSIENT MODEL

                          JUNE  1986

to LAKE ST.  CLAIR  (SCS)
                               24.0 HOUR TIME INCREMENTS
                                         179 REACHES
       RIVER DISCHARGE VERSION
I	RIVER  PROFILE	|   (--TOTAL FLOW---J
                                    AL
                                   FLO*

                                   6090.
                                   6690.
                                   6561 .
                                   6602.
                                   6631.
                                   6557.
                                   6490.
                                   6627 .
                                   6587 .
                                   6487.
                                   6476.
                                   6325.
                                   6565.
                                   6513.
                                   6454 .
                                   6666.
                                   6657.
                                   6544 .
                                   6553.
                                   6643.
                                   6438.
                                   6497.
                                   6626.
                                   6808.
                                   6681 .
                                   6511 .
                                   6651 .
                                   6699.
                                   6806.
                                   6746.

                                   6592.
SCS
MEAS.
1
2
3
4
6
6
7
8
9
10
11
12
13
14
16
16
17
18
19
20
21
22
23
24
26
26
27
20
29
30
AVE
176
17S
175
176
175
175
175
176
175
175
176
175
175
175
175
176
176
175
176
175
175
176
176
176
176
176
176
176
175
175
176
.66
.00
.00
.64
.68
. 70
.69
.69
.69
.69
.71
.74
.76
.77
.78
.74
.76
.76
.76
.78
.77
.74
.74
.73
.74
.72
.71
.73
.73
.73
.72
SC
COMP.
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
176
.26
.26
.23
.22
.26
.26
.24
.26
.26
.24
.26
.26
.30
.31
.31
.30
.32
.29
.30
.33
.29
.28
.30
.32
.31
.27
.28
.30
.32
.31
.28
FG
MEAS.
177
177
177
177
177
177
177
177
177
177
177
176
177
177
177
177
177
177
177
177
177
177
177
177
177
177
177
177
177
177
177
.06
.00
.01
.02
.05
.03
.01
.06
.04
.00
.02
.98
.07
.06
.06
.09
. 11
.06
.06
.11
.03
.04
.08
.14
. 10
.03
.08
. 10
. 14
.12
.06
FG
FLOW
6690.
0690.
6639.
6604 .
6647.
6552.
6485.
6638.
6582.
6480.
6488.
6319.
6690.
6607.
6464.
6664 .
6666.
6527.
6562.
6658.
6414.
6600.
6633.
6816.
6671.
6497.
6663.
6704.
6812.
6740.
6693.
, 	
STAG E
q
2676.
2676.
2616.
2639.
2656.
2621.
2693.
2662.
2633.
2592.
2593.
2529.
2631.
2605.
2682.
2664 .
2666.
2613.
2623.
2661.
2B70.
2598.
2652.
2724 .
2671.
2600.
2662.
2681.
2724.
2698.
"--- AJ1_«I
STAG W
q
4016.
4016.
3928.
3963.
3987.
3934.
3893.
3980.
3952.
3891.
3892.
3796.
3947.
3907.
3873.
3997 .
3998.
3920.
3936.
3991.
38S4.
3899.
3978.
4086.
4006.
3902.
3995.
4022.
4086.
4046.
l_un;> 	 •
FAWN E
q
838.
838.
820.
828 .
830.
820.
812.
830.
824.
812.
810.
790.
821.
812.
805.
833.
831 .
817.
819.
829.
803.
812.
828.
861 .
834.
813.
832.
837.
860.
842.
FAWN W
q
6852.
5862.
6730.
6773.
6805.
6737.
6677.
6798.
6763.
6674.
5669.
6635.
5747.
6702.
6649.
6830.
5829.
5724.
6736.
6817.
6632.
6684.
5799.
6967.
6847.
6696.
5820.
5864.
6956.
6903.
1 	
N.CH.
q
2119.
2119.
2077.
2087 .
2090.
2069.
2047.
2091 .
2074 .
2048.
2040.
1988.
2049.
2038.
2024.
2094 .
2070.
2061.
2053.
2077.
2022.
2040.
2074.
2136.
2094.
2046.
2086.
2097.
2138.
2119.
	 l/CL. 1 «
M.CH.
q
901.
901 .
683.
886.
889.
881.
872.
890.
883.
872.
869.
849.
878.
872.
867.
896.
886.
881.
877.
889.
865.
872.
886.
912.
894.
873.
890.
896.
913.
906.
S.CH.
q
912.
912.
897.
896.
890.
849.
828.
857.
843.
836.
829.
771.
728.
714.
708.
740.
716.
748.
766.
734.
709.
736.
764.
788.
779.
766.
791.
789.
809.
808.
	 i
CUTOFF
q
2768.
2768.
2096.
2735.
2756.
2768.
2744.
2789.
2786.
2732.
2734.
2716.
2910.
2887.
2866.
2936.
2982.
2869.
2866.
2940.
2844.
2853.
2912.
2974.
2913.
2829.
2886.
2917.
2946.
2912.
i 	 "«-
MEAS.
176.27
176.26
176.26
176.20
176.28*
176.28
176.28
176.29
176.29
176.27
176.30
176.32
176.35
176.36
176.34
170.36
176.36
176.33
176.33
176.36
176.32
176.32
176.34
176.35
176.34
176.32
176.34
176.35
176.36
176.34
¥ — 	 |
DEV
C-M
-0.02
0.00
-0.02
-0.03
-0.02
-0.02
-0.03
-0.03
-0.03
-0.03
-0.04
-0.07
-0.06
-0.04
-0.03
-0.0S
-0.03
-0.04
-0.04
-0.03
-0.02
-0.04
-0.04
-0.03
-0.04
-0.06
-0.06
-0.04
-0.03
-0.03
                                             2637.
                            3956.
                                                               824.
                                               5769.
2072.
                                                               884.
                                                                       797.
                         2639.
                                                                                       176.32  -0.03

-------
                          Table A-6.   Daily  model output - river velocities option.
             FT.  GRATIOT
                                        ST.CLAIR RIVER  TRANSIENT  MODEL

                                                    JUNE  1986

                          to LAKE  ST.  CLAIR  (SCS)
                                                                                             RIVER  VELOCITY  VERSION
                                    24.0 HOUR  TIME  INCREMENTS
                                                                    179  REACHES
    |	RIVER PROFILE	I
                          I--TOT.  VEL.--I
                                                                                                                      	DEV	1
 1
 2
 3
 4
 6
 e
 7
 B
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
26
29
27
28
29
30
  SCS
 MEAS.

175.68
176.66
175.66
175.64
176.68
176.70
175.69
175.69
175.69
175.69
176.71
175.74
175.7S
176.77
175.78
175.74
175.76
175.76
176.76
176.78
176.77
176.74
175.74
175.73
176.74
175.72
176.71
176.73
175.73
175.73
  SC
 COMP.

176.26
176.25
176.23
176.22
176.26
176.26
176.24
176.26
176.26
176.24
176.26
176.26
176.30
176.31
176.31
176.30
176.32
176.29
176.30
176.33
176.29
176.28
176.30
176.32
176.31
176.27
176.28
176.30
176.32
176.31
  FC.
 MEAS.

177.06
17 7.06
177.01
177.02
177.05
177.03
177.01
177.06
177.04
177.00
177.02
176.98
177.07
177.06
177.05
177.09
177.11
17 7.06
177.06
177.11
177.03
177.04
177.08
177.14
177.10
177.03
177.08
177.10
177.14
177.12
 FC
VEL.

1 . 11
1.11
1 .00
1 . 10
1 .10
1 .09
1.08
 .10
 .09
 .08
1.08
 .05
 .09
 .08
1.07
1.10
1.10
1
1
1
08
09
10
1.06
 .08
 .10
1.12
AVE 176.72   176.28  177.06
1 . 10
1.08
1.10
1.11
1.12
1.11
                           1 .09
        AL
       VEL.

       0.71
       0.71
       0.69
        ,70
        .70
         0
0.
0.
0.69
0.69
  70
0.70
0.69
0.68
0.67
0.69
0.68
0.68
0.70
0.70
0.69
0.69
0.70
0.67
0.68
0.70
0.72
0.70
0.69
0.70
0.70
0.72
0.71

0.69
1 	 »
STAG E
V
1.03
1 .03
1 .01
1 .02
1 .03
1 .01
1 .00
1 .02
1 .02
1 .00
1 .00
0.98
1 .01
1 .00
0.99
1 .02
1 .02
1.01
1 .01
1 .02
0.99
1 .00
1 .02
1 .04
1 .03
1 .00
1 .03
1 .03
1 .04
1 .03
1.02
«1U 13U«WU
STAC W

0.
0.
0.
e.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
V
90
90
89
89
90
89
88
90
89
88
88
86
89
87
87
89
89
88
88
89
86
88
89
91
90
88
90
90
91
90
89
VC.LUV-1 1 1
FAWN E
V
0.51
0.61
0.60
0.50
0.50
0.60
0.49
0.60
0.60
0.49
0.49
0.48
0.60
0.49
0.48
0.60
0.60
0.49
0.49
0.60
0.4B
0.49
0.60
0.61
0.60
0.49
0.60
0.51
0.61
0.61
0.B0
CO 	 |
FAWN W

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
' 0
0
0
0
1
0
0
V
.99
.99
.97
.97
.98
.96
.96
.98
.97
.95
.96
.93
.96
.95
.94
.98
.97
.96
.96
.97
.94
.96
.97
.00
.98
.96
.98
.98
.00
.99
.97
| 	 nnx
N.CH.
V
0.82
0.82
0.80
0.81
0.81
0.80
0. 79
0.80
0.80
0.79
0.78
0.76
0.78
0.78
0.77
0.80
0.79
0.79
0.78
0. 79
0.77
0.78
0.79
0.82
0.80
0.78
0.80
0.80
0.82
0.81
0. 79
V 1/U.U 1 r-*
M.CH.
V
0.56
0.56
0.64
0.66
0.66
0.54
0.53
0.54
0.54
0.63
0.53
0.52
0.53
0.63
0.62
0.64
0.54
0.53
0.63
0.64
0.62
0.63
0.54
0.66
0.64
0.63
0.64
0.64
0.66
0.66
0.64
T »_t_ V^ A » *
S.CH.
V
0.31
0.31
0.31
0.31
0.30
0.29
0.28
0.29
0.29
0.28
0.28
0.26
0.26
0.24
0.24
0.26
0.24
0.25
0.26
0.25
0.24
0.26
0.26
0.27
0.26
0.26
0.27
0.27
0.27
0.27
0.27
WW |
CUTOFF
V
0.88
0.88
0.86
0.87
0.88
0.88
0.87
0.89
0.88
0.87
0.87
0.86
0.92
0.91
0.90
0.93
0.94
0.90
0.90
0.93
0.89
0.90
0.92
0.94
0.92
0.89
0.91
0.92
0.93
0.92
0.90
1 	
SC
MEAS .
176.27
176.26
176.26
176.26
176.28*
176.28
176.28
176.29
176.29
176.27
176.30
176.32
176.36
176.36
176.34
176.36
176.36
176.33
176.33
176.36
176.32
176.32
176.34
176.36
176.34
176.32
176.34
176.36
176.36
176.34
176.32
DEV
C-M
-0.02
0.00
-0.02
-0.03
-0.02
-0.02
-0.03
-0.03
-0.03
- .03
- .04
- .07
- .06
- .04
- .03
- .06
- .03
- .04
- .04
- .03
- .02
- .04
- .04
- .03
-0.04
-0.06
-0.09
-0.04
-0.03
-0.03
-0.03

-------
             ST.  CLAIR AND DETROIT RIVER CURRENT MEASUREMENTS
         Jan A. Derecki, Kathleen A. Darr, and Raymond N. Kelley
                                 ABSTRACT









     Velocities in the unregulated Great Lakes connecting channels,  the St.




Clair and Detroit Rivers, were continuously measured with current meters




during an experimental field program.  The program was initiated to  improve




determination of winter flows, when the accuracy of normal flow




determinations is affected by ice.  This study describes the experimental




results of continuous flow measurements using electromagnetic current meters




and an acoustic Doppler current profiler meter during the 1983-87 period of




data collection.  Verification of current meter results was provided by




model-simulated flows during open-water periods and flow transfer between




the rivers during winter, when at least one of the rivers was ice-free.




Results indicate that accurate estimates of mean river velocities (and




consequently discharge) can be obtained with a single well-placed current




meter.  However, the electromagnetic current meters are a direct-contact




single-point sensors that are affected by frazil ice during winter and weed




effects during most of the year, producing frequently questionable or




erroneous data.  The acoustic profiler is a remote sensor of velocities in




the overhead water column and is not affected by the frazil ice and weed




problems, producing superior data.
                                     92

-------
                               INTRODUCTION









     Flows in the unregulated Great Lakes connecting channels, the St. Clair




and Detroit Rivers (Figure 1), are normally determined by either stage-fall-




discharge equations or unsteady flow numerical models.  The calibration of




both the equations and models is based on periodic discharge measurements




taken over the years by the Corps of Engineers (COE) during the open-water




seasons (spring, summer, and fall).  Consequently, the calculated flows




normally exhibit good accuracy during ice-free periods, but may contain




large errors during winter months with extensive  ice cover.  The St. Clair




River is particularly prone to large ice jams because of practically




unlimited ice flow supply provided by Lake Huron  and an extensive river




delta that retards the passage of these ice flows.  Large  ice retardation of




flows in the St. Clair River  is relatively frequent during winter months.




The magnitude of larger ice retardations generally approaches about 20% of




normal flow  (1,100 c3/s), but in extreme cases has been observed to approach




50%.  The ice conditions  in the St. Clair and Detroit Rivers are different,




because of large difference in the upstream lakes and ice  supplies, and the




ice problem  also contributes  to large discrepancies in the simulated  flows




for these rivers.  In some cases,  these flow discrepancies exceed 20% of




total flow,  which  exceeds acceptable errors by an order of magnitude.









     The  St.  Clair and  Detroit River flows are determined  at  the Great Lakes




Environmental Research  Laboratory  (GLERL) with unsteady flow models,  with




the current  model  versions described by Derecki  and Kelley (1981) for the
                                     93

-------
St.  Clair River,  and Quinn and Hagman (1977) for the Detroit River.  These




rivers generally do not freeze over and are frequently free of ice during




winter.   During such ice-free periods the models are adequate for the




simulation of winter flow rates.  However, the models are not calibrated for




additional flow resistance due to ice, because of lack of proper data, and




tend to greatly overestimate the river flows during heavy ice accumulations.




The ice covers in these rivers are transient in nature, formed by the




consolidation of ice flows supplied by the upstream lakes due to ice break-




up by winter storms or spring  thawing.   In both instances proper




meteorological conditions are  required to produce heavy ice concentrations




in the rivers.  Generally southern storms are needed  to destroy  ice bridges




(Figure  2) which normally form at the heads of  the rivers and help keep  the




rivers free of ice, with subsequent shift to northerly winds that can  force




large amounts of ice flows  into  the river channels.









      Knowledge of  accurate  flows during  both  open-water and  ice-covered




periods  in  the St.  Clair and Detroit  Rivers  is  needed for a variety  of




scientific  and water resource  studies,  ranging  from water balance  and




chemical/biological loadings to  lake  regulation,  lake level  forecasts, and




winter navigation.   Large  discrepancies  in  winter flows are  associated with




 abnormal river profile on  the St.  Clair  and/or  Detroit Rivers  that  exist




 during  ice  conditions.   A  lack of measured  data on river velocities  and




 water levels  at  critical interim points  makes it impossible  to  determine




 whether  one or both river  models are in error during winter.   To address




 this problem,  a  field measurement program was implemented  in the St.  Clair




 and Detroit Rivers.  The program tests the  applicability  of  using
                                     94

-------
continuously  recording current meters to provide accurate velocity




measurements  on  an ongoing basis,  independent of river ice conditions.
                        FIELD MEASUREMENT PROGRAM









     The velocity and thus flow (discharge)  of the upper St. Clair and




Detroit  Rivers  were continuously measured using current meters.  Practical




requirements  dictated the  use of current meters without moving parts (to




avoid clogging),  that are  capable of prolonged operation (six months) at




frequent sampling rates.   The initial phase  of the field program on the St.




Clair River contained a pilot study, started in 1981, which provided for




familiarization and field  testing of equipment.  The actual data collection




program  started in 1983,  following additional resolution of encountered




problems.  During the Upper Great Lakes Connecting Channels Study (UGLCCS)




activities, begun in 1985, the field program was in its second phase,




started  in 1984,  with simultaneous measurements of point-velocities in both




rivers and selective measurements of vertical velocity profiles in one of




the rivers.  The current-meter stations were located in the upper portions




of both  rivers, close to the COE flow measurement sections.  These sections




of the rivers have fairly  steep hydraulic gradients and are normally free




from consolidated ice cover.  The meters were permanently deployed at the




river bottom and connected by cable to shore-located data recording




stations.  This arrangement permitted remote access by telephone via




teletype-recorder to both the meters and their individual data records.  The




operation of the current meters was monitored daily to detect and correct
                                    95

-------
any instrument problems in order to eliminate or reduce data gaps.  Ice




conditions in the rivers were also monitored during winter and checked as




needed by periodic ice surveys conducted by car or plane.  Collected




velocity data were stored in computer files and subjected to routine




preliminary analysis, including comparison with model-simulated flows.









Point Measurements









     After examination of the types of meters available, an electromagnetic




(EM) current meter was selected (Marsh McBirney, Model 585) for the first




phase of the program limited to the St. Clair River.  The standard meter was




modified to include an externally located recording system (Figure 3), which




provided unlimited continuous operational capacity at a cable-connected




recording system  (cassette tapes) located on the shore.  After field  testing




and several meter modifications, the in situ field operations were started




in September 1981, with  the deployment of two EM current meters in the upper




St. Clair River,  near  the head  of the river at Port Huron, MI (Figure 2).




The meters were  installed on the United States side of the river, outside of




navigation channel about 50 and 70 m from shore, in 13 and 15 m of water.




Meter  sensors were positioned 2 m above the bottom.  Deployment and




subsequent removal of  meters took place with the assistance of the USCGC




Bramble  and  a commercial diver, who guided the underwater operation.  The




Detroit  District  of  COE  also participated in the project by making discharge




measurements  during  the  open-water seasons.  These measurements were




intended to  provide  data for calibration of  the point-velocities  measured by




the meters with  the  mean river  velocity at the meter location.  They  were
                                     96

-------
not used in this study, because several conducted measurements either




encountered operational problems or indicated considerable discrepancy in




the data.










     The EM current meters in this study sampled ambient river velocity at




one-second intervals for Y- and X-axis velocity components, and an azimuth




angle.   These raw data were converted to the north and east velocity




components,  which were recorded with an accompanying azimuth angle at




15-minute intervals.  The 15-minute input data were stored in a computer




file and converted to hourly and/or daily resultant velocity magnitude and




direction.  The field seasons during the first phase of the program normally




covered late fall, winter, and spring months (November-June).  The meters




were redeployed for the 1982-83 and 1983-84 winter seasons.  However,




velocity measurements during the first two seasons contained some unresolved




problems and questionable data, and were excluded from this study.  High




quality river velocity measurements during the 1983-84 season (obtained with




one of the meters) coincided with the record St. Clair River ice jam of




April 1984.   This jam, which lasted nearly the entire month (April 5-29),




established records for both magnitude and lateness of occurrence, and




provided an excellent opportunity for testing the current-meter program.




This record ice jam and other aspects of the field measurement experiment




are discussed in previously published papers (Derecki and Quinn, 1986a,




1986b,  and 1987).









     The second phase of the study included simultaneous velocity




measurements in both rivers, starting in November 1984, with redeployment of
                                    97

-------
meters in the St. Clair River for the 1984-85 winter season.  The Detroit




River installation consists of two EM current meters, which were deployed in




August and tested during the summer of 1984 in the upper portion of the




river at Fort Wayne COE Boatyard (Figure 4).  The two meters were installed




outside of navigation channel about 60 and 90 m from the United States




shore.  Meters were placed in 12 and 14 m of water with upward positioned




sensors 2 m above the bottom.  Similar operation and deployment procedures




were used on both rivers, with the USCGC Mariposa or the USCGC Bristol Bay




providing assistance in the Detroit River.  During this phase of the study




the meters were not removed for the summer but were left operating




throughout the year to test the effects of weed transport and accumulation




on the velocity measurements.









Vertical Profile Measurements









      Initial point-velocity measurements indicated a need for vertical




distribution of velocities, and recent advances in acoustical




instrumentation  (Doppler-shift sensors) made such measurements practical.




Consequently, the St. Clair River installation was augmented during the




November 1984 redeployment with one acoustic Doppler current profiler (ADCP)




meter  (RD Instruments, Model 1200 RDDR), which permits measurements of




velocities at approximately 1 m intervals in almost the entire vertical




water  column  (Figure 5).  The ADCP meter was installed between the two EM




current meters,  about 60 m from shore in 14 m of water (Figure 3).  The




meter  housing was oriented horizontally and the upward-looking sensor was




connected by  a 90-degree elbow about 0.5 m above the bottom.  The ADCP meter
                                     98

-------
samples remotely vertical velocities in the overhead water column with




continuous sound waves (pings) from four beams at a rate of five times per




second, starting about 1 m above the sensor.  The raw data from the four




beams are averaged to produce Y- and X-axis velocity components, along with




an azimuth angle, for approximately 1 m increments of depth to the surface.




These data are converted to the north and east velocity components for the




1-m progressive data segments, and indicate velocities at the mid-points of




each vertical segment.  In a total water depth of about 14 m, this procedure




provided vertical velocity and direction values for 11 levels between




approximately 2.5m above the bottom and 0.5 m below the surface.  The




surface readings are eliminated because of  large data scatter at the air-




water  interface  (sound speed is about 5 times faster in water than in air).




The data were recorded at a cable- connected shore station at 15-minute




intervals (similar to the EM current meters).









     These remote-sensing instruments are expensive but provide continuous




measurement capability that can not be duplicated with a string of point-




measuring meters because of navigation and  ice problems near the surface.




The ADCP meter was removed from the St. Clair River in April 1986.  It was




used during summer on the Detroit River in  a demonstration of moving-boat




measurements  (in June),  and later (November) deployed in a normal-bottom




position  on that river in place of  the outer EM meter, 90 m  from shore




 (Figure 4).   The ADCP meter change  was made to provide vertical velocity




profile measurements  in  both  rivers.  Additional requirement for such data




 in the Detroit  River  are the  reversals of  its flow, which occur occasionally
                                     99

-------
because of the combined effects of storms on Lake Erie and ice jams in the




St. Clair River.
                           ST. CLAIR RIVER DATA









Electromagnetic Current Meter Records









     During the period of study, data were collected from the EM current




meters for nearly three and a half years on the St. Clair River.  These data




underwent preliminary analysis and comparison with model simulated flows.




The meters' operation was monitored daily to detect and correct any




instrument problems in order to eliminate or reduce data gaps.  The water




level gages on the river were also monitored daily to detect ice effects on




the river's profile and several ice surveys were conducted, when ice




problems were indicated by this process.









     Operation of the current meter program and monitoring of the meter




records indicated that frazil ice affects the operation of the EM current




meters.  Although frazil ice episodes in the St. Clair River (later




confirmed on the Detroit River) are relatively infrequent (about 5 to 10




occurrences on each river per winter), they drastically affect the meter




data, which have to be corrected by elimination of bad data records.  The




formation of frazil ice is a supercooling phenomenon, with distinct




characteristics, and can be easily identified.  During cold spells in the




winter months  (December-February), an additional sudden drop in temperatures
                                    100

-------
causes the formation of frazil ice.  This jelly-like ice formation is sticky




and adheres to objects; it coats the meter sensors, reducing their




sensitivity and producing low readings, at times approaching zero.  The




sudden drop in the EM meter velocities associated with frazil ice normally




starts after sunset (before midnight) and disappears rapidly after sunrise




(before noon).  However, severe episodes of frazil ice may last continuously




for a few days at a time.









     Serious weed effects on the EM current meter operations were not at




first apparent during  the initial phase of the program, limited to the St.




Clair River, because the meters were deployed in late fall (November), the




Lake Huron water is relatively clean, and the water velocity is high in the




upper river.  These factors contributed to reduced weed accumulation around




the sensors.  However, definite weed problems were encountered during the




subsequent prolonged operations, particularly in the summer and fall seasons




during continuous annual operations of the meter program.  Weed accumulation




reduces meter readings and requires divers to inspect and clean the sensors




at frequent  intervals  for reliable data records.  The EM current meter




velocity  records taken immediately before and after cleaning of sensors by




divers (on both rivers)  indicate that weed accumulation may reduce meter




velocities by as much  as 25-50%.   The records also show that this weed




accumulation may occur in only a few days, following deployment or cleaning




of meters.   However, weed accumulation is generally gradual and difficult to




identify  during initial  stages.  Since diver operations are expensive and at




times not feasible, this type of meter is not generally suitable  for
                                    101

-------
prolonged/continuous operations in rivers with high weed content,




particularly during the high weed transport season.









     Discussion and presentation of the EM current meter results in this




report are limited to selected episodes which illustrate the nature and




quality of data.  The first drastic episode of this type is the record April




1984 ice jam on the St. Clair River.  This jam vividly demonstrates the




effectiveness of the in situ current meter velocity measurements in




estimating the river flows.  The collection of high quality-current meter




data during the ice jam represents a major accomplishment and invaluable




information on the winter flow regime of the St. Clair River.  Results from




one of the current meters in operation at that time are indicated in Figure




6, which shows the effect of the ice jam on the upper river flows (velocity




and direction).  The meter velocity was reduced by about 50% during most of




April, changing near the river bottom at the meter location from about 1.0




to 0.5 m s~l.  Higher velocities at the beginning of May, following the jam




breakup, were produced by the increased head (water level difference)




between Lakes Michigan-Huron and St. Clair.  Records from the second meter




during this period showed other/additional effects, which were later




determined as weed effects.









     Verification of the current meter results on  the St. Clair River during




the ice jam episode is provided by  flow transfer from the Detroit River,




which was free  of ice  during April  and provided accurate flow simulation




with a numerical model.  Conversely, good agreement in derived flows by two




independent methods demonstrates that the St. Clair-Detroit River flow
                                    102

-------
transfer method is a very useful technique, provided that one of the rivers




is free of ice problems.   Comparison of flows transferred from the Detroit




River with the St. Clair River flows derived from the current meter




measurements is shown in Figure 7.  Extrapolation of the average river




velocity and discharge from the current-meter point-measurements is




discussed in the following paragraphs.  The transfer factor, shown in the




figure, represents a summation of the hydrologic factors that determine the




difference between the flows in the St. Clair and Detroit Rivers, namely,




the precipitation on Lake St. Clair plus tributary runoff minus evaporation




from the lake and the storage of water on the lake.  The agreement between




the meter and transferred flows is good during most of the March-May period,




particularly during the ice jam in April.  In the few instances when the two




sets of flows deviate substantially, it is probably the transferred flows




that are in error.  Thus, the high peak in transferred flow at the beginning




of May is caused by an extremely high storage of water on Lake St. Clair,




which appears to be overestimated.  Larger deviations at the beginning and




during the second week of March appear to be caused by model oversimulation




of the Detroit River flows, probably due to the presence of some ice in the




river  (March ice cover was not observed).









     Comparison of the current meter velocities with the St. Clair River




numerical model results during the first deployment period  (November 1983




-July  1984 field season), expressed as a ratio of model to meter velocity,




is shown in Figure 8.  As expected, the figure shows a complete breakdown of




the St. Clair River model following the development of the  ice jam in April.




During other times, the normal model-meter relationship is  reasonably
                                    103

-------
consistent and first-cut estimates of the average river velocity at the




meter location could be obtained by applying the velocity ratio to the




point-measurements of the meter (CM#1) .  The other meter (CM//2) shows weed




effects during March-June period, with reduced meter readings and




exaggerated model ice effects  in April.  The relationship between the normal




model and the weed-free meter  velocities for the 1983-84 deployment period,




after elimination of the bad model results in April, is indicated in Figure




9.  The two equations shown in the figure are for a linear regression of the




data points (least squares) and for a velocity forced through a zero-




intercept.  The equations agree closely and either one could be used to




produce acceptable average river velocities.  The equation constant from the




zero-intercept equation also agree very closely with the reciprocal of the




average model-to-meter ratio  (Figure 8).  The high correlation coefficient




(0.94) indicates  that over 88% (R  squared) of the variation between the




average river velocity  (simulated  by model) and the current-meter velocity




measured  at a single point near  the river bottom is explained by a simple




regression.









     Determination of  river  discharge,  based  on measurements  (Figure 7), was




made by multiplying derived  average river velocities from current meters by




the corresponding cross-section areas,  obtained from model computations.




These  areas were  readily  available, since  in  either velocity  extrapolation




method (ratio or  regression)  the flows  (discharge or velocity) were also




simulated by  the  models.   At the meter location, most changes  in the river




discharge are produced by corresponding changes  in velocity,  and errors




 introduced in the derived discharge due to  omission of  the corresponding
                                     104

-------
cross-section area changes are relatively small.   In the most extreme cases,




connected with prolonged-massive ice jams, the velocity and corresponding




discharge changes (reduction) could exceed SOX; similar cross-section area




and corresponding discharge changes would be under 5X.   During large ice




jams, to which the St. Clair River is particularly prone, the above meter-




derived flows represent a tremendous improvement over uncorrected model




results, which may oversimulate actual flows by a factor of two (CM#1 in




Figure 8).   Availability of similar measurements during such ice jam




episodes (provided meter readings are not affected by weeds), especially in




conjunction with flow transfers (if feasible), may provide acceptable flow




estimates.









     Simultaneous operation of the current meter program on both rivers




throughout the year and monitoring of the meter records indicated the




seriousness of weed effects on the EM current meters.  Severe weed effects,




especially after storms or other sudden surges, can be as dramatic as those




of frazil ice, but generally weed accumulation is gradual and may fluctuate




in severity.  An attempt was made to keep the EM current meters free of weed




problems with periodic cleaning of meter  sensors by divers, but was




generally unsuccessful.  Primarily because of weeds, the EM current meter




field program was generally unsuccessful  on the St. Clair River for




prolonged periods during  the second continuous deployment spanning several




field seasons and a few meter  changes because of instrument problems




(November 1984-June 1987).  This  is indicated by the model-meter velocity




ratios  shown  in  Figure  10.   Drastic weed-effect problems during a summer




season  (May-October,  1986)  are indicated  in Figure 11, which shows the
                                    105

-------
effects of small but gradually increasing weed accumulation during May and




June,  some recovery in July, and a massive-sudden weed clogging of the




meters' sensors in mid-August that remained in effect until the cleaning of




meters in November.









Acoustic Current Profiler Records










     The ADCP meter was deployed in the St. Clair River in November 1984 and




operated until April 1986, for nearly a year and a half long data period.




Data collected with this instrument are unaffected by the frazil ice and




weed problems, most likely because of the meters' physical characteristics.




Both the outgoing and reflected sound waves travel through any frazil ice




coating the sensor.  The same applies to weed accumulation.  Meter




characteristics also permit its deployment in a low-profile horizontal




position on a support structure designed to reduce weed accumulation.  Very




little weed accumulation was actually observed by divers during inspections.




This eliminates data gaps during winter and questionable or outright




erroneous data periods during heavy weed transport/accumulation (summer-fall




and after storms).  The upper St. Clair River vertical velocity profile




measurements obtained with this meter represent high quality,  unique data




not previously available on the Great Lakes connecting channels.  The




profiler is expensive but produces a data set which could not be duplicated




with a dozen of the EM current meters, since they could not be deployed at 1




m intervals and operated continuously near the surface throughout the year




(navigation and ice problems).  The quality of ADCP meter data is also




better.  The following discussion and presentation of the profiler results
                                    106

-------
is limited to a few data samples that illustrate the nature and quality of




collected data.









     The vertical distribution of velocity in the water column measured with




the profiler during June 1985 is indicated in Figure 12.  It gives the




progression of daily velocities at 11 levels with 1-m depth increments




between approximately 2.5m above the bottom and 0.5 m below the surface,




the practical  limits of vertical measurements for the water depth of about




14 m.  The figure shows a high degree of consistency between velocities at




different depths throughout the month.  This consistency indicates that good




estimates of velocities in the entire water column or at different depth




levels could be obtained with single point-measurements, such as those made




with the EM current meters (provided problems are eliminated) .  Highest




velocities normally occur near the surface, with a smooth progression of




increasing velocities from the bottom towards the surface, unless surface




flow is  opposed by  substantial wind shear.  With strong counter-current




winds  (southerly),  which are  generally  limited  to relatively  short periods,




the velocity near  the surface  is  occasionally retarded  sufficiently  so that




the highest  velocity occurs  2-3 m below the surface.  A more  frequent




occurrence  is  the  nearly uniform  velocity in the top water  layer  spanning  a




few  (occasionally  several) meters.









     The smooth transition of velocities between progressive  water  layers  is




 indicated even more vividly  in Figure 13, which shows  two  vertical velocity




profiles.   A typical high-velocity profile  is  shown by June 10,  1985,  which




was  selected because of sharp increase  in velocities on that  day (Figure
                                     107

-------
12);  the March 5,  1985, profile was added to show a typical low-velocity




profile.  Despite rapid change in velocities on June 10, the graph shows an




extremely smooth transition in the vertical distribution of velocities.  The




use of daily velocities provided some smoothing of the graphs, but generally




similar profiles are obtained for shorter periods (hourly and 15-minute




data).  To extend the profiles to the bottom and the surface, where




velocities could not be measured, these points were estimated and




incorporated  in the graphs.  The surface point was estimated by extending




the curve indicated by the preceding three measured points to the surface.




The bottom point was estimated by forcing a similar curve near the bottom




through a maximum-depth and zero-velocity intercept.  The profiles show  that




the vertical  velocity  distribution  is definitely exponential  (logarithmic),




which agrees  with  theoretical  derivations for  turbulent flow  (Prandtl, 1925;




vov Karman,  1934).  This  includes most  of the  depth but excludes  the




boundary  layer, where  the distribution  can  not be  logarithmic because  of




 theoretical  considerations.









      Verification of  the  high consistency of velocities at different  levels,




 indicated in the preceding figure,  is  shown in Figure 14 by a statistical




 relationship between  profiler velocities near the  bottom (bin 1)  and  the




 integrated average velocities (11 bins) for the eighteen-month period




 (November 1984 - April 1986).  The two equations shown in the figure  are for




 a linear regression of the measured data points (least squares)  and for a




 velocity forced through zero-intercept, which are nearly identical.   Either




 equation could be used to provide good estimates of the average vertical




 velocity.  The extremely high correlation coefficient for the least squares
                                     108

-------
linear regression (R-0.99) indicates that almost all (nearly 992) of the




variation between the average vertical velocity and a single point




measurement near the river bottom is explained by a simple regression.









     Comparison of the profiler velocities (near-bottom and integrated




average) with the St. Clair River average values (at this location), derived




with the unsteady flow numerical model for the November 1984 - April 1986




period and expressed as ratios of these velocities, is shown in Figure 15.




Larger variations or disagreements are seen during January and February,




when the model results contain substantial errors because of ice effects.




During other times the agreement is reasonably good and first-cut estimates




of the average river velocity (or eventually discharge) could be obtained by




applying the velocity ratios to the profiler measurements.  The relationship




between normal model and profiler velocities (excluding the bad ice-affected




model results) for the same period is shown in Figure 16.  Comparison of




results presented in Figures 14 and 16 indicates a considerable loss of




accuracy (23Z) for the estimates of average river velocity.  However, the




correlation coefficient for these estimates is still reasonably high




(R=0.87).  Even these estimates, obtained with a single meter,  represent a




large improvement over the model-simulated results during winter months with




significant ice problems.  The 1984-85 and 1985-86 winter seasons were




relatively uneventful (without large ice jams) and the ice effect indicated




in Figure 15 for the model-simulated flows represents approximately average




ice conditions.
                                    109

-------
                            DETROIT RIVER DATA









Electromagnetic Current Meter Records









     The period of record for the EM current meter data on the Detroit River




covered about 3 years (August 1984-June 1987).  Problems with weed




accumulation for the EM current meters became readily apparent on the




Detroit River during the second phase of the program, when continuous meter




operation throughout the year were begun in the summer of 1984.  The weed




content in  the Detroit River is higher and the river velocities are lower




than in the St. Clair River, contributing to more weed accumulation and




higher weed effects.  With  the higher weed content, more problems were




encountered in the operation of the  EM current meters on the Detroit River.




Primarily because of weeds, the EM current meter field program was generally




unsuccessful  on the St. Clair River  during most of the year  (summer and




fall), but  at least partially successful during winter, while on the Detroit




River  it was  completely unsuccessful throughout the  three annual periods.




This  is  indicated  in  Figure 17, showing  the model-meter velocity ratios  for




the period  of record.  Generally,  these  ratios are not stable for any




extended period  of  time.









Acoustic Current  Profiler Records









      The quality of the  ADCP meter data collected on the  Detroit River




 remained high, similarly to that  from the  St.  Clair  River.   Profiler




 operations were similar on both rivers,  with about the  same  water depths and
                                     110

-------
the vertical velocity measured at 11 levels or bins approximately 1 m in




depth.  The quality of the Detroit River profiler data is indicated in




Figure 18, showing the model-profiler velocity ratios for the period of




study (November 1986 - June 1987).  The figure shows small ice-effect




problems affecting model-simulated velocities during winter, which is




typical for the Detroit River.  Large ice jams occurred during this winter




on the St. Clair River, but its EM current meters were generally affected by




weeds and indicate biased ice effect.









     The current meter field experiment was terminated in June 1987 with the




removal of the EM current meters in both rivers.  The ADCP meter was left in




place in the Detroit River to continue the study of its flow reversals.




More detailed data analysis from the experimental field program will be




conducted next year.  Its primary purpose will be to develop a method for




correcting the unsteady flow model simulation during winter periods with




substantial ice problems.
                        SUMMARY AND RECOMMENDATIONS









     Flows in  the  St.  Clair-Detroit River system, the outlet from the upper




Great Lakes, are needed  for  a variety of hydraulic and water resource




studies.  Applications include hydrologic water balance, lake regulation,




lake level forecasts,  navigation,  transport of pollutants, recreation, and




consumptive water  use.   During the open-water season, acceptably accurate




estimates for  these  flows  are provided with available mathematical unsteady-
                                    Ill

-------
flow models.   However, these models may produce large errors during winter




months when rapid transport of ice flows causes formation of ice jams in the




lower river reaches.   The St. Clair River is particularly prone to large ice




jams because of the potentially large ice flow supply from Lake Huron and an




extensive river delta which retards the passage of these ice flows.  Flow




estimates during ice conditions can best be obtained from in situ current




meter measurements.









     Analysis of data collected during the 1983-87 period indicates that




acceptable estimates of river flows can be obtained with a single, well-




placed current meter.  However, the EM current meters are susceptible to




frazil ice problems during winter, and to weed effects during most of the




year, making them of dubious value on rivers with high weed content,  such as




the Detroit River.  These problems can be avoided with the ADCP meter,  which




is not affected by the frazil ice and weed effects and produces better




quality data for nearly the entire water column.   The vertical velocity




profiles measured with the ADCP meter show a high consistency in an




exponential vertical distribution of velocities.









     Because of the high quality of data for the  overhead water column,




deployment of the ADCP meters should be considered by agencies responsible




for flow measurements in large rivers,  such as COE for the Great Lakes




connecting channels.   Data from such meters could be collected either




continuously or on demand.  With proper calibration, the ADCP meters  may




provide a suitable substitute for the labor-intensive periodic measurements
                                    112

-------
now conducted by these agencies.  The quality of such measurements would




also be higher.
                             ACKNOWLEDGEMENTS









     From the beginning of  the current-meter field experiment a number of




people contributed to the successful operation of the project.  The authors




express their thanks to all  the contributors and acknowledge specifically




their contributions.  Several people from GLERL Instrument Lab, namely, H.K.




Soo, R.D. Kistler, R.W. Muzzi, T.C. Miller, and J.E. Dungan are acknowledged




for  their extensive  involvement in  instrumentation and field operations




throughout  the  project or its portions.  The study was conducted by GLERL's




Lake Hydrology  Group with Dr. F.H.  Quinn as Head, initially by A.J. Potok




with assistance from J.J. Kolodziejczak.  The initial program was




extensively modified and field tested prior to data  collection phase.  At




different times throughout  the study assistance in data analysis was




provided by M.  Moliassa, S.R. Bonema, W.P. Moore, B.M. Slizewski, B.E.




Short,  and  D.A. Buckwald, all part-time temporary employees.  Finally, the




authors  thank the U.S. Ninth Coast  Guard District for the deployment  support




provided on both rivers, without  which  this project  would not have been




possible.
                                    113

-------
                             LITERATURE CITED









Derecki, J.A., and R.N. Kelley.  1981.  Improved St. Clair River dynamic




     flow models and comparison analysis, NOAA Tech. Memo. ERL GLERL-34.









Derecki, J.A., and F.H. Quinn.  1986a.  Natural regulation of the Great




     Lakes by ice jams: a case study.  Proceedings, 4th Workshop on




     Hydraulics of River Ice,  Service Hydraulique, Ecole Polytechnique de




     Montreal, Quebec, June  19-20, Vol.  1, pp. F4.1-F4.24.









Derecki, J.A., and F.H. Quinn.  1986b.   The record St. Clair River ice jam




     of 1984, J. Hyd.  Eng..  112(12):1182-1194.









Derecki, J.A., and F.H. Quinn.  1987.  Use of current meters for continuous




     measurement  of  flows  in large  rivers.  Water Resour. Res.,




     23(9):1751-1756.









Prandtl,  L.   1925.    Bericht uber utersuchungen  zur ausgenbildeten




      turbulenz,  Z.  angew.  Math, u.  Mech..  5(2):136.









Quinn,  F.H.,  and J.C.  Hagman.  1977.  Detroit  and St. Clair  River transient




      models,  NOAA Tech.  Memo. ERL GLERL-14.









 von Karman,  Th.   1934.  Turbulence  and skin friction, J.  Aeronaut.  Sci..
                                     114

-------
                             LIST OF FIGURES
1.    St.  Glair-Detroit River system with location of water level gages.




2.    The point-measuring EM current meter and support.




3.    Locations of St. Glair River current meters and ice bridge.




4.    Location of Detroit River current meters.




5.    The ADCP meter with diagram showing its remote-sensing operation.




6.    St. Glair River EM meter data centered around ice jam, March-May, 1984.




7.    St. Clair and Detroit River flows, March-May, 1984.




8.    St. Clair River ratios of model to EM meter velocity during first




     deployment period, November 1983  - July 1984.




9.    St. Clair River ratios of model to EM meter velocity during second




     deployment period, November 1984  - June 1987.




10.  Relationships between St. Clair River normal model and EM meter




     velocity, November 1983  - July 1984.




11.  St. Clair River EM meter and model velocities showing weed-effect




     problems, May  - October  1986.




12.  Vertical distribution of daily velocity, June 1985.




13.  Vertical velocity profiles, March 5 and June 10, 1985.




14.  St. Clair River profiler relationship between near-bottom and average




     vertical velocity, November 1984  - April 1986.




15.  St. Clair River ratios of model to profiler velocity during deployment




     period,  November  1984  -  April 1986.




16.  Relationship between  St. Clair River normal model  and profiler




     velocity, November  1984  - April 1986.
                                    115

-------
17.   Detroit River ratios of model to EM meter velocity  during deployment




     period, August 1984 - June 1987.




18.   Detroit River ratios of model to profiler velocity  during the period  of




     study,  November 1986 - June 1987.
                                    116

-------
                                              .LAKE HURON
                                       Ft. Gratioti
                                     Dunn Papers
                              Mouth of Black River<

                                       Dry
                                                       fPt Edward
                                     Marysvillei
SCALE IN MILES
     •:


KILOMETERS
                Michigan
                                          St. Claire


                                                  \o

                                                  JK
                                                   CO

                                     ^ '.-• Roberts;//".'.
                                         ^Landing//. .
                                                 'Port Lambton
                                         3\lgonacJ
K
        St. Clair Shores^

         Grosse Point <


      Windmill Point.
                                       LAKE

                                     ST. CLAIR
    Ft.



             >La Salle


WyandotteW. Mpr
               Tecumsetv-
         ERIE
        \
                               FBelle River .
                           Ontario

-------
      PORT
     HURON
Blue Wtter
  Current
              113

-------
           • 4 Probe
            EM Sensor

           EM Meter

           - Support Clamp


             Support Legs

              Power/Data
               Cable
3 RR Wheel Support
       119

-------
DETROIT




   Fort Wayne
           120

-------
                  Horizontal
                   VWocJty
                 Components
Resolution
   and
 Bin Range
Cells (-
                      121

-------
                                                                                 200
                                      08
                                  U
                                  O
i i
i
                                  >   0.4
                                      0.0
VELOCITY
DIRECTION
                                                                                 12O
                                             11   II   11   10   20
                                             MARCH       APRIL
                                                           1984

-------
    8000
    4000-
—     0

^  3000
2
o
rr1  -3000
"•   8000
    4000-
            DETROIT (MOOEL)
             11    21

             MARCH
A ^ A
V v
TRANSFER FACTOR

y— . .-w_..
10    20

  MAY
                           123

-------
or
LJ
U_

O
5.0





4.5





4.0





3.5





3.0





2.5





2.0





 1.5





 1.0-





0.5-





0.0-
                                                CM*{
          1 6 11 16 2126 1 6 11 16212631 5 10 1520?5304 9 M 1924295 10 15202530 4 9 14 192*29 4 9 U 1924293 8 13182328 3 8 13182328

             NOV       DEC        JAN       FEB       MAR       APR       MAY       JUN       JUL

                  1983                                             1984

-------
                       1.5
                    o>
> i
                   U
                   o
                   _l
                   ui
                   ac
                   OJ
            = '0.941

            0.867*X
                                            0434- 0.832*X
                          o.o
 0.6       1.0       1.5

MODEL VELOCITY (m/s)
2.0

-------
LJ


cr
LJ

LJ
d

§
b
Q
     5.0
     4.5-
     4.0-
3.5-
     3.0-
>

r^   2.5 H
2.0 H





 1.5





 1.0 H





0.5
     0.0
                          MODEL/C.MJ1

                          MODEL/C.M.#2
         NDJFMAMJJASONDJFMAMJJASONDJFMAMJ

         1984              1985                           1986                   1987

-------
 1.75
 1.50-
 1.25-
 1.00-
0.75-
0.50-
0.25
0.00
                                         A  \ *V    *   I'S \
                                        / \ A- .V/-.Y • •   y .
	CM#1 VELOCITY
	CM#2 VELOCITY
	MODEL VELOCITY
       i  i  i  i   i  i  i  T  r  r  T   i  T  i  i   i  i  I  i   i  i  i  r  i  i  i  r  i  i  I   i  i  v i  T  r
     1  6  11  16 21 26  31 5  10  15 20 25 30  5  10 15 20 25 30 4   9  14  19 24 29  3  8  13 18 23 28 3  8  13 18 23 28
          MAY          JUNE          JULY         AUGUST      SEPTEMBER     OCTOBER
                                            1986

-------
u
3
IU
                  NEAR SURFACE
                   THRU
                  NEAR BOTTOM
11 II  1?
  JUNE
                               21  » It 27 2ft
                       128

-------
.  MARCH 5, 1985
                   JUNE |10, 1985
       0.6       1       1.5
         VELOCITY (m/s)
           129

-------
 <0   2.0-
O
o
LU
O

h-
o
CO
oc
LU
z
    1.5-
                R = 0.993
               Y = 0.879*X
            Y =-0.011 4- 0.888*X
0.0      0.5      1.0      1.5       2.0

 AVERAGE VERTICAL VELOCITY (m/s)

-------

-------
    2.0n
t  1.6H
O
O
—I
uu
>  1.0H
cc
uu
0.5-
LL
O
oc
Q.
    0.0
       0.0
           x
          X
    R = 0.869
    Y = 0.973*X
Y = 0.117  + 0.882*X
           0.5       1.0       1.5
          MODEL VELOCITY (m/s)
                      2.0

-------
      5.0




CO    4.5
UJ
t:


Q    4.0




      3.5-




      3.0-
 LJ
 0

 3
 LU


 _J
 LU
 Q
 O
U_
O

-------
5.0
4.5-
4.0-
3.5-
3.0-
2.5-
2.0-
 1.5-
 1.0-
0.5-
0.0
SOLID LINE - MODEL/BIN #1 RATIO
DASHED LINE - MODEL/INTEGRATED AVG. RATIO
    1 6 11 16 21 26 1 6
       NOV
             1986
    11 16 21 26 31 5 10 15 20 25 30 4 9  M 19 2* 1
    DEC         JAN        FEB
11 16 21 26 31 5 10 15 20 25 30 5 K) 15 20 25 30 * 9 M 19 24 29
 MAR        APR        MAY       JUN
       1987

-------
  DEVELOPMENT OF A SHALLOW WATER NUMERICAL WAVE MODEL FOR LAKE ST. CLAIR
                      David J.  Schwab  and  Paul  C.  Liu
     In the summer and fall of 1985, the National Water Research Institute




(NWRI) and GLERL participated in an extensive field measurement program in




Lake St. Clair.  The availability of high-quality, over-lake meteorological




data and the other physical measurement systems in a relatively flat and




shallow basin with sufficient fetch to generate substantial waves provided a




perfect opportunity to measure wave dissipation and the effect of waves on




resuspension in shallow water.  On June 19, 1985, scientists from GLERL,




NWRI, and AES  (Atmospheric Environmental Service) met and planned a joint




program entitled Wave Attenuation, Variability, and Energy Dissipation in




Shallow Seas (WAVEDISS '85) for the fall of 1985 in Lake St. Clair to make




these measurements.









     An inventory of readily available equipment limited the number of




recording stations to six, three each from GLERL and NWRI.  The GLERL




stations consisted of a single Zwarts transmission wave staff and a Datawell




Waverider radio transmitter.  The NWRI stations consisted of a triangular




array of capacitance wave gauges, a cup anemometer, and a SeaData recording




package.  The  instrumentation systems were attached to towers which were




guyed to a base anchored with railroad wheels.  The towers were deployed




along a transect parallel to the prevailing storm wind direction (Fig.  1).




One tower was  deployed on a perpendicular leg to give an indication of
                                    135

-------
cross-transect variability.  The separation distances between the towers




were based on expected differences in energy for a 10 m/s wind compared to




the sampling variability of the estimates of wave energy.  A sampling rate




of 4 Hz was selected based upon expected wave periods.  Wave record length




was set at 4096 samples, or about 17 minutes at 4 Hz sampling rate.  This




record length was a compromise between recording capacity of the SeaData




recorders and the statistical reliability of wave energy estimates.









      The GLERL instruments transmitted continuously to a shore station at




Stoney Point, Ontario where a microcomputer stored one wave record from each




station per hour.  The microcomputer was periodically interrogated by




telephone to retrieve the  data on the GLERL VAX computer.  The NWRI




instruments recorded burst samples on even-numbered hours when the




anemometer reading exceeded a present threshold value.  The triangular array




of capacitance gauges provided wave direction estimates  in addition to the




wave  energy spectrum.  The experiment ran  from September 20 - December 2,




1985. Nearly  continuous measurements of significant wave height and wave




period (at hourly intervals) were  obtained from  the GLERL towers.  The NWRI




towers recorded  data only  when  the wind  speed exceeded a preset  threshold




 (usually  7 ms'l)  so the  data covers only the higher wave height  periods.




Meteorological data consisting  of  hourly values  of wind  speed, wind




direction,  air temperature, and water  temperature were obtained  from  the




NWRI  meteorological buoy at 42.5 degrees north,  82.8  degrees east.  The




meteorological data was  used  to drive  the  GLERL-Donelan  numerical  wave




 prediction model on a  1.2  km  numerical  grid.  This model was developed by




 Donelan (1977) and used successfully to  predict  wave  height and  wave
                                     136

-------
direction in Lake Erie (Schwab et al., 1984) and Lake Michigan (Liu et. al.,




1984) .   The model is a parametric model based on a momentum balance equation




for the wave field.  The model predicts the two components of the wave




momentum vector and the phase speed of the peak energy waves.  From these




variables, significant waveheight, wave period, and wave direction are




derived.  In the mathematical formulation of the numerical model, the waves




are assumed to obey the deep water dispersion relation.  Refraction and




bottom  dissipation are ignored.









     As part of  the UGLCCS,  the  GLERL-Donelan model has been modified  to




account for the  effect of finite water depth on wave propagation by




incorporating  the  Kitaigorodskii et  al.  (1985) shallow water wave spectrum




along  with  a depth-dependent group velocity and a  simple  form  of bottom




friction.   This  shallow  water version of the model was also  run  with  the




same grid and  same wind  input as the deep water version.  The results  for




hourly values  of significant waveheight  are compared to  observations  at  the




 six towers in Figures 2  and 3.   The  statistical  comparison in  terms  of root




 mean square error and correlation coefficient  is  presented in  Table  1.









      As can be seen in Table 1 and Figures 2  and 3,  the  deep water version




 of the model provides quite acceptable estimates of waveheight,  even for the




 largest waves at the shallowest  stations.  The shallow water version of the




 model  tends to underestimate the highest waves at all stations.   The shallow




 water  model could be adjusted to better match the observed waveheights by




 decreasing the bottom friction  parameter, but the best it could do would be




 no better than  the deep  water model.
                                     137

-------
     Work is underway now to determine why the deep water model works in




Lake St. Clair.  One possibility is that the wind momentum input function in




the model is oversimplified and if it were formulated more realistically,




the deep water model would tend to overestimate the highest waves.  This




possibility is being investigated, but for now, the deep water model appears




to be quite acceptable for providing wave height estimates in Lake St.




Clair.
                                    138

-------
                             LITERATURE CITED









Donelan, M. A., 1977.  A simple numerical model for wave and wind stress




     prediction.  Unpublished manuscript.  National Water Research




     Institute, Burlington, Ontario, Canada. 28pp.









Kitaigordskii, S.A., V.P. Krasitskii and M.M. Zaslavskii, 1975:  On




     Phillips' theory of equilibrium range  in the spectra of wind-generated




     gravity waves.  J. Phys. Oceanogr.. 5, 410-420.









Liu, P.  C., Schwab,  D. J.  and Bennett,  J. R., 1984.  Comparison of a two-




     dimensional wave prediction model  with synoptic measurements in Lake




     Michigan.  J.  Phys.Oceanogr.   14:1514-1518.









Schwab,  D.J.  Bennett, J.R.,  Liu, P.C.  and Donelan,  M.A., 1984.  Application




      of a simple  numerical wave prediction  model  to Lake Erie.  J. Geophys.




      Res., 89(C3):3586-3592.
                                     139

-------
Table 1.  Comparison of Deep and Shallow Water Model Predictins with
          Measured Significant Waveheight
STATION
C3
Depth (m)
Data Points
Deep Water Model
rmse (m)
corr. coeff.l
Shallow Water Model
rmse (m)
corr. coeff.l
6,
.7
244

0,
0,

0,
0,

.11
.89

.13
.88
U2
7
.0
1366

0
0

0
0

.11
.93

.13
.93
C2
6
.4
237

0
0

0
0

.16
.88

.20
.85
Ul
5
.5
1478

0
0

0
0

.09
.93

.11
.91
Cl
3.
7
153

0.
0.

0.
0.

10
94

18
92
U3
4.4
1539

0.08
0.94

0.09
0.93
•"•correlation between predicted and measured values.
                                    140

-------
                               LIST  OF  FIGURES

Fig. 1.    Station locations for GLERL and NWRI wave towers during WAVEDISS
           '85.

Fig. 2.    Comparison of observed and predicted wave height for the deep
           water model.

Fig. 3.    Comparison of observed and predicted wave height for the shallow
           water model.
                                    141

-------
                 LAKE ST CLAIR
                 WAVE STATIONS
(
11
                                                     GLERL

                                                      U2".NWRI
                                                          C2-

-------
    Lake St. Clair 1985 Station VI
   CM
   10
•g
  m
13 O
CL
1=
     0     0.5    1      1.5     2
      Observed Waveheight (m)
    Lake St. Clair 1985 Station C /
   
-------
    Lake St. Clair 1985 Station
   CM
*'
 Q)
.n
O
     0     0.6    1      1.6     2
      Observed Waveheight (m)
    Lake St. Clair 1985 Station Cl
 \A^

I-
s.,
 3 O
 a
 E
          T*^
                •».
     0    0.6    1      1.6    2
     Observed Waveheight (m)
                                        Lake St Clair 1985 Station o
                                    .g>52
                                         s
                                         0    0.6     1     1.6    2
                                         Observed Waveheight (m)
                                        Lake St. Clair 1985 Station CZ_
                                                              ''
                                        0     0.5     1     15    2
                                         Observed Waveheight (m)
    Lake St. Clair 1985 Station

S1" U5
  O
Q.
    0     0.5    1     1.6    2
     Observed Waveheight (m)
    Lake St. Clair 1985 Station
                                                                           .
                                                                           0)
rs o
Q.
     0     05     1     1.5    2
     Observed Waveheight (m)

-------
                MODELING PARTICLE TRANSPORT IN LAKE ST. CLAIR
                        D. J. Schwab and A. H. Clites
     The results of a numerical circulation model of Lake St. Clair are




used to describe particle transport pathways in the lake in terms of




residence time and variability due to wind-induced circulation.




Specifically, we address the following questions. 1) What path does water




entering Lake St. Clair from one of the tributaries follow through the lake




before leaving at the Detroit River?  2) How long does it take?  3) How is




the path changed by wind-induced circulation in the lake?  4) For the




meteorological conditions during the summer and fall of 1985, what are




typical statistical distributions  of these pathways?









      In order  to  answer  these  questions,  a numerical circulation model of




the  lake was used.  The  numerical  model  is the same time-dependent rigid-




lid  model  developed by  Schwab  et al.  (1981) and used by  Schwab  (1983)  in




Lake Michigan  and Schwab and Bennett  (1987) in Lake Erie.  The basic




assumptions of the model are that  the  circulation is barotropic  and non-




divergent,  and that  nonlinear acceleration and horizontal  diffusion of




momentum are negligible compared to first-order  acceleration and Coriolis




 forces.   A simple quasi-linear formulation is  used to  describe bottom




 friction.   The model is forced by the hydraulic  flow  through the lake and




 by time-dependent wind stress at the surface.   The hydraulic flow is




 assumed to be constant in time at 5700 m3 s'l  and the  inflow is divided
                                       145

-------
among the tributaries as follows:  North Channel, 35Z; Middle Channel, 20Z;




St. Clair Flats, 20X; St. Clair Cutoff, 20X; Bassett Channel, 5X.









     Currents are calculated on a 1.2 km grid approximating the shape of




Lake St. Clair  (Fig. 1).  Simons and Schertzer  (1986) and Ibrahim and




McCorquodale  (1985) have also developed numerical circulation models for




Lake St. Clair  and have obtained essentially similar results for the wind-




induced  circulation.  However, the circulation  patterns generated by the




numerical models often differ considerably from the results of Ayers'




(1964) physical model of the lake  for  some wind directions.









     After  currents  are  calculated for each  grid square in the numerical




model,  another  model  is  used to  move  tracer  particles  through the lake.




The  particles are  assumed  to follow exactly  the vertically uniform  currents




without  sinking or diffusing.   The numerical model  for particle trajectory




calculations was  developed by  Bennett et  al.  (1983)  and used by Schwab  and




Bennett (1987)  in Lake  Erie.   The  model uses a  second-order method  to




compute particle  trajectories  and  takes special care  to realistically




represent the currents  near the shoreline.   As  shown by Bennett and Clites




 (1987) ,  this method is  far more accurate  than simple  first order  methods




yet is  only slightly more  complex computationally.   This  particle




 trajectory model is also used in the "Pathfinder" trajectory prediction




 system  (Schwab et al.,  1984)  that is used by the National Weather Service




 and the U.S. Coast Guard for tracking hazardous spills and  search and




 rescue missions on the Great Lakes.
                                       146

-------
     In order to answer questions about residence time for water entering




the lake from the various tributaries, several model runs were carried out




with idealized wind conditions.  First, the hydraulic flow in the absence




of wind was calculated (Fig. 2).  This pattern was then used as the initial




condition for a series of simulated storms with wind directions from the




eight compass points and peak wind speeds of either 10 or 20 ms'l.   Figure




3 shows a graph of the time dependence of the wind for the storms.  The




wind speed increases linearly  from zero to its peak value over a period of




two days and  then ceases.   Particles  from the five tributaries listed  above




plus the Clinton River,  the Clinton River Cutoff, and the Thames River were




released into the resulting circulation pattern  starting at  the beginning




of  the  second day, when  the wind speed was at exactly half its peak value.




One particle  was  released every three hours  for  one  day.  These particles




were  then  tracked through the  lake  and their residence  time  was compared  to




 the residence time  for particles in the  absence  of  wind (purely hydraulic




 flow).   The  results  for the 10 ms"1 storms  are  summarized in Figure 4. The




 dashed line  in the panel for  each tributary  represents  the  residence time




 for water  entering there in the no-wind case.   It can be seen that even




 though the average hydraulic residence time  for Lake St. Clair is about




 nine  days  (based on hydraulic flow of 5700 m3s-l-, surface area of 1100 km2,




 and a mean depth of 4 m), residence time for water from the individual




 tributaries  ranges from 4.1 days for the Middle Channel to over 30 days for




 water from the Thames River.  The calculated residence times for all  the




 tributaries  in the no-wind case are:
                                       147

-------
           North Channel:  4.5 days




           Middle Channel: 4.1 days




           St. Clair Flats: 5.0 days




           St. Clair Cutoff: 8.3 days




           Bassett Channel: 24.1 days




           Clinton River:  6.9 days




           Clinton Cutoff: 9.5 days




           Thames River: over 30 days









     The effect of wind on the residence time is greatest for the Thames




River and the Bassett Channel where a 10 ms'l wind from the SE, E, or NE




can decrease the expected residence time to less than 10 days for Bassett




Channel or 15 to 20 days for the Thames.  A 20 ms'l wind reduces the




residence time even more.   Winds from W, NW, or N tend to increase the




residence time of water from St. Clair Flats and more considerably from the




St. Clair Cutoff and Bassett Channel by moving the water eastward into the




relatively stagnant eastern part of the lake.  The residence time of water




entering the lake from the Clinton River is increased by NE, E, SE, and S




winds and decreased by SW, W, and NW winds.  Wind from almost any direction




except NW tends to decrease the residence time of water from the Clinton




Cutoff.









     The idealized wind condition calculation give some indication of the




variability in residence time that can be expected in Lake St.  Clair due to




wind-induced circulation,  but what kind of variability actually occurs




during the summer and fall?  To answer this question, hourly values of wind
                                      148

-------
speed and wind direction recorded at the CCIW met buoy at 42.5 degrees




north, 82.8 degrees east for the period 23 May, 1985 - 1 December, 1985




were used to drive the numerical circulation model.  The resulting




circulation patterns were stored and then used with the particle trajectory




model to calculate the paths of tracer particles released at the mouth of




each tributary once every six hours during the entire six month period.




The calculated tracks of the particles were then used to develop




probability plots of the likelihood of a parcel of water emanating from one




of the eight tributaries passing through a given area of the lake during




this period.  These plots quantify the wind-induced variability in the




pathway that water from one of the tributaries takes through the lake.  The




results of these calculations are presented in Figure 5 in terms of




probability contours.  The outermost line is the 99.9X contour, i.e., 99.9X




of the conservative particles released from that tributary remained within




this  contour for the 6 months simulation.  The next contour delineates the




area  in which 90X of the particles remained.  Remaining contours are at 10%




intervals.









      Most  of the water from the St. Glair River enters the lake through the




North Channel (35X).  According to the calculations, this water tends to




flow  down  the western shore of the lake and never  gets into the central or




eastern parts of the basin.  Water from the Middle Channel tends to remain




in  the western  third of  the lake, almost never entering the eastern half.




Water from St.  Clair Flats and the St. Clair Cutoff can be dispersed almost




anywhere  in the lake to  the south of the shipping  channel which connects




the  St. Clair Cutoff with the Detroit River.  A small amount of the St.
                                      149

-------
Clair inflow (5X) enters through Bassett Channel.  This water can pass




through any part of the eastern half of the lake depending on the wind




conditions.  The Thames inflow tends to be confined to the eastern and




southern shores before reaching the Detroit River and it can take a very




long time to get there (see Fig. 4).  Water from the Clinton River and




Clinton Cutoff is most likely to follow the western shore of the lake




southward with the most probable paths within 3 km of the western shore.









     Water quality measurements made in Lake St. Clair by Leach (1972 and




1980) showed two distinctly different areas in the lake.  In the




southeastern part of the lake, the water quality is dominated by the Thames




inflow, which is a major source of phosphate and other dissolved and




suspended material.  The central and western parts of the lake were more




similar to Lake Huron in terms of water quality than to the southeastern




part of the lake.  The pattern of water mass distribution mapped in Leach's




(1980) Figures 1-4 is very close to the combined patterns of the four main




St. Clair River  inflows and the Thames inflow in our Figure 5.  Bricker et




al.  (1976) examined the distribution of zooplankton in the western half of




the lake.  They  distinguished an area of biological and physiochemical




similarity along the western shore of the  lake that appeared to be




influenced more by the Clinton River than  the St. Clair River.  The shape




of this area matches quite well with the distribution pattern for water




from the Clinton River in Figure 5 here.









     To verify the circulation model and lend credence to the calculated




currents used in this study, the model was tested by comparing model output
                                      150

-------
to actual current data measured in Lake St. Clair in 1985.  Two separate




current data bases were gathered.  One involved the use of 5 drifting buoys




which were repeatedly launched and tracked in the lake.  The other was the




result of several synoptic current surveys utilizing electromagnetic




current meters.









     Currents predicted by the circulation model were used to simulate 16




drifter tracks.  Most of the tracks are about 2 days in length from various




portions of the lake.  In most cases, the model simulated the tracks




extremely well.  For the entire data set, the mean root mean square (rms)




of the drifter was 25Z greater than that of the calculated current track.




The directions compared favorably except for a few tracks near the mouth of




the Bassett Channel, where the model prediction was over 90 degrees




different in direction when compared with the observed track.









     The comparisons between current meter measurements and model-predicted




currents were even better.  In nearly 100 comparisons, 60% of the variance




is explained by the model prediction.  The model again seems to under-




predict the current speeds, here by about 30%.









     The results presented in this report are not water quality




calculations.  They only track conservative, non-dispersive tracers from




the mouths of the tributaries through the lake under various wind




conditions.  Work in another GLERL UGLCCS project couples the circulation




patterns calculated here with the TOXIWASP water quality model for Lake St.




Clair.  However, based on the comparisons with actual current measurements
                                      151

-------
presented above, the calculated currents provide a realistic depiction of




the wind-induced circulation in Lake St. Clair.
                                      152

-------
                               LITERATURE CITED









Ayers, J. C. 1964. Currents and related problems at Metropolitan beach,




     Lake St. Clair. Great Lakes Res. Div. Spec. Rpt. No. 20, Univ. of




     Mich., Ann Arbor, Michigan.









Bennett, J. R. and elites, A. H. 1987. Accuracy of trajectory calculation




     in a finite difference circulation model. J. Comp. Physics,




     68:272-282.









Bennett, J. R.,  Clites, A. H., and Schwab, D. J. 1983. A two-dimensional




     lake circulation modeling system: programs to compute particle




     trajectories and the motion of dissolved substances. NOAA Tech. Memo.




     ERL-GLERL-46, 51pp.









Bricker, K. S.,  Bricker, F. J., and Gannon, J. E. 1976. Distribution and




     abundance of zooplankton in the U.S. waters of Lake St. Clair, 1973.




     J. Great Lakes Res. 2:256-271.









Ibrahim, K. A. and McCorquodale, J. A. 1985.  Finite element circulation




     model for Lake St. Clair. J. Great Lakes Res. 11:208-222.









Leach, J. H. 1972. Distribution of chlorophyll a and related variables in




     Ontario waters of Lake St. Clair, pp. 80-86. In Proc. 15th Conf.  Great




     Lakes Res.  Internat. Assoc. Great Lakes Res.
                                      153

-------
Leach, J. H. 1980. Limnological sampling intensity in Lake St. Clair in




     relation to distribution of water masses. J. Great Lakes Res.




     6:141-145.









Simons, T. J. and Schertzer, W. M. 1986. Hydrodynamic models of Lake St.




     Clair. NWRI Contribution #86-xxx, National Water Res. Inst.,




     Burlington, Ontario.









Schwab, D. J. 1983. Numerical simulation of low-frequency current




      fluctuations in Lake Michigan. J. Phys. Oceanogr. 13:2213-2224.









Schwab, D. J. and Bennett, J. R.  1987. Lagrangian comparison of objectively




      analyzed and dynamically modeled circulation patterns in Lake  Erie. J.




      Great  Lakes Res.  (in press)









Schwab,  D.  J.,  Bennett,  J. R., and Jessup, A. T. 1981. A  two-dimensional




      lake circulation  modeling system. NOAA Tech. Memo. ERL-GLERL-38,  79pp.









Schwab,  D.  J.,  Bennett,  J. R., and Lynn, E. W.  1984.  "Pathfinder"--a




      trajectory prediction  system for the  Great Lakes. NOAA Tech.  Memo.




      ERL-GLERL-53,  37pp.
                                      154

-------
                                LIST OF FIGURES









Figure 1.   The 1.2 km numerical grid for Lake St. Clair. Arrows show the




           inflows and outflow used in the numerical model.









Figure 2.   Modeled hydraulic flow in Lake St. Clair. Streamlines indicate




           10Z increments of stream function from -2850 to 2850 ms'l.









Figure 3.   Wind history for residence time calculations.









Figure 4.   Calculated residence time for water from eight tributaries for




           simulated storm winds of 10 and 20 ms'l from eight compass




           points. Dashed lines indicate no-wind residence time.









Figure 5.   Probability distributions of water masses from eight tributaries




           in Lake St. Clair based on model simulations of particle




           trajectories for the period 23 May - 1 Dec, 1985.









Figure 6.   Verification of circulation model based on drifter and current




           meter measurements gathered in 1985.  Directions compare




           favorably except near the mouth of the Bassett Channel.   Model




           current speeds seem to be slightly underestimated based on the




           data.
                                      155

-------
Numerical Grid for
  Lake St. Clair
   No Wind
                     (2)
       156

-------
         Wind History for Residence Time Simulations
Ui
                                 Particles Released
                     1        2        3
                         Time (days)

-------
    Residence Time (days) in Lake St. Clair
              for a 10 ms"1 storm
   (dashed line indicates no-wind residence time)
North J
Channels
o
m M+ » it S
Middle g
Channels
o

St. Clair "
_. «
Flats o
o
/H» XNI * S
St. Clair g
Cutoff o
o




•


-


-•










...










—










_,










-










...


...







...


"


"1




-1





...


™

...
Bassett "
Channels


Clinton j
River s

x^i. O
Clinton ;
Cutoff o

-»-i O
Thames "
River s
o

.


-


-


-










'


















••4









































.«










i«_





     Residence Time (days) in Lake St. Clair
               for a 20 ms"1 storm
   (dashed line Indicates no-wind residence time)
            E  S W  N
North ;
g
Channels
o
fc 4* _1 Jl «
Middle p
Channels
Q
fH. /^| • S
St. Clair ;
Flats o
o

-


-


•





























;
..j

,


;





>






.~>







i
i»"l





!


"


:— •



Bassett "
~>uw W >_r kV Q
Channels

-*... o
Clinton J
River o


Clinton ;
Cutoff o
o

*


-


-
























•I


i

















!








I





...








""""



St. Clair
 Cutoff
Thames
 River
8 -
                            158

-------
North Channel
                                                       BasseU. Channel
   SL Clair Cutoff
                      159

-------
LAKE ST. CLAW
DRFTER STUDY
                                                            -Observed
                                                           D- Predicted
 3-6 September 1985
 LAKE ST. CLAIR
 CURRENT SURVEY
                                                             Observed

                                                             Predicted
  5 June 1985
  Depth Integrated Currents

-------
         TOTAL PHOSPHORUS BUDGET FOR LAKE ST. CLAIR: 1975 - 1980
      Gregory A. Lang, Julie A. Morton, and Thomas D. Fontaine, III
                                 ABSTRACT









     As part of the U.S. - Canadian Upper Great Lakes Connecting Channels




Study a total phosphorus budget was developed for Lake St. Clair.  An




unbiased ratio estimator technique was used to estimate annual loads and




variances from monitored hydrologic areas.  During the 1975-80 period, Lake




Huron was the major source of phosphorus to Lake St. Clair, accounting for




approximately 52% of the total annual load.  Hydrologic area loads, which




include diffuse and indirect point sources, contributed approximately 43% of




the total annual load.  The remaining 5% came from the atmosphere, shoreline




erosion, and direct point sources.  Of the hydrologic area loads, 85% could




be attributed to diffuse sources.  The Thames area contributed 58% of the




total hydrologic area  load, followed by  the Sydenham  (17%), the  Clinton




(9%),  the Ruscom (7%),  the Black (6%), the St. Clair  (3%), and the Rouge




(0.4%).  Over the  entire  six year period examined,  the lake's total input




and  output  of phosphorus were  nearly equal.   It was concluded that there was




no significant  net source or sink of phosphorus  in  Lake  St. Clair during the




1975-80 period.
                                     161

-------
                                INTRODUCTION









     Phosphorus budget calculations for the Great Lakes have generally




overlooked the dynamics of phosphorus transport into and through Lake St.




Clair.  The sum of Lake St. Glair's point and non-point inputs have simply




been included as part of the total load to the Western Basin of Lake Erie; the




assumption has been that the entire input of phosphorus to Lake St. Clair is




transported, unaltered, through the lake and into the Detroit River.  Recent




attention to the connecting channels of the upper Great Lakes, however, has




served as the impetus for  calculating a phosphorus mass balance for Lake St.




Clair.  The intent of this study was two fold: 1) to estimate the  total




phosphorus budget for 1975-80, and 2) to determine whether or not  the lake is




a net source or sink for phosphorus.









      Lake St. Clair is unique  against the pelagic backdrop of the  Great  Lakes




because  it  is shallow  (average depth = 3.3 m)  and has a very short average




hydraulic retention time  (9  days).  Lake St. Clair has a surface area of about




 1100 km^  and  a  drainage  area of  about 15500 km^.  We calculated the annual




 phosphorus  loads  entering this shallow, quick  flushing lake from a variety of




 sources  and compared  them to the amount of phosphorus  leaving the  lake  through




 the Detroit River.  From these comparisons along with  estimates of data  and




 load uncertainty,  we  were able to draw certain conclusions  concerning  the




 likelihood of net sources or sinks of phosphorus in Lake St. Clair.
                                     162

-------
                                  METHODS









     The total phosphorus budget for Lake St.  Clair includes external loads,




internal sources/sinks, outflow losses, and change in mass storage.  At steady




state, the sum of the external loads and internal sources/sinks should balance




the outflow loss.  External phosphorus loads to Lake St. Clair come from Lake




Huron, shoreline and streambank erosion, atmospheric sources, direct point




sources, and hydrologic  area loads.  Internal sources and sinks include




particle settling and  resuspension,  groundwater  input, bioturbation, and




aquatic macrophyte uptake  and  release.   The outflow  loss  is  through  the





Detroit River.









 External  Loading Estimates









      Lake Huron loads were taken directly from Yaksich et.  al.  (1982).  They




 estimated the average daily loading for the head of the St. Clair River by




 multiplying the flow weighted mean yearly phosphorus concentration by the




 reported average daily  flow per month.  This daily load was adjusted by the




 number of days  in the month to yield  a monthly  load.  Loading from shoreline




 erosion was  estimated by  multiplying  the length of Lake St. Clair shoreline by




 the  annual loading rate of phosphorus per kilometer of shoreline  for  the Lake




 St.  Clair basin (Monteith and Sonzogni  1976).   Loading from streambank erosion




  (along the St.  Clair  River) was  assumed to be negligible because  the  total




  loading to the  Great  Lakes  from streambank erosion  is <  4% of  that  contributed




  by shoreline erosion on the U.S.  side of the Great  Lakes (Knap and  Mildner





  1978).
                                      163

-------
     Direct atmospheric loading was estimated to be the average (14.0 MT/yr)




of values obtained during the field seasons of 1975 (Delumyea and Petel 1977,




measurements made at stations around Southern Lake Huron) and 1981




(Klappenbach 1984, measurements made at Mt. Clemens, Michigan).  Direct and




indirect point source load estimates were compiled from the International




Joint Commission  (1982),  the Great Lakes Basin Commission  (unpub. data), and




the  Ontario Ministry of  the Environment  (1985).









Hydrologic Area  Loads









      Seven hydrologic  areas were defined using the convention of Hall et al




 (1976): the Black,  St.  Clair complex, Clinton,  and Rouge complex (does not




 include the Rouge River) areas in Michigan, and the Ruscom,  Thames, and




 Sydenham areas in Ontario (Fig. 1).  Hydrologic area loads include diffuse




 loading from land areas  that drain into a  tributary or directly into Lake St.




 Clair, and indirect point source inputs.   Indirect point source loads are




 those  which discharge upstream  of a  river  monitoring station.  Their entire




 load was  assumed to be  transported to Lake St.  Clair.  Direct point sources




 were  defined  as  those which discharge downstream  of a river monitoring  station




  or  directly  into the  St. Clair River or Lake St.  Clair.   Input  from  these




  sources were  not included as  part of the hydrologic area  loads.   Several  small




  (< 1 MGD) point source  discharges may have been omitted from our hydrologic




  area analysis since scant information existed for them during 1975-80.




  However, analysis of recent municipal point source data from STORET (U.S.  EPA)




  and industrial point source data from the Ontario Ministry of the Environment




  (1985) shows that these sources presently represent about IX of all of the
                                      164

-------
point source flows.  Therefore, their omission should have little or no effect

on calculation of loads from hydrologic areas.



Each hydrologic area load equaled the sum of  its monitored and unmonitored area loa



Monitored areas.  Hydrologic area loads from  monitored  areas were calculated using

estimate the average daily  load at  the mouth, adjusted  to minimize  the

variance associated with the flow component,  as  follows:



                                          Sxv/(nMvMx))
                         Ux •  My/Mx •                                        (1)
                                       (1 + Sx2/(nMx2)
 where Uy is  the unbiased estimate of the daily load at the mouth,  Ux is the

 mean daily flow for the year,  My is the mean daily loading for the days for

 which concentration data exists, Mx is the mean daily flow for the days for

 which concentration data exists, n is the number of days for which

 concentration data exists, and
                                n
                                        - nMyMx
                                  _
                         Sxy =                                              (2)
                                       n-1
                                     165

-------
                               n
                               S  Xj.2 - nttx2
                              i-1     _
                        SX2 - -
                                     n-1
where Xj. is the individual measured  flow  for each day for which  concentration

data exists and Yj.  is  the loading for  each  day  for which concentration  data

exists  (calculated  as  the product of the  individual  measured flow and the

total phosphorus  concentration) .   Since this  load includes  any indirect point

sources,  the  diffuse component of the  hydrologic area load  from a monitored

area was  estimated by subtracting the  indirect point source inputs from the

total  calculated load.



      The estimated mean square error  of  the estimated load  (the  square root of

 which  is the estimated standard  error of the mean) was also calculated using

 the ratio estimator method,
             MSE - My2 -  [  1/n •  (SX2/MX2 + Sy2/My2  -  Sxy/(MxMy))

                   + l/n2 • (2-(Sx2/Mx2)2 - 4.Sx2/Mx2.Sxy/(MxMy)

                   + (Sxy/MxMy)2 + Sx2/Mx2.Sy2/My2)]                         (4)


  where E is the estimated mean square error of the load estimator and Sy2 is

  calculated analogously to SX2 .  In an attempt to quantify some of the

  uncertainty of  these loads  and  to statistically compare the total annual

  external load with  the outflow  loss, the root mean  square error (RMSE) is used

  in a later section  to  estimate  90% confidence intervals around the individual

  tributary  loads,  the  total  external  loads,  and  the  outflow  losses.
                                       166

-------
     Three hydrologic areas (Clinton, Thames, Sydenham) were 100X monitored




during the entire study period.  Percent monitored represents the percent of




the hydrologic area that is monitored.  The Black area was not monitored for




years 1975-77 and 100X monitored for years 1978-80.  The St. Clair and Ruscom




areas were 36% (Belle R.) and  18Z  (Ruscom R.) monitored, respectively, during




the study period.  The Rouge area  was not monitored.  On an areal basis, 832




of the Lake St. Clair basin was monitored during the study period.









     Flow data were obtained from  the U.S. Geological Survey  (1976-82) and  the




Water Survey  of Canada  (1976-81).   Flow measurements recorded at  gaging




stations located  upstream of the  river  mouth were  corrected to  the river mouth




by multiplying the  gaged flow  by  the ratio of the  entire drainage area to the




gaged drainage area as  described  in Sonzogni et al.  (1978).   Phosphorus




concentrations were obtained from the U.S. Environmental Protection  Agency's




STORET  system,  the  U.S.  Geological Survey (1976-82), and the  Ontario Ministry




of  the  Environment  (1975-80).   In most  cases, water quality monitoring




 stations were located at or near the river mouth.   It  was  assumed that  these




 phosphorus  concentrations  equaled the concentrations at the mouth.









      Unmonitored areas.  Some hydrologic areas and individual tributaries were




 not sufficiently monitored (fewer than 6 samples  per year) over some or all of




 the 1975-80 period.  In addition, some land areas drain directly into Lake  St.




 Clair or the St. Clair River, and therefore cannot be monitored.   Diffuse




 loads for these watersheds were estimated by calculating a diffuse unit area




 loading (UAL) (the diffuse load per unit area per unit time)  for a monitored




 area with basin characteristics similar to the unmonitored area.   The
                                     167

-------
selection of a representative monitored area was based on soil texture,




surface geology, runoff characteristics, land use, and proximity to the




unmonitored area.  The diffuse load for the unmonitored area was estimated as




the product of the diffuse unit area load from the monitored area and the




unmonitored area as described in Sonzogni et al.  (1978).  Indirect point




source discharges in unmonitored areas were added to the estimated diffuse




load to yield total loads for the unmonitored area.
Total External Load









     The  total external phosphorus  load  to Lake St. Clair equaled the sum of




loads from the total hydrologic  area, atmospheric sources, shoreline erosion,




direct point sources, and  the Lake  Huron.  The variance associated with the




total external load was not known because some hydrologic areas were partially




or  totally unmonitored and variance estimates were not available for the




atmospheric, erosion, direct point,  or Lake Huron loads.  However, an estimate




of  the variance associated with  the total external load was made using the




following procedure.  Annual load probability distributions for each monitored




tributary and the  outflow  from Lake Huron were constructed with means set




equal to  their estimated annual  loads.   The variances of the tributary load




distributions were calculated from  their estimated RMSE (eq. 5).  The variance




of  the Lake Huron  load distributions was assumed to be constant and equal to




the variance of the Huron  load averaged  over the six-year period.  For each of




the six years, values from each  distribution were randomly selected and summed




to  give a single estimate  the total annual external load (Figure 2).  This
                                    168

-------
procedure was repeated 12 times to yield a combined probability distribution




for each year's total external load.  A sample size of 12 corresponds to the




average number of times per year that each tributary was sampled.  The




standard error of the total annual external load was calculated from each




year's combined probability distribution for the purpose of statistical




comparison with the  outflow data.









Outflow  Loss









      Annual  phosphorus  loss  through the Detroit River and the mean square




 error associated with this loss were estimated using the unbiased ratio




 estimator technique.  Phosphorus concentrations were measured along a transect




 at Windmill Point near the head of the Detroit River by the Michigan




 Department of Natural Resources (DNR) and recorded in EPA's STORET data base.




 There are ten stations, each representing 10X of the flow of the river.  The




 data from these stations were limited to non-winter months (usually April




 through October).   Phosphorus data  for the winter months were measured at the




 Detroit municipal water treatment facility at Water Works Park and recorded  in




 the  U.S. Geological Survey  (1976-82).  Water quantity data were  taken  from the




 results  of  a hydrodynaroic model of  the Detroit River  (F. H.  Quinn, Great Lakes




 Environmental  Research Laboratory,  personal communication).










  Internal Sources and Sinks









       Scant quantitative information exists concerning the annual internal




  input or loss of phosphorus in Lake St.  Glair.  It is estimated that some
                                      169

-------
sources and sinks vary seasonally (e.g., particle settling and resuspension




and aquatic macrophyte uptake and release), while others remain more or less




constant throughout the year (e.g., groundwater input and bioturbation).  For




the purpose of annual budget calculations, the amount of phosphorus added to




or lost from the lake via internal sources and sinks was estimated as the




difference between annual external loads  and annual outflow losses.
                                   RESULTS









      The annual estimated river mouth phosphorus  loads  from monitored




 hydrologic areas to Lake St.  Clair as well as the annual  estimated Detroit




 River outflow losses are presented in Table 1.   Table 1 also includes the




 estimated RMSE of the estimated load and the 90%  confidence intervals around




 the estimated load.  The RMSE and the Student's t-statistic were used to




 calculate the 90% confidence intervals (Remington and Schork 1985).









      The total annual phosphorus budget for Lake St. Clair is presented in




 Table 2.  On the average, Lake Huron contributed a major  portion (about 52%)




 of the external total phosphorus load.  The contribution of hydrologic areas




 was also significant and equaled approximately 43% of the external phosphorus




 load to Lake St. Clair.  The remaining 5% of the total annual load came from




 the atmosphere  (0.5%),  shoreline erosion  (2.6%), and direct point sources




 (1.91).  Average phosphorus dynamics of Lake St. Clair during the 1975-80




 period  are summarized  in Figure  3   In the  diagram,  the relative proportion of




 inflows  and  outflows are approximated by  the thickness of  the arrow  shafts.
                                     170

-------
     Hydrologic area loads originating from Canadian sources averaged 82% of




the total hydrologic area load to Lake St. Clair.  The Thames and the Sydenham




areas contributed 92% of the total Canadian hydrologic area load.  The Black




and Clinton areas were responsible for 83% of the total U.S. hydrologic area




load.  The largest individual hydrologic area loads originated from the Thames




area.  In all years except 1975, loading from the Thames exceeded 50% of the




total hydrologic area load (the  six-year average contribution was 58%).  The




Thames area load was followed by the  Sydenham (17%), the Clinton (9%), the




Ruscom (7%), the Black  (6%), the St.  Clair  (3%), and the Rouge (0.4%).









      Over  the  six  year  period  investigated,  about  85%  of the total hydrologic




area load  was  calculated to  originate from  diffuse  sources.  The remaining  15%




came from  indirect point sources.   The diffuse  portion of  the Thames  and




Sydenham area  loads accounted  for  most of the  total diffuse load (62% and 20%,




respectively).   The Clinton and Thames areas contributed a majority  of the




total indirect point source load (43% and 35%,  respectively).









 Internal Sources and Sinks









      Assuming estimated loads presented in Table 2 are an accurate




 representation of the actual loads and assuming steady state conditions, then




 the difference between external loads and outflow losses is a measure of the




 internal sources/sinks of phosphorus  in Lake St. Clair.  The total annual




 incoming load falls within  the  90% confidence  interval around the Detroit




 River outflow loss for all  years  except 1976 (Figure  4).  The six-year mean




 external  load (3133 MT/yr)  and  loss  (3148 MT/yr) were not  found to be
                                     171

-------
statistically different at the 10X level of significance (t-test, Remington




and Schork 1985).  The root mean square difference between annual external




loads and outflow losses was only 7X of the mean external load (5* excluding




1976 values).  Therefore, given the above assumptions and results, there




appeared to be no net internal source or sink of total phosphorus in Lake St.




Clair during 1975-80.
                                  DISCUSSION









      The  second objective of this study was  to determine  whether  or not  Lake




 St.  Clair is  a net source or sink for phosphorus.   The  total  incoming load




 fell within the 90% confidence interval around the Detroit River  loss for all




 years except  1976; and,  averaged over 6 years, external loads were not




 significantly different from outflow losses;  implying no  net  internal source




 or sink of phosphorus in the lake during 1975-80.   Because the net internal




 sources/sinks were calculated as the difference between the incoming loads and




 outflow losses, their validity is limited by the accuracy and precision of the




 load/loss estimates.  If loads and losses are sufficiently accurate and




 precise,  then the conclusion that net internal sources/sinks  were negligible




 during the study period seems reasonable.









      It  is difficult to determine how accurate the estimated loads and losses




 are  since actual  loads and  losses can never be known exactly.  Sonzogni et.




 al.  (1978) note that the RMSE terms  in Table  1 are useful for statistical




 comparisons, but  they do not necessarily reflect how close the estimated load
                                     172

-------
is to the true load.  The RMSE is an estimate of the error determined from a




limited number of daily samples, based on the premise that the true annual




load can be determined by sampling flow and concentration at the river mouth




each day of the year.  In addition, the method assumes that instrument and




measurement errors can be neglected and that instantaneous flow/concentration




measurements are true representations of the tributary conditions on that day.









     As briefly stated previously, Dolan et. al.  (1981) evaluated 10 tributary




load estimator methods and concluded  that the ratio estimator method used in




the present study was the most  suitable for application in the Great Lakes




basins.  They recommended that  this method be used to estimate tributary loads




for  total  phosphorus when concentration data are  limited  and the daily flow




record is  available.  The accuracy of the streamflow data used in this study




was  rated  "fair"  to "good" by the  U.S. Geological Survey.  "Good" means that




about  95%  of  the  reported daily discharges are  within 10% of the actual values




and  "fair" within 15%.   Chemical data represent (as much  as possible) water




quality conditions  at  the  time  of sampling,  consistent with available sampling




techniques and methods  of  analysis.   The  phosphorus  loads from Lake Huron,




estimated by  Yaksich et. al. (1982),  were consistent  (trends and magnitude)




with other external loads  to Lake Erie for 7 water  quality  constituents during




 1970-80 and are thus,  for  lack of better  criteria,  considered representative




 of the true Lake Huron loads.  Therefore,  given that 90%  of the hydrologic




 area load was from monitored tributaries  and that 95% of  the  total  external




 load was from hydrologic areas and Lake Huron,  we conclude  that  the calculated




 total external phosphorus loads and losses are at least a fair  representation
                                     173

-------
and at best an accurate estimate of the actual loads and losses to and from




Lake St. Clair.









     Finally, does this conclusion (i.e., negligible net internal




sources/sinks during the study period) seem reasonable given the nature of the




internal sources and sinks?  Groundwater input of phosphorus is assumed to be




insignificant compared to that entering the lake from other sources.  The




biological release of phosphorus from sediments and mussels to the overlying




water has also been shown to be negligible (Nalepa et. al.  1987).  Assuming an




average apparent settling rate of 16 m/yr (Chapra 1977) and a range of lake-




averaged phosphorus concentrations measured during eight cruises in 1975 (US




EPA's STORET data base) yields a range of 253-321 MT of phosphorus potentially




lost  to the sediments annually.  These values represent only about 10X of the




total external load for any one year and would be lost in the variability




between input  and output.  Given the shallowness and high wave energy of Lake




St. Clair, sediment resuspension would reduce the impact of particle settling.




Robbins  (1987) indicates that the net deposition of particulate matter to the




lake's  sediments  is small; the sediment  thickness above post-glacial clay




ranges  from 0  to  30 cm, corresponding to a net sedimentation rate of only




0.1-0.2 cm/yr.  At this rate, a range of 0.01-0.04 MT of phosphorus/year would




be lost to the sediments,  representing less than 1% of the total incoming load




for any one year.  Therefore, it does seem reasonable that over a six year




period  there would be no significant net source or  sink of phosphorus in Lake




St.  Clair.
                                    174

-------
                                   SUMMARY









     The objectives of this study were two fold: 1) to estimate and present




the total phosphorus budget for 1975-80, and 2) to determine whether or not




the lake is a net source or sink for phosphorus.  Lake Huron contributed over




half (522) of the lake's load while the seven hydrologic areas contributed 43%




of the remainder.  About 92% of the total Canadian hydrologic area load is




attributable to the Thames and Sydenham areas  of Ontario.  The Clinton and




Black areas of Michigan were responsible for 83%  of the total U.S. hydrologic




area load.  Were reduction of phosphorus loads to Lake St. Clair deemed




desirable, control efforts might be best focused in these four hydrologic




areas.   Because 85% of the total hydrologic area load is from diffuse sources,




a non-point source reduction plan  might be most appropriate.  Reduction of




municipal point sources along the  Thames River and the Clinton River may also




be  important as these sources contributed a majority of the remaining 15% of




the  total hydrologic area  phosphorus  load.









     Over  the  six  year  study period,  1975-80,  the mean external  load and




outflow loss of phosphorus were  not  found to be statistically different at  the




10% level  of significance.  Assuming accurate  estimates of  the loads/losses




and steady state conditions, we  conclude  that  there was no  apparent net source




or  sink of phosphorus  in  Lake  St.  Clair during the  study  period.
                                     175

-------
                              ACKNOWLEDGEMENTS









     This is GLERL contribution No. 511.  This work was partially funded by




interagency agreement DW 13931213-01-0 with the Great Lakes National Program




Office, U.S. Environmental Protection Agency, Chicago.  We thank Barry M.




Lesht, Douglas Salisbury, and two anonymous reviewers for reviewing an earlier




version of the manuscript.
                                     176

-------
                              LITERATURE CITED









Beale, E.M.L.  1962. Some uses of computers in operational research.




     Industrielle Organisation. 31:51-52.









Chapra, S.C. 1977. Total phosphorus model for the Great Lakes. J.




      Environmental Engineering Div. ASCE. 103(2):147-161.









Delumyea,  R.G,  and Petel,  R.L. 1977.  Atmospheric Inputs of Phosphorus  to




      Southern  Lake Huron.  April-October 1975. U.S.  Environmental Protection




      Agency, Report  number 600/3-77-038, Environmental Research Laboratory,





      Duluth, Minnesota, 53 pp.









 Dolan, D.M.,  Yui,  A.K., and Geist, R.D. 1981. Evaluation of river load




      estimation methods for total phosphorus. J. Great Lakes Res. 7:207-214.









 Hall, J.R., Jarecki,  E.A., Monteith, T.J., Skimin, W.E.  and Sonzogni, W.C.,




       1976.  Existing  River Mouth  Loading Data in the U.S. Great Lakes Basin.




       Pollution From Land  Use Activities Reference  Group Report, International




       Joint Commission,  Great  Lakes Regional  Office, Windsor, Ontario, 713 pp.









  International Joint Commission.  1982.  1981 Municipal and Industrial  Phosphorus




       Loadings to  the  Great Lakes. Report of  the Great Lakes  Water Quality




       Board to the International Joint Commission,  Windsor,  Ontario.
                                      177

-------
Klappenbach, E. 1984. 1981 Atmospheric Loading for Lake Huron. Report to the




     U.S. Environmental Protection Agency Great Lakes National Program Office,




     Chicago,  Illinois.









Knap, K.M., and Mildner, W.F.  1978. Streambank Erosion in the Great Lakes




     Basin. Report of the  International Reference Group on Pollution from Land




     Use Activities  to the International Joint Commission, Windsor, Ontario.









Monteith, T.J., and  Sonzogni,  W.C. 1976. United States Great Lakes Shoreline




     Erosion Loadings.  Report of the  International Reference Group on Great




     Lakes  Pollution from  Land Use Activities to the  International Joint




     Commission, Windsor,  Ontario.









Nalepa,  T.F.,  Gardner, W.S.,  and Malczyk, J.M. 1987.  Phosphorus  release from




     sediments and mussels in Lake St.  Clair, with notes on mussel abundance




     and biomass.  Upper Great Lakes  Connecting Channel Final Report.  Great




     Lakes  Environmental Research Laboratory, Ann Arbor, Michigan.









Ontario  Ministry of  the  Environment.  1975.  Water Quality Data: Ontario Lakes




     and Streams. Volume  10.  Report of the  Water Resources Branch, Toronto,




     Ontario.
     .  1976.  Water Quality Data: Ontario Lakes and Streams.  Volume  11.  Report




      of the Water Resources Branch, Toronto,  Ontario.
                                     178

-------
     1977. Water Quality Data: Ontario Lakes and Streams. Volume 12. Report




    of the Water Resources Branch, Toronto, Ontario.









     1978. Water Quality Data: Ontario Lakes and Streams. Volume 14. Report




    of the Water Resources Branch, Toronto, Ontario.









     1979. Water Quality Data: Ontario Lakes and Streams. Volume 15. Report




    of the Water Resources Branch, Toronto, Ontario.









     1980. Water Quality Data: Ontario Lakes and Streams. Volume 16. Report




    of the Water Resources Branch, Toronto, Ontario.









      1985. Upper Great Lakes Connecting  Channels Study:  Canadian Point Source




     Discharge and Combined Sewer Overflow Activities.  Report of the




     Environmental  Protection Service, Ontario  Region,  Environment  Canada,




     Ontario Ministry of the Environment, Windsor,  Ontario.
Remington, R.D. and Schork, M.A. 1985. Statistics with Applications to the




     Biological and Health Sciences. Prentice-Hall, Inc. Publ.,  Englewood




     Cliffs, New Jersey. 415 pp.









Robbins, J.A.  1987. Accumulation of  fallout cesium-137 and chlorinated organic




     contaminants  inn recent sediments of Lake St. Clair. Can. J. Fish, and




     Aouatic Sci.  In  Press.
                                    179

-------
Sonzogni, W.C.,  Monteith, T.J., Bach, W.N.,  and Hughes, V.G. 1978. United




     States Great Lakes Tributary Loadings.  Great Lakes Pollution from Land




     Use Activities Reference Group, International Joint Commission, Technical




     Report, Great Lakes Regional Office, Windsor, Ontario.









U.S. Geological Survey. 1976. Water Resources Data for Michigan. Water Year




     1975. Report of the U.S. Geological Survey, Water Resources Division,




     Lansing, Michigan.
     . 1977. Water Resources Data for Michigan. Water Year 1976. Report of the




     U.S. Geological Survey, Water Resources Division, Lansing, Michigan.
     . 1978. Water Resources Data for Michigan. Water Year 1977. Report of the




     U.S. Geological  Survey, Water Resources Division, Lansing, Michigan.









     . 1979. Water Resources Data for Michigan. Water Year 1978. Report of the




     U.S. Geological  Survey, Water Resources Division, Lansing, Michigan.









     . 1980. Water Resources Data for Michigan. Water Year 1979. Report of the




     U.S. Geological  Survey, Water Resources Division, Lansing, Michigan.









     . 1981. Water Resources Data for Michigan. Water Year 1980. Report of the




     U.S. Geological  Survey, Water Resources Division, Lansing, Michigan.









     . 1982. Water Resources Data for Michigan. Water Year 1981. Report of the




     U.S. Geological  Survey, Water Resources Division, Lansing, Michigan.
                                    180

-------
Water Survey of Canada. 1976. Surface Water Data. Ontario. 1975. Report of the




     Inland Waters Directorate, Water Resources Branch, Water Survey of




     Canada, Ottawa, Quebec.
     . 1977. Surface Water Data. Ontario. 1976. Report of the Inland Waters




     Directorate, Water Resources Branch, Water survey of Canada, Ottawa,




     Quebec.
     .  1978.  Surface  Water Data. Ontario. 1977. Report of  the  Inland Waters




      Directorate,  Water Resources  Branch, Water survey of Canada,  Ottawa,




      Quebec.









     .  1979.  Surface  Water Data. Ontario. 1978. Report of  the  Inland Waters




      Directorate,  Water Resources  Branch, Water survey of Canada,  Ottawa,




      Quebec.









    _.  1980.  Surface  Water Data. Ontario. 1979. Report of  the  Inland Waters




      Directorate,  Water Resources  Branch, Water survey of Canada,  Ottawa,




      Quebec.









     . 1981.  Surface  Water Data.  Ontario.  1980. Report of  the  Inland Waters




      Directorate,  Water Resources  Branch, Water  survey of Canada,  Ottawa,




      Quebec.
 Yaksich, S.M., Melfi, D.A.,  Baker, D.B. and Kramer,  J.W.  1982.  Lake Erie




      Nutrient Loads. 1970-1980. Lake Erie wastewater management study.  U.S




      Army Corps of Engineers District, Buffalo, New York.






                                     181

-------
Table 1.  1975-80 River Mouth Loadings.  The load is presented in metric tons
per year, followed by the root mean square error in metric tons per year,
followed by the 90X confidence interval in Metric tons per year, followed by
the number of phosphorus samples.
Monitored
Tributary
Name 1975
Black NA1




Belle2 31.1
4.9
[22.2,
40.0]
11
Clinton 198.4
22.3
[158.3,
238.4]
12
Ruscom 3.8
0.1
[3.7,
3.9]
7
Thames 418.3
60.4
[309.9,
526.8]
12
Sydenham 196.0
28.6
[142.9,
249.1]
9
Detroit 2769.3
289.7
[2244.2,
3294.4]
11
1976
NA1




28.7
8.7
[13.1,
44.4]
12
143.8
24.0
[100.8,
186.9]
12
6.0
0.5
[5.1,
6.9]
11
690.8
160.6
[396.5,
985.1]
10
195.5
46.9
[110.5,
280.5]
12
3935.9
301.5
[3383.3,
4488.6]
10
Year
1977 1978
NA1




7.1
0.9
[5.5,
8.6]
12
118.3
16.9
[88.0,
148.7]
12
16.3
1.3
[14.0,
18.6]
11
1391.8
427.4
[625.1,
2158.4]
12
494.3
137.1
[248.4,
740.2]
12
3304.7
393.4
[2603.6,
4005.9]
13
47.4
2.9
[41.9,
52.9]
8
10.4
0.9
[8.8,
12.0]
12
112.7
23.3
[70.9,
154.5]
12
22.3
6.0
[11.5,
33.1]
12
643.3
58.1
[537.9,
748.7]
11
94.4
18.6
[61.1,
127.8]
12
3090.2
250.3
[2644.1,
3536.2]
13
1979
90.8
38.9
[21.0,
160.6]
12
11.9
2.1
[8.0,
15.7]
12
77.8
8.9
[62.0,
93.7]
12
9.0
1.8
[5.7,
12.3]
12
917.7
193.3
[587.0,
1248.4]
25
241.1
27.1
[195.0
287.2]
27
2879.7
247.9
[2448.5,
3310.9]
18
1980
88.4
29.0
[34.5,
142.4]
9
17.4
4.7
[8.7,
26.2]
9
114.0
28.9
[62.2,
165.8]
12
36.3
20.7
[0.0,
75.4]
8
663.8
102.1
[493.8,
833.8]
75
170.3
19.2
[137.7,
202.9]
31
2908.7
273.3
[2427.3,
3390.1]
15
  ^Phosphorus data not available for years 1975-77,
  2within St. Clair Hydrologic Area.
                                     182

-------
Table 2.  Annual phosphorus budget (MT/yr) for Lake St. Glair (1975-80).
Values in parentheses represent percent of total load from diffuse sources,
Percent monitored refers to percent of hydrologic area that is monitored.
Source
Hydrological Areas
Black
0,100% monitoredb
St. Clairc
36% monitored
Clinton
100% monitored
Rouged
0% monitored
Ruscom6
18% monitored
Thames
100% monitored
Sydenham
100% monitored
Total HA Load
Atmospheric
Erosion
Direct Point
Lake Huron
1975
IHAI
115. 5a
(74)
75.4
(90)
198.4
(48)
10.8
(100)
21.8
(100)
418.3
(85)
196.0
(97)
1036.2
(80)f
14.0
82.5
55.8
2022
EXTERNAL LOAD 3211
90% C.I. [3014,
3408]
1976
84. 5a
(64)
68.7
(89)
143.8
(26)
4.2
(100)
34.1
(100)
690.8
(90)
195.5
(97)
1221.7
14.0
82.5
64.5
1373
2756
[2485,
3027]
1977
41. 9a
(28)
7.8
(3)
118.3
(21)
2.9
(100)
92.2
(100)
1391.8
(94)
494.3
(99)
2149.2
(90)f
14.0
82.5
58.9
1187
3492
[3070,
3914]
1978
47.4
(36)
17.0
(56)
112.7
(37)
4.7
(100)
126.4
(100)
643.3
(91)
94.4
(92)
1045.9
(83)f
14.0
82.5
57.7
1613
2813
[2648,
2978]
1979
90.8
(67)
21.2
(64)
77.8
(18)
1.6
(100)
51.1
(100)
917.7
(92)
241.1
(99)
1401.3
(87)f
14.0
82.5
59.2
1703
3260
[2987,
3533]
1980
88.4
(63)
38.2
(80)
114.0
(39)
5.0
(100)
205.7
(100)
663.8
(90)
170.3
(99)
1285.4
(86)f
14.0
82.5
54.0
1827
3263
[3136,
3390]
 OUTFLOW LOSS
     90% C.I.
  IN  -  OUT

  aUnit Area  Load
   UAL  of Clinton
   2769      3936      3305      3090      2880       2909
[2244,     [3383,     [2604,     [2644,     [2449,     [2427,
    3294]      4489]      4006]      3536]      3311]      3390]
                      441
            -1180
186
                                                   -277
380
                                                      354
(UAL) assumed to equal average of UAL of St.  Clair HA and
HA.
                                     183

-------
bOX for years 1975-77, 100Z for years 1978-80.
CUAL assumed to equal UAL of Belle R. within St. Glair HA.
dUAL assumed to equal UAL of Clinton HA.
eUAL assumed to equal UAL of Ruscom R. within Ruscom HA.
^Percentages weighted with respect to percent of total annual input that is
 attributable to a given area.
                                     184

-------
                               LIST OF FIGURES









Figure 1.  Lake St. Clair Basin showing hydrologic areas used in calculating




           load estimates.  Note:  the Thames hydrologic area extends to




           approximately 43" 30'  latitude, 80°  30' longitude.









Figure 2.  Schematic  displaying procedure used to estimate the combined




           probability  distribution for each year's total external  load.









Figure 3.  Summary of Lake  St. Clair average phosphorus dynamics during  the




           1975-80 period.  Values are  inMT/year.  The relative proportion of




           inflows and  outflows  are approximated by the thickness of the arrow




           shafts.









Figure A.  External phosphorus loads to Lake St. Clair  (closed circles and




           vertical bars  represent 90X confidence  interval around annual load,




           MT/yr) and Detroit  River outflow loss  (shaded  region  represents 90%




            confidence interval around annual  loss, MT/yr).
                                     185

-------
            83°00'
                         82°30'
 43°30'
 43°00'
42°30'
                                      82°00'
42°00'
    Complex
                                        136

-------
Clinton
Ruscom
Thames
              Load

-------
Lake St. Clair average phosphorus loads and losses
during the 1975-'80 period (metric tonnes per year)
U.S.
Hydrologic
Areas
            Atmospheric,
            Erosion,     Lake Huron
            Direct Point     1,621
              155™
 Black
        78
St. Clair
Complex 381
               LAKE ST. CLAIR


                 Net Loss =
   Canadian
   Hydrologic
   Areas

i232 Sydenham

i 788 Thames
             Detroit River Outflow 3,148
                           188

-------
    5000
  ca
tn *•
2«
O o
    4000-
« o
O -5=
  p 3000
  "5
                                  _j	i
           Outflow Loss i

                                         Load
           1975    76    77     78    79
                       Time (years)
                                           '80
                    189

-------
     PHOSPHORUS RELEASE FROM SEDIMENTS AND MUSSELS IN LAKE ST.  GLAIR,




                WITH NOTES ON MUSSEL ABUNDANCE AND BIOMASS










             T. F. Nalepa,  W. S. Gardner, and J.  M.  Malczyk
                               INTRODUCTION









     Since phosphorus is known to be the critical element in controlling




eutrophication, a thorough understanding of phosphorus dynamics is essential




for effective control strategies.  The sediments play an important role in




phosphorus cycling, serving as either a sink or a source of phosphorus to




the overlying waters.  The processes affecting the net flux of phosphorus




from the sediments are complex, depending on such factors as resuspension,




sedimentation, sorption, oxygen concentrations, and invertebrate activities.




Under oxic conditions, as found in most near-bottom waters of the Great




lakes, benthic invertebrates play a major role in phosphorus release from




the sediments  (Gallepp 1979; Graneli 1979; Holden and Armstrong 1980;




Quigley and Robbins 1986).  By their constant burrowing and feeding




activities, benthic invertebrates increase the rate of exchange between




nutrient-rich pore waters and overlying waters.  Also, these organisms




ingest organic material and subsequently excrete remineralized nutrients in




forms readily available for further use by phytoplankton.  In nearshore Lake




Michigan, excretion by benthic invertebrates was sufficient to account for




all the phosphorus released from the sediments (Gardner et al.  1981).  Of
                                    190

-------
the various invertebrate groups, unionid bivalves (mussels) in particular




can have a significant impact on nutrient cycling in a given body of water




(Lewandowski and Stanczykowska 1975; Walz 1978; Stanczykowska and Planter




1985; Kasprzak 1986; James 1987).  These large filter-feeders have the




capacity to remove great amounts of organic material from the water.  Thus,




they enhance nutrient mineralization either directly through excretion, or




indirectly by depositing the material on the sediment surface as faeces or




pseudofaeces and making it available to deposit-feeding forms.









     The purpose of this study was  (1) to quantify the rate of phosphorus




flux between the sediments and overlying waters in Lake St. Clair and  (2) to




determine  the rate of phosphorus excretion by  the mussel population.  The




significance of both sediment release and mussel excretion was subsequently




assessed by comparing these phosphorus sources to other sources of




phosphorus into  in  the  lake.  In addition to measuring excretion, the




abundance, biomass, species composition, and production of mussel




populations in  the  lake were also  determined.  Accurate estimates of mussel




biomass were, of course,  essential to assessing  the  importance of phosphorus




excretion  on a  lake-wide  basis.
                           METHODS AND MATERIALS









      Intact sediment cores were collected by divers at  five  sites  in Lake




 St.  Clair in May and September, 1985.   The sites were chosen to be broadly




 representative of different areas and sediment types (Fig. 1).  The  core
                                     191

-------
tubes (4.2 cm diameter and 10 cm long) were inserted into the sediment about




5 cm, stoppered at both ends, and carefully brought to the surface.  The




cores were kept upright in a cooler during transport back to the laboratory




and then placed in an incubator set at the in situ temperature.   Aeration




lines were placed through the top stopper and air was slowly bubbled into




the overlying waters.  This kept the water well-mixed and also kept




dissolved oxygen concentrations at near-saturation levels.   All core tubes




and aeration lines were made of high-density linear polyethylene to minimize




phosphorus adsorption.  Samples for phosphorus determinations were taken




every 3-4 days by drawing out 1 ml of water through a sampling port in the




top stopper.  Phosphorus concentrations (SRP) were determined with an




AutoAnalyzer as described by Gardner and Malczyk (1983).   Phosphorus levels




in lake-water controls were also measured on each sampling day.   The volume




of overlying water was kept constant by adding 1 ml of lake water after each




sample was drawn.  The incubation period lasted between 65 and 70 days.









     A total of 6-8  replicates cores were collected at each station on each




sampling  date.  Since Lake St. Clair  is shallow and bottom sediments are




easily resuspended,  the  impact of resuspension on sediment phosphorus




release was estimated by mixing one-half of the replicates at the beginning




of the incubation period to create a  sediment slurry with the overlying




waters.   The sediments were mixed again every 10 days until the end of the




incubation period.   Phosphorus release rates in these mixed cores were




compared  to release  rates in cores that were left undisturbed.
                                    192

-------
     Mussels for phosphorus excretion determinations were collected on a




monthly basis from May to October in both 1985 and 1986.  An epibenthic sled




was towed behind the vessel until enough individuals were collected.  In




1985, mussels were collected from only one site (Station 72) and excretion




measured on several different species.  In 1986, there were two collection




sites (Station 72 and  Station 24) and rate measurements were made on only




one species, Lampsilis radiata siliquodea.   The two sites had contrasting




substrate types, with sandy silt the dominant substrate at Station 72 and




silt dominant at Station 24.  Excretion rates were determined on at least




four individuals from each of the two stations except in May when rates were




determined on only two individuals from each station.   Individual mussels




were gently scrubbed and immediately placed in polyethlene containers having




2 liters of low-nutrient culture water (Lehman 1980).  The containers were




placed in large coolers and the culture water was maintained at the in situ




temperature.  The incubation period lasted 4 hours with 1-ml samples drawn




at 0, 2, and 4 h.  Phosphorus concentrations were determined as in the




sediment cores.  Dry weights of the mussels (soft tissue) were determined




after drying at 60 C for at least 48 h.









     To determine the density, biomass, and species composition of mussels




in the lake, a population survey was conducted in September, 1986.  Divers




placed a 0.5 m^ frame on the bottom and all shells within the frame area




were placed in a mesh bag.  A total of 10 separate replicate samples were




collected at random at each of 28 different stations (Fig.  1).   All live




mussels were immediately shucked and the soft tissue placed in preweighed




aluminum planchets.   Dry weights of both the shell and soft tissue were
                                    193

-------
obtained after drying at 60 C for at least 48 h.  Individuals of the two




most abundant species, Lamosilis radiata siliquodea and Leptodea fragilaris.




were aged by counting the number of annual growth rings on the external




shell.  The annual production rate of L. r. siliquodea was estimated from




the sum of the increase in weight of each of the different age groups




(Magnin and Stanczykowski 1971).
                           RESULTS  AND DISCUSSION









 Phosphorus  Release  From the Sediments








      Sediment phosphorus release rates at each of the five stations on the




 two sampling dates  is given in Table 1.  Rates were calculated from the net




 increase in phosphorus concentrations in the overlying waters from day 10 to




 day 65-70 for the May cores, and from day 3 to day 65-70 for the September




 cores.  In some instances, the increase in phosphorus was most rapid at the




 beginning of the incubation period and then remained relatively constant




 thereafter.  For these cores, rates were recalculated based on the time




 interval of greatest release and are included in Table 1 to provide an




 estimate of maximum potential release.  The mean release rates in this table




 include values  from all replicates at  a given station since there were no




 significant  differences  (t-test;  P < .05) between release rates of mixed and




 unmixed cores  at any  of  the five  stations.
                                     194

-------
     In both May and September, release rates at Stations 71 and 84 were




generally lower than release rates at the other three stations (Stations 4,




14, and 24).  The  former two stations were located in the northwestern




portion of the lake near the mouth of the St. Clair River where both




nutrient levels and algal productivity tend to be lower then areas more to




the southeast (Leach 1972).   This portion of the lake is dominated by low




nutrient water from Lake Huron, while areas farther south are more




influenced by enriched waters from Ontario tributaries (Leach 1980).  Also,




given the dominant current patterns and wind direction, very little




deposition of suspended material occurs in the northwestern portion of the




lake (Anne Clites, GLERL, per. commun.).  Release rates at Stations 71 and




84 were similar in both May and September, but release rates at the other




stations were higher in May than in September.  The settling and subsequent




mineralization of the spring phytoplankton bloom likely contributed to the




higher release rates at these stations in May.  Overall, the highest release




rates occurred at Station 24; this station was located in the area of




greatest deposition.









     The release of SRP from Lake St. Clair sediments was generally lower




than sediment release rates in other areas of the Great Lakes.  The mean




release rate in this study was 19 ugP/m^/day with a mean maximum release




rate of 47 ug P/m^/day.  This compares to release rates of 170-570




ugP/m^/day  in nearshore Lake Michigan (Quigley and Robbins 1986) and 30-800




ugP/m^/day  in Lake Ontario  (Bannerman et al. 1974).
                                    195

-------
     To determine the significance of sediment phosphorus release in Lake




St. Clair, the annual net release from the sediments was compared to other




input sources, i.e. Lake Huron, tributaries, the atmosphere, and direct




point sources.  The total mean load from these latter sources during the




1975-80 period was about 3,100 MT/year (Tom Fontaine, GLERL, per. commun.).




Of this amount, about 40% or 1200 MT can be considered bioavailable; that




is, available for algal uptake and not bound to particulates (Sonzogni et




al. 1982).  Assuming that the mean release rate of phosphorus at the five




stations  is representative  of the entire lake, release from the sediments




amounts to about  8 MT/year  or less than 12 of the total bioavailable




phosphorus load.   Maximum sediment release amounts to only about 20 MT/year




or 22 of  the  total bioavailable load.  Based on these calculations, the




sediments appear  insignificant as a source of phosphorus  in Lake St. Clair.









Phosphorus Excretion by Mussels









      Mean rates of phosphorus excretion  on  the  eleven sampling  dates  in  1985




and  1986  are  given in Table 2.    Rates from the two  stations sampled  in  1986




were combined since  significant  station  differences  were  not apparent  (t-




 test;  P < .05)  for any of the sampling dates.   Seasonal  trends  in  phosphorus




 excretion were similar for the two  years.   Rates  were high  in  the  spring,




 declined in the summer, and then increased in the fall  to reach peak  values.




 Reasons for this seasonal trend are not clear,  but may  be related  to  changes




 in the gametogenic cycle of the  organisms.   An increase in  ammonia excretion




 in the summer (Table 2) indicates an increase in  gamete production at this




 time.  Active protein catabolism (and hence increased ammonia  excretion)
                                     196

-------
occurs when the glycogen normally used for metabolism is used instead for




gamete production (Gabbott and Bayne 1973).   While ammonia excretion would




increase, phosphorus excretion would likely decrease, since a greater




portion of assimilated material would be used for reproductive activities




and not metabolism.  Seasonal changes in phosphorus excretion were




apparently unrelated to the nature of available food; the amount of




particulate phosphorus in the near-bottom water was constant throughout the




sampling period (Nalepa, unpublished).  Mussel excretion rates were




generally lower than those of other Great Lakes benthic organisms (Table 3).




This may be expected, however, since the rate of phosphorus excretion per




unit weight increases as body weight decreases (Johannes 1964).  The dry




weight of mussels in this study ranged from 1 to 4 g, while dry weights of




the other organisms shown in Table 3 have dry weights of less than 2 mg.









     Based on  estimates of biomass from the September 1986 population survey




(see below), the mussels in Lake St. Clair excrete about 59 MT of phosphorus




per year or 5% of the annual bioavailable load from other sources.  In




addition to excreting phosphorus, mussels are active filter feeders and may




remove large amounts of particulate phosphorus from the lake water during




feeding.  Based on preliminary estimates of mussel filtration rates




(Vanderploeg and Nalepa, unpublished) and amounts of particulate phosphorus




in Lake  St Clair water, mussels are capable of filtering 220 MT of




phosphorus from the water on an annual basis; this amounts to 7% of the




total annual load.
                                    197

-------
Mussel Population Survey









     The overall mean abundance of mussels was 2/m2 (range 0-8/m2) and the




mean biomass was 4.3 g/m2 (range 0-19.4 g/m2).   In general, both abundance




and biomass increased from the mouth of the St. Clair River to the head of




the Detroit River.  This corresponded to previously noted trends in water




column productivity.  A total of 287 individuals representing 20 different




species were collected.  The three most abundant species, Lampsilis radiata




siliquodea. Leptodea fragilaris. and Proptera alata accounted for 45Z, 13%,




and 10Z of the total population.  The former species was the most widely




distributed, being collected at 22 of the 28 stations.  Abundances found in




this survey were lower than abundances reported from other areas in the




Great Lakes.  For instance, densities of 7/m2 (Wood 1963) and 10/m2 (McCall




1979) have been reported from western Lake Erie, while Pugsley (unpublished




data) reported a mean density of 7/m2 in the southwestern portion of Lake




St. Clair.  A direct comparison can be made between this survey and the




Pugsley survey since three of the sampling stations were the same and




sampling techniques were similar.  Abundances in this survey were




significantly lower at two of the three stations (Table 4).  It is not clear




whether these lower abundances are are a result of an actual decline in the




population or an artifact of horizontal patchiness.  Considering the




stability for mussel populations over the short-term,  such a decline in




abundances over just a 3-year period seems unlikely unless, of course,




environmental conditions have recently become unfavorable.  Although an




unusually high number of dead mussels have been observed on Lake St.  Clair




beaches over the past few years (Tom Freitag, US Army Corps of Engineers,
                                    198

-------
per. commun.) only through long-term  monitoring efforts can definite trends




in abundances be discerned.









     In western Lake Erie, mussel density and diversity have apparently




declined over the past few decades  (Mackie et al. 1980).  Unfortunately,




historical records of mussel densities in Lake St. Glair are lacking.




However, mussel diversity and composition appear little changed since 1893.




Reighard (1894) reported finding 20 species  in Lake St. Clair with  Lampsilis




radiata siliquodea being very " widespread and abundant" and Proptera alata.




Liguinea nasuta. Anodonta grandis being found "frequently".  The most




apparent difference between this survey and  the  1893  survey of Reighard was




the  relative  abundance of Leptodea  fragilaris: this species was reported




being  "scarce" by Reighard but  in this survey, it was the second  most




abundant species.









     The age  structure of  L.  r.  siliquodea and L. fragilaris  is given  in




Figure 3.   For  L.  fragilaris.  the age structure  of  the population was  quite




similar to that found in other  freshwater systems  (Strayer et al. 1981;




Paterson 1985)  and reflects  low adult mortality  and yearly variation  in




recruitment.  However, for L. jr.  siliquodea. the average individual was




almost 10  years  of age and few  younger  individuals  were  found.  The reason




for this lack of recruitment  is not clear, but may  indicate that  either  the




adult  population is under  some  sort of  stress  (low  reproductive capacity)  or




that mortality  of the young  is  increasing.     Populations of  fish species




which  serve as  host for  the  glochidia of L.  r.  siliquodea  (yellow perch,




 smallmouth bass,  largemouth  bass,  bluegill,  and  crappie  among others)  have
                                     199

-------
remained stable over the years (Bob Haas, Michigan DNR, per.commun.).




Because the population is dominated by older individuals,  the annual




turnover rate (production/biomass) of L. r. siliquodea was only 0.13;  this




value is lower than found for mussels in most other freshwater lakes (Table




5).
                                    200

-------
                              LITERATURE CITED







Bannerman, R. T., D. E. Armstrong, G. C. Holdren, and R. F. Harris.  1974.




     Phosphorus mobility in Lake Ontario sediments (IFYGL), pp. 158-178.




     Proc. 17th Conf. Great Lakes Res., Int. Assoc. Great Lake Res.









Gabbot, P. A. and B. L. Bayne.  1973.   Biochemical effects of  temperature




     and  nutritive  stress  on Mytilus edulis L.   J. Mar.Biol. Ass.  U. K.




     53:269-286.









Gallepp,  G.  W.   1979.   Chironomid influence and phosphorus release in




     sediment-water microcosms.   Ecology  60:  547-556.









 Gardner,  W.  S.,  T.  F.  Nalepa,  M.  A.  Quigley,  and J.  M.  Malczyk.   1981.




     Release of phosphorus by certain benthic invertebrates.   Can. J.  Fish.




      Aquat.  Sci.  38:978-981.









 Gardner,  W.  S. and J. M. Malczyk.   1983.   Discrete injection flow analysis




      of nutrients  in small-volume water samples.  Anal. Chem.  55:1645-1647.









 Graneli, W.  1979.  The influence of Chironomus plumosus on the exchange of




      dissolved  substances between sediment and  water.  Hydrobiologia




      66:149-159.









 Holden,  G.  C. and  D.  E. Armstrong.   1980.  Factors  affecting  phosphorus




      release  from  intact  sediment cores.  Environ.  Sci. Technol.  14:79-87.
                                     201

-------
James, M. R.  1987.  Ecology of the freshwater mussel Hyridella menziesi




      (Gray) in a small oligotrophic lake.  Arch. Hydrobiol. 3: 337-348.









Johannes, R. E.  1964.   Phosphorus excretion and body size in marine




      animals: microzooplankton and nutrient regeneration.  Science




      146:923-924.









Kasprzak,  K.   1986.   Role of Unionidae and Sphaeriidae  (Mollusca,  Bivalvia)




      in the eutrphic lake Zbechy and its outflow.   lint.  Revue ges.




      Hvdrobiol.  71: 315-334.









 Leach  J.  H.   1972.  Distribution of chlorophyll a and related variables in




      Ontario waters of  Lake St. Clair, pp.80-86. In Proc. 15th Conf. Great




      Lakes Res.,  Int. Assoc.  Great Lakes Res.









 Leach, J. H.  1980.  Limnological sampling intensity in Lake St. Clair in




      relation to  distribution of water masses.  J. Great Lakes Res.  6:





      141-145.









  Lehman, J. H.   1980.   Release and cycling of  nutrients between planktonic




       algae and  herbivores.  Limnol.  Oceanogr.  25:  620-632.









  Lewandowski,  K. and A.  Stanczykowska.  1975.   The occurrence and role of




       bivalves of the family Unionidae inMiklajskie Lake.   Ekol.  Pol. 23:





       317-334.
                                      202

-------
Mackie, G. L.,  D. S. White, and T. W. Zdeba.  1980.  A guide to freshwater




     mollusks of the Laurentian Great Lakes with special emphasis on the




     genus Pisidium.   EPA-600/3-80-068, Environmental Protection Agency,




     Duluth, Mn.  144p.









Magnin, E. and A. Stanczykowska.  1971.  Quelques donnees sur la croissance,




     la biomass.et la production annuelle de trois mollusquesUnionidae de la




     region de Montreal.  Can. J. Zool. 49: 491-497.









McCall, P. L.,  M. J. S. Tevesz, and S. F. Schwelgien.  1979.  Sediment




     mixing by Lampsilis radiata siliquodea (Mollusca) from western Lake




     Erie.  J.  Great Lakes Res.  5:105-111.









Nalepa, T. F.,  W. S. Gardner, and J. M. Malczyk.  1983.  Phosphorus release




     by three kinds of benthic invertebrates: effects of substrate and water




     medium.  Can. J. Fish. Aquat. Sci.  40:810-813.









Paterson, C. G.  1985.  Biomass and production of the unionid, Elliptic




     complananta (Lightfoot)  in an old reservoir in New Brunswick, Canada.




     Freshwat.  Invertbr. Biol. 4: 201-207.









Quigley, M. A.  and J. A. Robbins.  1986.  Phosphorus release processes in




     nearshore southern Lake  Michigan.  Can. J. Fish. Aquat. Sci. 43:




     1201-1207.
                                    203

-------
Reighard, J. E.  1894.  A biological examination of Lake St. Clair.  Bull.




     Mich. Fish Comm. No.4.  61p.









Sonzogni, W. C., S. C. Chapra, D. E. Armstrong, and T. J. Logan.   1982.




     Bioavailability  of phosphorus  inputs to lakes.  J. Environ. Qual.  11:




     555-563.









Stanczykowska, A.  and M. Planter.   1985.  Factors  affecting nutrient budget




     in lakes  of  the  R.  Jorka  watershed (Masurian  Lakeland, Poland)  X.  Role




     of the mussel Dreissena polmorpha (Pall.)  in  N  and  P cycles  in  a  lake




     ecosystem.   Ekol.  Pol.   33:345-356.









 Strayer, D. L.,  J. J. Cole,  G. E.  Likens,  and  D. C.  Busco.  1981.  Biomass




      ans annual production of the freshwater mussel  Elliptic  complanata in




      an oligotrophic lake.  Freshwat.  Biol.  11: 435-440.









 Walz,  V. N.  1978.  Die produktion der Dreissena-population und deren




      bedeutung im stroffkreislauf des Bodensees.   Arch.  Hydrobiol.




      82:482-499.









 Wood,  K. G.   1963.  The bottom fauna of western Lake Erie,  1951-52.   Great




      Lakes Res. Div., Univer. Michigan, Publ.  No.  10, pp.  258-265.
                                     204

-------
Table 1. Mean (+ SE) rates of phosphorus release from
Lake St. Clair sediments at each of the stations on
the two sampling dates in 1985.  Maximum mean release
rates are given in parentheses.  Rates are given as
ugP/m2/day.
 Station
                         Sampling Date
May1
  ^-Water  temperature =  13 C.
  2water  temperature =  22 C.
September^
4
14
24
71
84
31
15
31
4
8
.4 ±
• 5 ±
-2 ±
.2 ±
• 3 +
4.9
6.0
9.4
0.8
3.0
(40
(49
(38
(18
(15
.0)
.0)
.2)
.6)
.0)
11
11
22
5
8
• 2 ±
• 2 ±
.9 +
.2 ±
• 0 ±
6
1
5
1
3
.3
.6
.5
.2
.9
(32
(78
(58
(16
(32
.3)
.4)
.5)
.3)
-3)
                       205

-------
Table 2. Mean  (± SE) phosphorus and ammonium excretion
rates of mussels in Lake St. Clair in 1985 and 1986.
Rates given in ug/gDW/h.
Sampling
Date
1985
May 9
May 14
Jul 16
Sep 3
Sep 19
1986
Apr 30
May 19
Jul 10
Aug 4
Sep 16
Oct 15
n

5
7
6
7
7

9
4
9
10
10
10
Excretion
Phosphorus

0.9 ± 0.1
0.9 + 0.3
0.7 + 0.5
3.9 ± 0.9
2.0 ± 0.4

1.1 + 0.3
0.6 ± 0.1
0.5 ± 0.1
1.5 + 0.3
1.9 ± 0.4
1.9 ± 0.4
Rate
Ammonium

24.6 ± 3.7
12.8 ± 2.1
49.1 + 3.2
25.3 ± 2.4
27.6 ± 4.5

23.3 ± 2.3
20.9 + 4.0
50.1 ± 3.7
52.3 ± 8.5
40.8 ± 6.2
21.3 ± 2.2
N:P
Ratio

27
14
69
7
14

20
35
109
35
21
12
                      206

-------
Table 3.   Mean (± SE) phosphorus and ammonium excretion rates
(nmol/gDW/h) for some common benthic invertebrates occurring in
the Great Lakes.  Data compiled from Nalepa et al. (1983), Gardner
et al. (1983), and Gauvin (unpublished).
Benthic             	Excretion Rate                 N:P
Organism            Phosphorus     Ammonium            Ratio

Chironomidae            690         11,300               16
Oligochaeta             150          8,100               54
Pontoporeia              90          1,090               12
Unionidae                50          1,020               23
                             207

-------
Table 4.  Comparison of mean mussel abundances (number per
square meter) at three stations in Lake St. Clair in 1983
(Pugsley, unpublished) and in 1986 (this study).   Standard error
in parenthesis. * - Densities significantly different at the
0.05 level (t-test).
                         1983             1986
          Station    (Pugsley 1986)	(This study)

             3         13.8  (1.7)         7.8 (1.5)*
            21          9.8  (2.1)         2.2 (0.9)*
            66          2.4  (0.6)         2.0 (0.7)
                             208

-------
Table 5.   Turnover ratio (production/biomass) of unionids from various
lentic environments.
Water Body
P/B
 Reference
Lake Zbechy
Lake Mikolajskie

Lac des Deux
  Montagnes
Morice Lake
Lake St. Clair
Mirror Lake
Lac Saint Louis
0.45
0.35

0.20

0.19
0.13
0.12
0.10
Kasprzak (1986)
Lewandowski and
  Stanczykowska (1975)
Hagnin and
  Stanczykowska (1971)
Paterson (1985)
This Study
Strayer et al. (1981)
Magnin and
  Stanczykowska (1971)
                             209

-------
                                LIST OF  FIGURES
Fig. 1.  Sampling stations in Lake St. Clair.   Sediment phosphorus
         release rates were determined at Stations 4,  14,  71,  84, and
         24.  Mussel populations were sampled at all the stations
         except the first four stations given above.
Fig. 2.  Age structure of the two most abundant species,  Lampsilis radiata
         siliquodea and Leptodea fragilaris.
                                    210

-------
                                              10
                                 m
Figure 1,
   211

-------
  25r-
£ 20

o
O
O
   10
       L r. siliquodea
     0   2  46   8  10   12  14  16

               Age (years)
   25
   20
O
JD

O  15
O


I  10
       L fragilaris
    o
•-•••••••—
4   6  8  10  12  14  16
  Age (years)
                Figure 2.

                   212

-------
                     SEDIMENT TRANSPORT IN LAKE ST. CLAIR




                        Nathan Hawley and Barry Lesht^







                                   ABSTRACT




     In order to study resuspension in Lake St. Glair, bottom-resting




instrumented tripods were deployed at various sites in the lake during 1985




and 1986.  The tripods recorded time series measurements of current




velocity, temperature, and water transparency.  The measurements were then




used to calculate the parameters of a simple flux model in order to




determine the criterion for sediment resuspension.  Since most resuspension




in the lake is due to wave action, wave orbital velocity was used as the




forcing function.  These velocities were calculated by running the GLERL




wave model using on-lake wind measurements to determine the wave climate,




and then calculating the orbital velocities using intermediate-water wave




theory.  The pattern of the calculated velocities shows good agreement with




both the measured standard deviation of the current velocity (although the




magnitudes are somewhat different) and the total suspended material.




Critical values of wave orbital velocity (the threshold for resuspension)




range from 0.1 to 0.9 cm/s.  The model shows good predictive capability and




is relatively insensitive to changes in the settling velocity and




resuspension coefficient.  Some of the variability in the critical velocity




may be due to changes in substrate characteristics.
^Argonne National Laboratory
                                      213

-------
                               INTRODUCTION









     As part of the Upper Great Lakes Connecting Channels study, we




undertook to develop an empirical relation between flow activity and




sediment resuspension in Lake St. Clair.  To do this we deployed bottom-




resting instrument packages in the lake at various times and sites which




recorded time series measurements of water transparency, flow velocity, and




water temperature.  The measurements were then modeled using a relation




suggested by Simons and Schertzer (1986) to develop a criterion for sediment




resuspension.  Measurements of total suspended material (TSM) and vertical




profiles of water transparency and water temperature were also made.  A




meteorological tower measured on-lake weather conditions from July to




September, 1986.  Plans to use a bottom-resting flume to measure erosion




thresholds had to be abandoned because a suitable vessel was not available




for  its deployment.  Several other programs, in particular the wave study by




the  Great Lakes Environmental Research Laboratory (GLERL) and the Canadian




Centre for Inland Waters  (CCIW)  and the sediment transport study by CCIW




have provided valuable supporting data.









     Lake St. Clair is a  large  (approximately 40 km wide), shallow (maximum




depth 7m) lake located between Lakes Huron and Erie.  As such it receives




the  entire outflow from the upper Great Lakes - approximately 5300 m^/s.




Since the residence time  of the  water in the lake is only about 7 days, much




of the sediment carried into the lake is almost immediately swept out again.




However much of the lake  is covered by sand and silt which can be




resuspended.  Because of  its shallow depth and large fetch, we felt that
                                      214

-------
wave action might be a prime cause of resuspension in Lake St. Clair.  Thus




most of our current velocity measurements were in burst mode.  This allowed




us to measure not only the mean current, but also its standard deviation,




which we felt could serve as a measure of wave action.
                              DATA COLLECTION









     One  or more  tripods were  deployed  five  times  during 1985 and  1986  (Fig.




1).  One  of these  tripods,  the one deployed  by Barry Lesht of Argonne




National  Laboratory  (ANL),  was equipped with a Marsh-McBirney current meter




and  25  cm pathlength Seatech transparency meter, as well as with a




temperature probe.   The other  two tripods were deployed by Nathan  Hawley




(GLERL) and were  equipped with only  a temperature  probe and 25 cm  Seatech




transparency meter.   Although  we had planned to have current meters on  these




two  tripods as well,  the meters we bought never worked properly.   Details of




the  tripod locations and deployment  periods  are given in Table 1.  In all




cases  the GLERL transparency meters  were 0.9m above the bottom.  Both they




and  the temperature  sensor, which was located 1.2m above the bottom, took a




60 second sample  at  1 Hz every 15 minutes.   The mean and standard  deviation




of the transparency  (TSD) were recorded along with the average temperature.




The  transparency  meter on the  ANL tripod was 0.9m  above the bottom, the




current meter  0.7m and the  temperature  sensor 1.0m.  During all but one of




 the  deployments  the current meter recorded  a 75 second burst at 3.4 Hz.




 During July-August of 1985  and May-June,  1986 these measurements were taken




 every 45  minutes, during July-August,  1986  every 60 minutes, and during
                                      215

-------
October, 1986 every 30 minutes.  For all but the last of these deployments




only an average transparency and temperature were recorded.  During the




October, 1986 deployment transparency was recorded in the same manner as the




current velocity.  During the September-October, 1985 deployment continuous




5 minute averages of current velocity, water transparency, and temperature




were made.









     In order to calibrate  the transparency meters, measurements of




transparency and TSM were made in triplicate at 25 different stations (Fig.




2).  For 110 measurements,  TSM (measured in mg/1) is related to water




transparency (measured as the fraction of the transmittance in air) by
                        TSM -  -8.33* Ln(Tr)-1.96          r2=0.92      (1)









The measurements  used to  establish  this  relationship show no geographic or




temporal  trends.   Use of  (1) allowed TSM to be calculated from the




transparency measurements.  Figure  3 shows the predicted values and the




measurements used in the  regression.   The two curved lines represent the 95%




confidence  interval  for the predicted  values. It  should be noted however,




that  all  of the calibration measurements were made  during fairly calm




conditions.  During  resuspension events  far more  sand-sized material is




likely to be  in  the  water column.   Because of its high density, a given




weight of sand will  attenuate  the light  beam far  less than the same weight




of more porous,  far  less  dense,  floes  which form  from cohesive material.




Thus, the TSM calculations must  be  underestimates of the true loading.  In
                                      216

-------
the model we used, however, concentration is essentially a surrogate for

transparency.  Since the model parameters are all internally estimated, they

will still be consistent.  However, the actual rates of settling and

resuspension measured on a mass basis will probably be somewhat different

from those calculated.
                             MODEL DESCRIPTION



     Simons and Schertzer (1986) have proposed a simple model which relates

changes in the vertically integrated (assuming vertical uniformity)

suspended sediment concentration (C) to changes in the balance between the

rates of sediment resuspension and  sediment settling.  This is just a flux

model in which the net flux  (the left-hand side of the equation) is equal to

the difference between the upward and downward fluxes. The resuspension rate

is assumed to be a linear function of the forcing function (F) above some

critical value (Fc) and zero otherwise, with R being the resuspension

concentration.  The settling rate is the product of the settling velocity

(S) and the ambient concentration (Ca) which is assumed to always be

present.  Thus the model is
                D*dC =          - S(C-Ca)   for  F < Fc                 (2)
                  dt
and
                D*dC = R(F-FC)  - S(C-Ca)    for F > Fc                 (3)
                   dt
                                      217

-------
Using the results from our deployments we could first solve equation 2 for S




and Ca for low values of dC, and then use these values in equation 3 to




solve for R and Fc.  However, we have used the time integrated form of




equation (3) to estimate the sets  of model parameters that best reproduce




the observed time series of sediment concentration.  The model results were




then evaluated using Wilmot's  (1984) criteria.  Obviously, since  the model




does not take advection into account,  sediment concentration changes due  to




advection must at  least be identified,  and  if possible removed from  the




record, prior  to  determining  the  parameters.  This requires a set of




criteria  to  differentiate  between advection and  local  resuspension events.









      One  obvious  criteria for local resuspension is that the  increase  in




 sediment  concentration should occur at the  same  time as  an increase  in the




 forcing function.   Thus, if resuspension is due  to wave  activity, increases




 in sediment concentration should occur when wave action is greatest  (during




 storms).   We have used high values of Speed Standard Deviation (SSD) as an




 indication of wave  activity.  Since waves are wind-generated,  in the absence




 of direct or indirect wave measurements wind records might be useful in




 predicting resuspension activity.  There should also be (in the absence of




 advection) a characteristic decay time of suspended sediment concentrations




 which depends on the  settling speed of the  sediment and the height to which




 the sediment was resuspended.  All of the observations  in Lake St. Clair




 indicate  that  it  is vertically well mixed,  particularly during resuspension




 events,  so  differences in the decay  time should  depend  only on the  settling




  rates.   These variations are expected to be small.  Chriss and Pak  (1976)




  have also suggested that the standard deviation  of water  transparency should
                                        218

-------
be higher during resuspension events than during advection because the water




has not had as much time to become well mixed









     Ironically, the data sets that most unambiguously show local




resuspension were collected last.  Figure 4 shows the sediment concentration




record measured at station 42 during October, 1986.  The record extends only




to day 301 because the other two stations were retrieved at that time.  Thus




the data collected on days 303-309 are not discussed here.  The gaps in the




record are due to instrument problems; the recording tape was filled on day




296 and not replaced until day 301.  The very high concentrations (off the




scale) occurred when the transparency was zero.  For zero transparency,




equation 1 gives a TSM of of infinity.  We have used a value of 68 mg/1 -




equivalent to the lowest transparency we could measure.  The TSM record is




very well correlated with both the mean speed and SSD (Figs. 5 and 6).  In




fact the plots of the mean speed and its standard deviation are almost




identical.  All three plots show major changes beginning on day 287 and




extending until the measurements were interrupted on day 289.  The high TSM




levels, however, continue for several more days - until day 293 - even




though both the mean speed and the SSD are at background levels.  Most




probably the high TSM values during this time are due to advection of




material resuspended elsewhere.









     Figure 7 shows wave orbital velocities calculated for the site.  These




velocities were calculated by running the GLERL wave model using wind data




collected by CCIW near the station.  The wave model results were then used




as  input  to calculate the maximum orbital velocity one meter above the
                                      219

-------
bottom.  The pattern of the results is very similar to that of the mean




measured speed and the SSD, although there is a difference in magnitude of




about two.  The calculated velocities also decay more quickly than do the




measured values.  This is probably because the measured parameters also




include the effects of currents generated by the storm.  In spite of these




differences, the generally good agreement suggests that at least most of the




TSM record can be accounted for by local wave-induced resuspension.  The




temperature record (Fig. 8) shows a marked decrease during and after the




storm, another indication that advection was present. The standard deviation




of the transparency (TSD) in Figure 9 shows a peak during the storm, but




also several other peaks when no wave action was present.  Apparently, this




measurement is not an unequivocal indicator of local resuspension, at least




in Lake St. Clair.  The very large peak on day 293 is particularly




interesting because it does not appear to be related to any measure of flow,




either observed or calculated.









     Figure 10 shows the results from the model when the calculated orbital




velocity data is used as the forcing function.  The fit is remarkably good,




as evidenced by the high values of Wilmot's (1984) index of agreement (d)




and the unsystematic mean-squared error (MSEU, Table 2).  Here d is a




relative  index of agreement between the model and the observations based on




the summed  square error.  MSEU is that portion of the root mean-squared




error  that  is unsystematic.  Note that the model predicts a higher




concentration than the maximum measured, but recall that the transparency




meter  was saturated during this period.  Also note that the elevated TSM




measurements on days  289-301 are not predicted by the model, as would be
                                      220

-------
true if they are due to advection.  Although the main peak (on days 287-299)




is predicted quite well, the peak on day 308 is substantially less than the




predicted concentration.  This suggests that either R is too high or that




the resuspension flux is not a linear function of Fc but possibly a power




function of some sort.  The results agree quite well with those obtained




using SSD as the forcing function (Fig. 11), so we have used the wave model




results at the other two stations as the forcing function since we have no




current measurements at those sites.









     Figure 12 shows the TSM record at station 71 for the same period.  The




large increase on day 287 is also present here, as is the long decay.  In




addition, however, there is a pronounced peak on day 283.  The calculated




orbital velocities (Fig. 13) also show a marked peak on day 283, in contrast




to the record at station 42, where the peaks in both sediment concentration




and wave orbital velocity are much smaller.  Again, as at station 42, cooler




water is present during the storm and its aftermath (Fig. 14).  The TSD




(Fig. 15) shows two peaks -  one during the storm and the second just prior




to the dramatic lowering of the concentration on day 291.  Since this second




peak is not associated with wave action, it may indicate mixing of clearer




water from upstream with the more turbid water in the lake.  Comparison of




the model results with the observations are shown in Figure 16.  Again the




model slightly overpredicts the peak concentration, although in this case




the  transparency meter was not saturated.  The model also underpredicts the




peak concentration during the storm on day 283, and predicts several small




peaks that were not observed.  Overall, however, the model does a good job




of predicting the actual measurements.
                                      221

-------
     Results of the TSM measurements made at station 1 are shown in Figure




17.  The biggest difference between this station and the other two is the




two peaks on day 285.  It is very reassuring to find that a peak in




calculated wave orbital velocities also occurs on that day (Fig. 18), in




addition to the peaks on days 283 and 287.  The effects of advection after




the storm on day 287 are not as evident at this station, and there is no




peak in the TSD on day 293  (Fig. 19).  There are however several peaks later




in the record which are not correlated with wave activity.  These may be due




to inhomogeneities in the water entering  the lake.  The temperature record




(Fig. 20) is similar to those at the other stations.  The observed and




modeled results are shown in Figure 21.   The model  accurately predicts all




three peaks, although it either overestimates  or underestimates their




magnitude,  and  the decay curves are reasonably close  to the observed ones.









     Model  results for  the  stations  are  tabulated  in  Table 2, along with  the




percent mud (less  than  60 microns)  of  the bottom sediments as measured by




the  University  of  Windsor  (1985).   The stations for this  deployment were,  in




fact,  chosen so that the  sand percentages were approximately  equal  since  one




of the goals was  to  investigate the effect of differing wave  climates  on




 similar substrates.    Although the model results from stations  1  and 71 are




 in good agreement, the results from station 42 are somewhat  different  - all




 the parameters but S are much higher.   Some of this discrepancy may be  an




 artifact of the data however.  The higher ambient concentrations  are




 probably due mostly to the outflow of the Thames River, which has a high




 suspended  load.  If Ca is artificially high, then bottom  resuspension may be




 occurring  at lower values of Fc than is evident from the  data.  An
                                       222

-------
artificially high value of Fc could in turn lead to a higher value of R, if




in fact the resuspension rate is not linear as assumed here but is a power




function of F.









     The spring, 1986 deployment was designed to examine the effects of




changes in substrate on resuspension.  During this period the tripods were




all located within  8 km of each other in the northwest portion of the lake.




We hoped that putting the tripods close together would minimize the




differences in wave climate between stations.  Since we did not yet have our




meterological tower available, we put the tripods near the St. Glair Shores




Coast Guard station, which makes weather observations.  However, a




comparison of the weather records from St. Glair shores with on-lake




observations  later  in the year showed substantial differences between the




two.  Until a valid transfer  function can be developed, wave climates




estimated from  the  St.  Glair  Shores weather data are not accurate.  This




means  that we could not use calculated wave orbital velocities as the




forcing function in the model for  this deployment.  Since  only the current




meter  at station 3  worked,  only  that  data  set  could be analyzed,  since  we




felt  that,  given the  variability in the  TSM measurements between  the




stations,  it  would not  be justifiable to apply the current data from one




station to  another site.   Once a suitable  transfer function is developed, we




will  be able  to assess  the effects of substrate variability, but  this cannot




be done yet.









      The TSM record for station 3 (Fig.  22)  shows  several  pronounced peaks,




 one of which lasted for several days.  Unfortunately  the mean speed record
                                       223

-------
(Fig. 23) is so noisy that it is hard to see any obvious correlation.  The




record of the SSD (Fig.24) shows a correlation with some, but not all, of




the TSM peaks and also shows peaks where no TSM increase occurred.  We




believe the three TSM peaks after day 152 and the one on day 136 are due to




wave resuspension since they correlate with peaks in the SSD and decay




quickly.  The long episode starting on day 142 is probably due to advection




since both  the  speed and  SSD are low when it begins.  The peak on day 137  is




hard to  explain because it is not correlated with a peak in SSD but  decays




quickly.    The  temperature record  (Fig.  25) does not help much in




interpreting  the  results.  The model  results, using SSD as the forcing




function,  show  a  reasonable agreement  with  the data  (Fig. 26) though not as




good as  for the fall.   The most obvious  failure  is on day 149  (May  29) when




the model predicts  a non-existent  peak in TSM.   This  is because of  a peak in




SSD that day.  The  model  parameters (Table  2)  are  fairly consistent with the




results from the fall deployment.   Ca is lower,  reflecting  the  relative




 absence of sediment in the water,  and although Fc  is higher,  it is  a




 different parameter than in the fall.   The values  of R and S  are  close to




 those for the other western stations.









      The results from station  5 (Fig. 27-29) are very much like those from




 station 3.   The  same TSM peaks are seen, although their form is slightly




 different, particularly  for the (assumed) advection event.  The TSD record




 correlates well  with  some TSM  peaks,  particularly those late in the record,




 but overall  TSD  does  not seem  to be a reliable  indication of local




  resuspension.   The temperature record show the  same general pattern as at




  station 3.
                                       224

-------
     The tripod at station 1 worked for only part of the deployment.  The




TSD peaks (Fig. 31) correlate extremely well with the early peaks in TSM




(Fig. 30) but not with the TSM peak on day 142.  This latter peak, although




it looks like the one in the middle of the records at stations 3 and 5,




actually begins about a day earlier.  The temperature record (Fig. 32) shows




a minimum just before the peak in TSM further supporting the hypothesis that




it is due to advection.









     Stations  1,  5,  and 71 were occupied during  the  summer of 1986.




Stations 1  and 5  had been occupied  earlier in  the year, and station 71 was




near both our meteorological  tower  and a wave  rider  deployed by NESDIS of




Canada.  However,  several problems  complicate  the interpretation  of the data




from this deployment.   First,  algal growth during the deployment  period




fouled  the  transparency meters.  We endeavored to calibrate the fouling by




taking  vertical  profiles with a  clean meter  once a week at each station, but




unfortunately two of the  tripods stopped during  the  deployment so we have




only one complete calibration curve, and the results vary from station to




 station.  The TSM data shown has been  corrected  to  the  best of our  ability,




but is  not perfect.  In addition,  we also  found  that although  there were




very few significant wave events,  movement of turbid bottom water occurred




 fairly persistently, particularly at station 71, which  is where  our  current




 measurements were made.  It is thus extremely difficult to  say with any




 great confidence when resuspension occurred during this deployment.









      The TSM record from station 71 is much noisier than that  during  the




 fall deployment  (Fig.  33) and does not correlate very well with  either the
                                       225

-------
mean speed (Fig. 34) or the SSD (Fig. 35).  Nor does it seem to correlate




very well with the calculated orbital velocities (Fig. 36).  In fact the




best correlation seems to be with temperature (Fig. 37) which shows




extraordinary variability.  In this case, sharp drops in TSM occurred




simultaneously with abrupt increases in  temperature.  These temperature




increases are in turn correlated with periods of high wave orbital




velocities.  Thus, rather than causing resuspension, wave action appears to




be associated with minimums in TSM.  The explanation appears to be that




during the summer there is a thin bottom layer of more cooler, more turbid




water underlying the warmer, clearer water.  Wave action mixes the two and




brings the warmer water down nearer the  bottom.  Vertical profiles taken




during this period frequently show this  cooler, turbid layer (Fig. 38).  In




addition, the water temperature sensor on our meteorological station, which




was  3m below the surface, recorded temperatures between 22 and 24 degrees




during the deployment.    It seems likely then that  most of the TSM signal is




not  due  to resuspension but to vertical  movement of the upper surface of




this bottom turbid layer  past the sensor.   Not surprisingly, the model does




not  do very well with  this data set.









      The results  from  stations  5  and  1 are  somewhat less noisy.  At station




 5 the TSM (Fig. 39)  shows 3 pronounced peaks, and  a noticeable minimum




beginning on day 201.   This minimum  correlates with a  rise in temperature




 (Fig.  40) so it may also be due  to a  thinning of a bottom  turbid  layer.  The




 first two peaks are associated with  peaks in orbital velocity  (Fig. 41), but




 the last is not:  it may be due  to advection.  Again,  the TSD record  (Fig.




 42) is not much help in distinguishing  resuspension events.
                                       226

-------
     The TSM record at station 1 (Fig. 43) shows several pronounced peaks




which decay very quickly.  These peaks are well correlated the the




calculated wave orbital velocities (Fig. 44) and also have high TSD values




(Fig. 45).  There does not appear to be a consistent correlation with




temperature (Fig. 46), so it seems likely that these events are in fact due




to wave resuspension.  We have run the model for this station using the




parameters from the fall deployment with the wave  orbital velocities as the




forcing function.  The results, shown in Figure 47) are surprisingly good.




The model accurately predicts the occurrence, if not the actual




concentrations, for several of the TSM peaks.  The most noticeable defect is




the overly long decay times, which indicates that  S should be increased.




The relatively poor values for the index of agreement and MSEU are not too




surprising when one considers that most of the summer data is in the range




of the ambient concentration for the fall deployment.









     Both of the 1985 deployments were exploratory, and neither has been




analyzed  in any detail.  During the summer deployment, one of the axes of




the current meter failed, so we have no record of  either mean speed or SSD.




The transmittance and temperature records are shown in Figures 48 and 49.




During  the fall deployment, the tripod was placed  near one of the wave




stations  deployed by CCIW.  The current meter and  transparency meter were




both  set  up to  log  continuous 5 minute averages.   These records and the




temperature are  shown  in Figures 50-52.  Although  preliminary examination of




both data sets  shows  a  correlation between TSM and wave orbital velocities,




 the model results  for  these deployments are not yet available.
                                      227

-------
                                DISCUSSION









     It appears that most, if not all, resuspension in Lake St. Glair is due




to wave action.  Although it is difficult in some of the records to




distinguish between resuspension and advection, those events which can be




unambiguously identified are almost always associated with wave activity.




The good fits obtained from a very simple model which totally ignores both




advection and resuspension due to currents also indicates that wave action




is the primary cause of resuspension.  The wave orbital velocities




calculated from the results of the GLERL wave model serve very well as the




forcing function in the model.  These velocities are the maximum values




calculated using intermediate-water wave theory for a height one meter above




the bottom in a total depth of 6.5m.  Since the wave model gives significant




wave height and period as the output, the orbital velocities are not the




absolute maximum velocities, but the maximums for the significant waves,




which are somewhat smaller than the peak waves.   It is thus not surprising




that the calculated orbital velocities are somewhat smaller than the actual




measured velocities.  The good agreement between the patterns of calculated




orbital velocities and the measured standard deviations of the speed




indicate that  in general  the latter is a good analog for the former.




However, there are times  when high values of SSD are not correlated with




high orbital velocities,  so  the analogy is not exact.









     We had hoped  that  the  standard deviation of the transparency (TSD)




would  be  a good  indicator of local resuspension, but we found frequent




 instances  of high  TSD values which did not correlate with resuspension
                                      228

-------
events.  It appears that in Lake St. Clair the time and length scales for




resuspension and advection are too similar for TSD values to serve as a




useful distinguishing criterion. This means that  resuspension was identified




primarily by the simultaneous occurrence  of a rise in TSM and in orbital




velocity.









     The various model  results  show the most  consistency  for the values of S




and R.  However the predicted concentrations  are  relatively insensitive to




the values  of  these parameters.   Examination  of  the  predictions  indicate




that a higher  value of  S may improve the  fit  in  several cases by shortening




the  time  required to return to  ambient conditions.   This  in turn would




require  an increase in R in order to keep the peak concentrations  the same.




Another  solution would to be to use a more complicated model, possibly  one




 in which the resuspension rate is a power function of F,  as proposed by




 Lavelle et al  (1984).  However, given the limitations of our  measurements




 and the good agreement between the model and measured concentrations, a more




 complicated model may not provide much more insight.









      When  wave orbital  velocities are  used as the forcing function in the




 model, the critical value above which  resuspension occurs is less than 1




 cm/s  (except  at station 42 where we believe  that the very high ambient




 concentrations mask the actual  initiation of resuspension).  Although an




 extrapolation of  F to  the bottom is fraught  with peril (recall that  it is  a




 calculated number - not measured),  it appears that  sand-sized material is




 unlikely to be resuspended at  these low  values.  Since sand is resuspended




  during at least some of the resuspension events  (as evidenced by the sand
                                       229

-------
found in the CCIW traps), there is probably a second, higher value of F,




which applies to the coarser material.  The value determined from the model




is for the finer material which causes the decrease in transparency.  The




variability in Fc between stations may be due to differences in substrate




characteristics, but the only  such measures available (X sand, gravel, and




mud) are not sufficient to  explain the differences.  Although there  is




considerable variation in Fc between stations,  the model parameters  appear




to be fairly constant  through  time.   The good agreement between the  model




results for the  summer deployment at station one, which were obtained using




the parameter values calculated  from the fall deployment, and the




observations, shows  that  the model has good predictive capability and lends




credence  to the  other  results.  Further  tests of  the model using the other




summer  and spring  data will be attempted in the future.
                                 CONCLUSIONS









      Our results show that resuspension in Lake St.  Glair is due mainly to




 wave action.  When wave orbital velocities are used as the forcing function




 in a simple model, the agreement between the predicted and observed




 instances of sediment resuspension is quite good.  Critical values of the




 orbital velocity are less than  one cm/s when calculated at one meter above




 the bottom.  These velocities can be calculated from the results of the




 GLERL wave  model  if on-lake wind records are available.  Variations in the




 critical velocity between sites may be due to differences in substrate




 characteristics, but there is no adequate data to test this.
                                       230

-------
                             LITERATURE CITED









Chriss, T.M. and H.J. Pak, 1978, Optical evidence for sediment




     resuspension-Oregon continental shelf, EOS, 59, p 410.









Great Lakes Institute, University of Windsor, 1985, A case study of




     selected contaminants in the Essex Region, Vol 1: Physical Sciences,




     contract report  for DSS contract UP-175.









Lavelle, J.W., Mofield, H.O., and E.T. Baker, 1984, An in  situ erosion




     rate  for a  fine-grained marine sediment, Jl. Geophys.




     Res.,89,6543-6553.









 Simons,  T.J.  and W.M. Schertzer, 1986, Modeling wave-induced  sediment




     resuspension in Lake St. Glair,  unpublished MS,  NWRI.









 Wilmot,  C. J.,1984, some  comments on  the evaluation of model  performance,




      Bull. Am.  Meteor. Soc.,  63,  1309-1313.
                                       231

-------
      Dates




 7/11/85  -   8/8/85




 9/10/85  -  10/9/85




 5/15/86  -  5/24/86




 5/15/86  -  6/6/86




 5/15/86  -  6/6/86




 7/08/86  -  7/27/86




 7/08/86  -  7/31/86




 7/08/86  -  8/8/86




10/10/86  -  10/28/86




10/10/86  -  11/11/86




10/10/86  -  10/28/86
                                   TABLE
                             TRIPOD DEPLOYMENTS
Station #
42

71

1

3

5

1

5

71

1

42

71

Location
42°23'45nN
82°42'03"W
42°24'55"N
82°41'45"U
42*31' 18-N
82°44'48"W
42"29'42"N
82°47'42"W
42°28'06"N
82°47'24"W
42°31'18"N
82°44'48"W
42°28'00"N
82°47'18"W
42'25'00-N
82"40'48"W
42'31'11-N
82°44'49"W
42°23'08BN
82-32'28-W
42°24'58'tN
82°40'38"W
Tripod
ANL

ANL

GLERL

ANL

GLERL

GLERL

GLERL

ANL

GLERL

ANL

GLERL

                 All moorings were in 6-7m of water.
                                232

-------
                              Table 2

                            Model Results
Station
% Mud
                      (mg/1)   (cm/s)   (cm/s)   (mg/1)
                                                                   MSEU
42-fall
71-fall
 1-fall

 3-spring
 5-spring
 1-spring

 1-summer
71-summer
 5-summer
  25
  35
  30

  39
  53
  30

  30
  35
  53
6.4
3.8
4.3

1.3
4.3
2.8
0.9
0.1

2.9
0.1
0.0055
0.0063
0.0100

0.0033
 0.100
0.17
0.05
0.05

0.02
0.05
.947
,969
.922

.870
.616
.999
1.00
.995

.980
                                           .757
                                233

-------
                              LIST OF FIGURES


Figure 1.   Deployment locations.

Figure 2.   Transparency calibration locations.

Figure 3.   Transparency-TSM calibration curve.

Figure 4.   TSM plot, station 42, October, 1986

Figure 5.   Speed plot, station 42, October, 1986

Figure 6.   Standard deviation of  the  speed, station 42, October, 1986

Figure 7.   Calculated wave orbital velocity, station 42, October 1986

Figure 8.   Temperature, station 42, October, 1986

Figure 9.   Standard deviation of  the  transparency, station 42, October, 1986

Figure 10.  Comparison of the observed and calculated TSM using wave orbital
            velocity as the forcing  function, station 42, October, 1986

Figure 11.  Comparison of the observed and calculated TSM using the standard
            deviation of the  speed as the forcing function, station 42,
            October, 1986

Figure 12.  TSM, station 71,  October, 1986

Figure 13.  Calculated wave orbital velocity, station 71, October, 1986

Figure 14.  Temperature, station  71,  October, 1986

Figure 15.  Standard deviation  of the transparency, station 71, October,
            1986

Figure 16.  Observed and calculated  TSM  using the wave  orbital velocity as
            the  forcing function,  station  71, October,  1986

Figure 17.  TSM,  station 1, October,  1986

Figure 18.  Calculated wave orbital  velocity, station 1, October, 1986

Figure  19.  Standard deviation  of the transparency, station 1, October, 1986

Figure  20.  Temperature,  station 1,  October,  1986

 Figure 21.  Observed and  calculated  TSM  using wave orbital velocity as the
             forcing function, station 1, October, 1986
                                     234

-------
Figure 22.  TSM, station 3, May-June, 1986

Figure 23.  Speed, station 3, May-June, 1986

Figure 24.  Standard deviation of speed, station 3, May-June, 1986

Figure 25.  Temperature, station 3, May-June, 1986

Figure 26.  Observed and calculated TSM using the standard deviation of the
            speed as the forcing function, station 3, May-June, 1986

Figure 27.  TSM, station 5, May-June, 1986

Figure 28.  Standard deviation of the transparency, station 5  May-June
            1986

Figure 29.  Temperature, station 5, May-June, 1986

Figure 30.  TSM, station 1, May-June, 1986

Figure 31.  Standard deviation of the transparency, station 1, May-June
            1986

Figure 32.  Temperature, station 1, May-June, 1986

Figure 33.  TSM, station 71, July-August, 1986

Figure 34.  Speed, station 71, July-August, 1986

Figure 35.  Standard deviation of the speed, station 71, July-August, 1986

Figure 36.  Calculated wave orbital velocity, station 71, July-August, 1986

Figure 37.  Temperature, station 71, July-August, 1986

Figure 38.  TSM profile, station 71, July 22, 1986

Figure 39.  TSM station 5, July-August, 1986

Figure 40.  Temperature, station 5, July-August, 1986

Figure 41.  Calculated wave  orbital velocity, station 5, July-August, 1986

Figure 42.  Standard deviation  of  the  transparency, station 5  July-August
            1986

Figure 43.  TSM,  station 1,  July-August, 1986

Figure 44.  Calculated wave  orbital velocity, station 1, July-August, 1986

 Figure 45.  Standard deviation  of  the transparency, station 1  July-August
            1986
                                      235

-------
Figure 46.   Temperature, station 1, July-August,  1986

Figure 47.   Observed and calculated TSM using the wave orbital velocity as
            the forcing function, station 1, July-August, 1985

Figure 48.   Transparency.station 42, July-August, 1985

Figure 49.   Temperature, station 42, July-August, 1985

Figure 50.   Transparency, station 71, September,  1985

Figure 51.   Speed, station 71, September, 1985

Figure 52.   Temperature, station 71, September 1985
                                    236

-------
237

-------
138

-------
 7)
 £
 I/O
                                                 TSM  CRLIBRRTION  CURVE
I)
•i'
                              til
                              li-
                              ft
                               r
                                 0.0
                                        5.0
                                                10. 0
                                                       15.0
                                                               20.0
                                                           TSM (MG/L)
                                                                       25.0
                                                                              30.0
                                                                                      35.0
                                                                                              •10.0

-------
                                                STA 42, OCT 10-NOV 8, 1986
                           o
                           Si


                           3H
                           q
                           81
                           q
                           n-
ho
*-
o
                           q
                           iri-
                           q
                           £H
                           q
                           a
                           q
                           id
                             283.0    285.0    287.0     289.0     291.0     293.0

                                                         JULIAN DAY 1986
295.0
        297.0
                299.0
                        301.0

-------
                   STA 42, OCT 10-NOV 8,  1986
 7)
cs'
283.0
       205.0
              287.0
                      269.0
  291.0     293.0

JULIAN DAY 1986
                                             295.0
                                                     297.0
                                                             299.0
                                                                     301.0

-------
                  STA 42, OCT 10-NOV 0, 1906
283.0   285.0    207.0     280.0    201.0     203.0
                            JUUAN DAY 1986
                                             305.0
                                                    207.0
                                                            209.0
                                                                    301.0

-------
                               STATION 42 10/1-11/15/86 STARTS AT 0000 ESTT
                         o
                         M-
                         o
                         -J-
                         o
                         b-
NJ
4^
LO
                        1SJ
                          3-
                          q
                          6-
                           263.0   285.0     287.0    289.0     291.0     293.0

                                                      JULIAN DAY 1986
                                                                      295.0
                                                                              297.0
                                                                                     299.0
                                                                                            301.0

-------
                       STA 42, OCT 10-NOV 0,1906
  q
  iri-
  q
  ri-
  q
  si-
  o
  6-
  q
  o>-
a,
  3-
  SH
  o
  evfH
  §.
—I	
 285.0
—I	
 287.0
—r
 289.0
 1	
297.0
    283.0
                                   201.0     293.0
                                 JULIAN DAY 1986
295.0
299.0
—l
 301.0

-------
                                                   STA 42, OCT 10-NOV 8, 1986
                              8-
                              S-i
                              o
                            Q o-
                            CO 6
                                283.0    885.0     207.0     269.0     291.0     293.0

                                                             JULIAN DAY 1986
295.0
297.0
299.0
                        301.0
-0

-------
                   80-i
                CK
                LJ
Ld
0,
CO
13
CO
               O
                    60
               O  40
               UJ
               O
                   20-
                    0
                           STATION H2 - OBSERVED AND MODELED TSM
                           FLOW PARAMETER - MODELED WAVE ORBITAL SPEED
                           MODEL PARAMETERS
                           CA=6.4 FPC=2.8 S=0.0055 R=0.17
                                    OBSFRVFD

                                    PREDICTED
      283    287    291    295    299    303

                     JULIAN DAY - 1986
                                                             307
311
o

-------
NJ
         80-i
      U)
      £
         60-
     Ld
     !<
     O
     UJ
     O
40-
     LJ
     Q_
     (f)
     ID
20-
     o
          0
           283    287
                       STATION Ml - OBSERVED AND MODELED TSM
                       FLOW PARAMETER - SPEED STANDARD DEVIATION
                       MODEL PARAMETERS
                       CA=8.0 FPC=6.5 S=0.006 R=0.225
                 I—i—i—i—|—i—i—i—|—i—i—i—|—r
                291    295    299    303

                  JULIAN DAY - 1986
307    311

-------
                                                             STATION S710CT
                               q

                               s
K3
4>

CO
                               q
                               a-
                               q
                               8
                               q
                               to
                               O
                               a
                                o
                                o
                                e
                                cd
                                O
                                ol
                                o
                                6
                                  383.0    285.0     287.0     289.0     201.0     293.0     295.0

                                                                 JULIAN DAY 1986
297.0
299.0
301.0

-------
                                STATION 7110/1-11/15/86 STARTS AT 0000
                            283.0
286.0
287.0
289.0
V?
  201.0    293.0
JULIAN DAY 1988
                                                                       296.0
                                            297.0
                                            299.0
                                            301.0

-------
                                                           STATION S710CT
                             1-
fO
^n
O
                                283.0    285.0     287.0     269.0     291.0     293.0     295.0
                                                              JUUAN DAY 1986
297.0
299.0
301.0

-------
                                                           STATION S710CT
ho
Ui
                             8
                             6
                                283.0    285.0     287.0     289.0     291.0     293.0

                                                              JULIAN DAY 1986
295.0
297.0
299.0
                                                                                                          301.0

-------
NO
l_n
KJ
                O)
               O
               LJ
               Q
               CL
               V)
               Q
                   25-1
                   20-
                    15-
                   10-
                     283
   STATION 71 - OBSERVED AND MODELED TSM
   FLOW PARAMETER - MODELED WAVE ORBITAL SPEED
   MODEL PARAMETERS
   CA=3.8 FPC=0.9 S=0.0063 R=0.050
                                           1 - 1
287
 ,

295
    291
JULIAN DAY - 1986
299
                              OBSERVED

                              PREDICTED
-•—I
 303

-------
                                                              STATION Sl-OCT
ISJ
                               o
                               M-
                               p
                               O
                                 283.0    285.0     287.0     289.0     291.0     293.0     2950

                                                                JULIAN DAY 1986
297.0
         299.0
                 301.0

-------
                                 STATION  110/1-11/15/86 STARTS AT 0000
NJ
(J\
ff
                           S-,
                           o
                          ffs.
                              283.0   2B5.0     287.0
280.0     291.0     203.0
      JULIAN DAY 1986
                                                                         205.0
297.0
200.0
301.0

-------
                                                           STATION Sl-OCT
Ol
                              3
                              d'


                              2
                              6
                                283.0
—I	
 285.0
287.0     289.0     291.0     293.0

               JULIAN DAY 1986
                                                                                 295.0
                                         297.0
                                                                                                 299.0
                                                          301.0

-------
                                                           STATION Sl-OCT
                             q
                             ri-
                             q
                             N-
                             o
                             j.
                             q
                             d-
                             q
                             o>
 Ul
                             P
                             
-------
               O>
               a:
               o
               Q
               Q_
               o
                   25-1
                   20-
                   15
                   10
               13
               00    5-
                    0
         STATION 1 - OBSERVED AND MODELED TSM
         FLOW PARAMETER - MODELED WAVE ORBITAL SPEED
         MODEL PARAMETERS
         CA=4.3 FPC=0.1 S=0.01 R=0.05
                    283
                                    OBSERVED.

                                    PREDldTED
287       29T       295
      JULIAN DAY - 1986
—i—
299
                                                                    303
u

-------
                                                STATION S3 MAY-JUNE, 1906
 CD
                             135.0   137.0
130.0
                                               141.0   143.0   145.0   147.0   149.0
                                                         JULIAN DAY 1986
                                      151.0
                                            153.0
                                                  155.0
                                                         157.0
N

-------
                                                  STATION S3 MAY-JUNE, 1986
                            o
                            0-1
K3
Ui
                                 _,	,	.	p.
13S.O   137.0   139.0    141.0    143.0    145.0   147.0    149.0
                              JULIAN DAY 1986
                                                                                   101.0
                                                                                          153.0
                                                                                                 155.0
 1
157.0

-------
                                                  STATION S3 MAY-JUNE, 1986
                              135.0   137.0
                                          139.0
                                                 141.0
                                                       143.0   145.0    147.0   149.0
                                                           JULIAN DAY 1986
151.0
      153.0
             155.0
                    157.0
N
-c

-------
                                                STATION S3 MAY-JUNE, 1986
                           q
                           oo
                           :
                            2-
                                         —I	1	
                                          139.0   141.0
—i	1	r
 143.0   145.0    147.0   149.0

    JULIAN DAY 1986
                              135.0  137.0
                                                                                 151.0    153.0    155.0   157.0
KJ

-------
12-j CONCENTRATION
 <

 4-1
     PRB5KTED CONCENTRATION (fflg/U
 a-
 4-
   16
  MAY
  1966
T
 19
23
                           i
                          27
31
 4
JUNE
                               8
                            262

-------
                                STATION  S5-MJ
S.
o
ri
o
Si-
O
CV1-
135.0   137.0    130.0    141.0    143.0    145.0    147.0    149.0
                                 JULIAN DAY 1986
                                                             151.0
                                                                    153.0
                                                                            155.0
                                                                                   157.0

-------
                                                                STATION  S5-MJ
 NJ
 7)
o'
                                   135.0  137.0   130.0    141.0    143.0    145.0    147.0    140.0

                                                                   JULIAN DAY 1986
                                                                                            151.0
153.0
                                                                                                           155.0
—l

 157.0

-------
                                                                STATION  S5-MJ
                                p
                                in-i
 N3
                                o

                                ri-
                                sl-
                                o
                                3-
                                   135.0   137.0    139.0    Ul.O
                                                               143.0    145.0    147.0    149.0


                                                                   JULIAN  DAY 1986
151.0
—I	1	1

 153.0    155.0    157.0
—n

-------
                                STATION  Sl-MJ
S.
   135.0   137.0    139.0    141.0    143.0    145.0    147.0    149.0
                                   JULIAN DAY 1986
151.0
       153.0
               155.0
                      157.0

-------
                                                                STATION   Sl-MJ
NJ
                                &u
                                o
                                CM
                                *•
                                O


                                o
                                o


                                §
                                d


                                s
                                d


                                s
                                6


                                3
                                o
                                  135.0   137.0    130.0    141.0    143.0    145.0    147.0    149.0

                                                                   JULIAN DAY 1986
                                                                                            151.0
                                                                                                   1530
155.0
                                                                                                                  157.0

-------
                                                                    STTATION  Sl-MJ
   i-o
   ON
   00
                                   £3-
                                 Si 9
 71
cs>
135.0  137.0    139.0    141.0    143.0    145.0    147.0     149.0

                                  JULIAN DAY 1986
                                                                                                  151.0
153.0
155.0
157.0

-------
                                                     STA  71, JULY  6- AUGUST 7,
NJ
ox
                                189.0 191.0  193.0  195.0  197.0 199.0  201.0  203.0  205.0  207.0 209.0  211.0  213.0  215.0  217.0 219.0

                                                               JULIAN DAY 1986

-------
                                                     STA 71, JULY  6- AUGUST 7,
to
•vj
o
                              SH
                              3j
                              o
                              6-
                              o
                              oi-
                            W o
                            U iC
                            (X,
                                189.0 191.0  193.0  105.0  107.0  190.0  201.0 203.0  205.0  207.0  209.0  211.0  213.0  215.0  217.0 210.0

                                                                JULIAN DAY 1986

-------
                         STA 71, JULY  6- AUGUST 7,
  9
  8-


  3.
  N
  o
  81
65*
     189.0 191.0  193.0  195.0  197.0  199.0 201.0 203.0  205.0  207.0  209.0  211.0 213.0  215.0  217.0  219.0

                                    JULIAN DAY 1986

-------
                                   STATION 71 7/10-9/27/86 STARTS AT 0600
Ni
•-J
N5
                           o
                           ri
                          fc
jjj
                                                                                              ..  it
                               18901010  1930  195.0  197.0 199.0  201.0  203.0 205.0 207.0  209.0  211.0  213.0 215.0 217.0  219.0

                                                           JULIAN DAY 1986
 1

-------
                    STA 71, JULY   6- AUGUST 7,
189.0 191.0  193.0  105.0 107.0  199.0  201.0  203.0  205.0  207.0 209.0  211.0  213.0  215.0  217.0  219.0
                               JULIAN DAY 1986

-------
KJ
                                                Station: STA
                                         • i
                                      1
                                          I-
                                                                                                                           - t
                                                                                                                               a
                                                                                                                               I
                                           o   a   10   i&   a>   a   so   an   40   4»   BO   u   to
                                                                               TRIMS M

-------
                                                                  STATION   S5JA
NJ
                                S-,
                                o
                                ci
                                o

                                N
                              §8-
                              S 9
                              -- r-
                                  189.0 191.0  193.0  105.0   197.0  199.0  201.0  203.0  205.0  207.0  209.0  211.0  213.0  215.0  217.0  219.0


                                                                    JULIAN DAY 1986

-------
                                    STATION   S5JA
  q
  n -
  N
  14
  q

  o»-
a,
   o
     189.0 191.0  193.0  195.0  197.0  199.0 201.0  203.0  205.0 207.0  209.0  211.0  213.0  215.0  217.0  219.0

                                       JULIAN DAY 1986

-------
                                       STATION 5 7/10-9/27/36 STARTS AT O600
                               o
                               rj
                               q
                               a
                                                            JLL
                                                                                              lA A.
                                                                                    *t*  "  ' '"f^1* '" iif -Tilt «T. ^ IT I .t-j- j ......

                                  189.0 191.0 193.0 195.0  197.0 199.0 201.0 203.0 205.0 207.0 209.0  211.0  213.0  215.0 217.0  219.0

                                                                JULIAN DAY 1986
es
 -t

-------
                                                                    STATION    S5JA
K)
vj

00
                                 2
                                 6
                                 o
                                 6


                                 8.
                                 8
                                 b


                                 3
                                 6
                                 o

                                 8
                                   189.0 191.0   193.0  195.0  107.0  199.0  201.0  203.0  205.0  207.0  209.0  211.0  213.0  215.0  217.0  210.0

                                                                      JULIAN DAY 1986

-------
                                                               STATION STA1-J
                               q
                               ri
                               q
                               Ji
to
                                   890 191.0  193.0  195.0  197.0  199.0  201.0  203.0 205.0  207.0 209.0  211.0  213.0  218.0  217.0  219.0

                                                                   JULIAN DAY 1986

-------
                                     STATION 1 7/IQ-9/27/Q6 STARTS AT 0600
                             o
                             n
NJ
00
O
                                109.0 191.0  193.0  195.0  197.0  199.0 201.0 203.0 205.0  207.0  209.0  211.0  213.0 215.0 217.0  219.0

                                                             JULIAN DAY 1986

-------
                                                                STATION STA1-J
ISi
oo
                                i
                                   890 191.0  193.0  195.0  197.0  199.0  801.0  203.0 205.0  207.0 209.0  211.0  213.0  215.0  217.0  219.0
                                                                    JULIAN DAY 1986

-------
                                                                 STATION STA1-J
                                o
                                id
                                q
                                sl-
00
INJ
                              Cu
                                   189.0 191.0  193.0  195.0  197.0  199.0  201.0  203.0  205.0 207.0  209.0  211.0  213.0  215.0  217.0  219.0

                                                                    JULIAN DAY 1986

-------
CO
LO
                   12
               D)
               E  10
                   8-
                   6-
UJ
I

o
              o
              z
              UJ
              OL
              CO
              ID
              CO
              O
                   0
                         STATION 1 - OBSERVED AND MODELED TSM
                         FLOW PARAMETER - MODELED WAVE ORBITAL SPEED
                         MODEL PARAMETERS
                         CA=4.3 FPC=0.1 S=0.01 R=0.05
                         DERIVED FROM FALL DATA
                                                                OBSERVED

                                                                PREDICTED
                    190
                  195          200

                      JULIAN DAY - 1986
205
210

-------
NJ
oo
          60-1
8  60
Z
<
      t
t:  40-
          30-
co
z

QC
h-

Z  20
LLJ
O
QC
LLJ  10
CL
            0
               LAKE ST. CLAIR
                   850203
              11
            JULY
            1985
              16
~r~r
  19
23
-J

27
31
 4
AUG
8

-------
           24-
00
Oi
       O  23
       UJ
       cc
       D
       !<
       n:
       IU
       CL
       LU
       h-
22-
21-
           20
                LAKE ST. CLAIR
                   860203
I " -" r T-
11
JULY
1085
1 1 '
15
1 1 '
10
• • , -r
23
1 ' I ' '
27
1 1 '
31
4
AUG
8

-------
00
ox
UJ
O
       LU
       O
       CC
       UJ
       O.
           75-i
       E  60 H


       2!
            25-
             0
D
               11
               FPT
                i
                16
21
 i
26
 1
OCT
i
6
11

-------
           15-1
oo
       Q
       in
       in
       Q.
       CO
                LAKE ST. CLAIR
                   860206
              n
            SEPT
            1086
— 1 — 1 — p^ — I — 1
16

i.ji.i
21

. I . .
26

1 ' I ' '
1
OCT
• • I • •
6

1 • i
11


-------
             22-1
00
CO
         o
         O

         LU
         CC.
         DC
         UJ
         UJ
18-
              14-
              10
                           LAKE ST. CLAIR
                              860206
                11
               SEPT
               1985
          T
           16
T"
 21
26
        T
                                     T—r
 1
OCT
"T
 6
11

-------
          ACCUMULATION OF FALLOUT CESIUM-137 AND CHLORINATED ORGANIC




              CONTAMINANTS IN RECENT SEDIMENTS OF LAKE ST.  GLAIR




                  John A. Robbins and Barry G. Oliver*
                                ABSTRACT









    Cesium-137 originating from atmospheric nuclear  testing,  lead-210,




potassium-40 and several chlorinated organic compounds  (HCB,




Hexachlorobenzene; OCS, Octachlorostyrene; PCBs,  Polychlorinated Biphenyls;




HCBD, Hexachlorobutadiene; QCB, Pentachlorobenzene; TCB,  Total




Trichlorobenzene; TeCB, Total Tetrachlorobenzene,  and Total DDT) have been




measured  in a set of sediment cores collected by  diver  from Lake St.  Clair




in 1985.  Distributions of Cs-137 often  show subsurface peaks  which




apparently correspond  to peak testing  in the mid  1960s.   Excess lead-210




distributions are similar to those previously encountered elsewhere in the




Great Lakes, possessing a zone of constant activity extending  about 3 cm




down from the surface  with exponential fall off below.  Mixed  depths and




sedimentation rates inferred from lead-210 profiles are consistent with Cs-




137 profiles and indicate sedimentation  rates of  the  order of  0.1-0.2 cm/yr.




Observed  profiles can  also be correctly  predicted by  mixing processes




(rather than sedimentation) in which mixing below the surface  region is
 National Water Research  Institute,  Burlington,  Ontario,  Canada
                                   289

-------
characterized by constant eddy diffusion coefficients of .2 to 4. cm2/yr.




Because of the evidence of extensive biological activity down to at least 15




cm,  mixing is favored as the mechanism producing observed radionuclide and




contaminant distributions.  Comparison of total loading of Cs-137 to the




lake (471 Ci in 1985) to actual storage (37 Ci) indicates a sediment




residence time of about 5 years.  This value is consistent with that




inferred from previously observed changes in surficial sediment levels of




mercury and pesticides between 1970 and 1974.  Changes in surficial sediment




Cs-137 concentrations between 1976 and 1985 are less than expected on the




basis of a five year residence time and suggest that the residence time of




particle-associated contaminants increases with the amount of time the




contaminant has remained in the system.  Preliminary measurements of Cs-137




in trap samples collected by others at two sites in the lake indicate that




the isotope may be used to distinguish between particle-associated




contaminants resuspended from the bottom and new contributions from the St.




Clair River.  Concentrations of chlorinated organic compounds in surface




sediments are well-correlated with each other and the pattern over the lake




bottom is closely related to the thickness of recent deposits.  In some




cores profiles of total DDT and PCBs (as well as compounds with significant




local sources such as HCB, QCB and DCS) apparently reflect the history of




loading to the lake.  Also, profiles of HCB/OCS and HCB/QCB ratios, used to




distinguish between alternative industrial sources of the chemicals,




indicate the changing history of waste management practices.  Total storage




of  contaminants shows that Lake St. Clair sediments are a significant




repository of chemicals passing through the Lake.  As of 1985 it is
                                    290

-------
estimated that 960 Kg of HCB,  870 Kg of PCBs and 210 Kg of DCS are contained




in the sediments.
                               INTRODUCTION









     The enormous flow of water water from the upper Great Lakes proceeds




out of Lake Huron through the St. Clair River and into a shallow heart-




shaped body of water, Lake St. Clair.  Urban and industrial activity on the




shores have made the river and this lake one of the most heavily




contaminated regions of the Great Lakes.   Although the lake has an extremely




short mean hydraulic residence time, about 9 days, sediments manage to




acquire significant burdens of contaminants originating from tributary




sources, especially the St. Clair River.   Previous studies of mercury




(Thomas et al., 1977) and chlorinated organic compounds (Frank et. al, 1977)




indicated that contaminants passing through the lake are temporarily




retained by sediments.   The marked reduction which took place during the




four year period between 1970 and 1974 in the concentration of these




constituents in surficial sediments indicated that the sediment reservoir




was not a permanent sink however probably because resuspension in this




shallow lake (3m mean depth) ultimately exports materials out of the lake




and down the Detroit river.









     These qualitatively characterized properties of the sediment reservoir




can be more quantitatively treated by determining the storage and




distribution of particle-associated radiotracers whose loading histories, in
                                    291

-------
contrast to the above contaminants, are well-known.  Fallout cesium-137 is




especially useful for this purpose since it is a tracer of fine-grained




(clay sized) constituents and has an accurately known input history.  A




second isotope, lead-210 is also of considerable use, since it is delivered




to the lakes at a virtually constant, well-determined rate and has no known




anthropogenic sources.  Thus Cs-137 can serve to illustrate the long-term




response of sediments to a pulse of radioactivity passing through the system




in the mid 1960s while lead-210 characterizes the steady-state response.









     The recent severe contamination of the St. Glair River with chlorinated




organics from industrial activity  in the Sarnia area has been documented




(Environment Canada, 1986, J. Water Poll. Res., 1986).  It is known  that




Lake St. Clair retains some portion of these contaminants since they have




been found at significant concentrations in the Lake's  sediments  (Frank et




al., 1977; Pugsley et al.,1985; Oliver and Bourbonniere, 1985).  This report




also examines the distributions of the contaminants  in  surficial sediments




of the  Lake, estimates the mass of contaminants stored  in sediments, and




discusses contaminant trends  in sediment cores.
                                  METHODS









      Sediment cores  were collected by diver at the sites shown in Fig la.




 during May and September 1985.   The locations were chosen to coincide with




 those occupied previously by Pugsley et al. 1986.  Locations possessing sand




 or coarser materials are unsuited for hand insertion of core tubes and, at
                                     292

-------
least for cesium-37,  are known to possess insignificant amounts of the




isotope (cf Robbins,  1986).  Two 10.1 cm i.d. cores were collected at each




location with no observed disturbance of the sediment-water interface and




were suitably spaced from one another (ca 50 cm) to avoid interferences.




Cores were extruded hydraulically in the field and sectioned in 1 cm




intervals to 10 cm and in 2 cm from 10 cm to the bottom of the core or in




many cases to the interface with gray glacial clay.  Sections from one core




were subdivided for radionuclide and metals analysis while sections from the




replicate core were subdivided for organics and additional metals analysis.




Because the accuracy of sediment tracer and contaminant inventories depends




directly on the extent of recovery of material collected in a tube of known




area, care was taken to collect all sediment from each section.  Subsamples




for organics analysis were stored frozen in precleaned glass jars prior to




analysis.









     Samples for radionuclide analysis were weighed, freeze dried and




reweighed to obtain the fractional dry weight of sediment and the mass of




dry sediment per unit area of the core.  Dried samples were lightly




disaggregated and placed in vials of standard geometry for gamma counting.




The activity of Cesium-137 (661.6 KeV) and K-40 (1460.7 KeV) were determined




using a lithium-drifted germanium detector coupled to a multichannel




analyzer.  Absolute detector efficiencies were determined by counting




sediment samples of equivalent geometry spiked with a mixture of gamma




emitting radionuclides of known activity.  Uncertainties in sample




activities were generally well under ten percent.  Samples for organics




analysis were soxhlet extracted prior to analysis by capillary, electron
                                    293

-------
capture  detector,  gas chromatographic analysis.   Procedure details have been




published (Oliver  and Bourbonniere,  1985).









     In  addition,  sediment trap samples were analyzed for Cs-137 from two of




nine sites occupied during 1985 by Charlton and Oliver (1986).  Trap




materials were collected at nine (about equally spaced) time intervals




during the period from early June to mid-November.  The sites 10 and 13 are




situated approximately 1 Km north of coring locations 24 and 28 respectively




(See Fig la.).  Samples were freeze dried and lightly disaggregated prior to




analysis. Details of the collection and sample preparation as well as




supplementary analytical results are given in the above reference.
                          RESULTS AND DISCUSSION









 Characteristics of Sediment Cores









     In general sediment cores collected from this Lake evidence significant




 stratigraphic inhomogeneity.  Gelatinous, flocculent material resides  on  the




 uppermost centimeter or so of sediment.  The surface, particularly  in  the




 deepest parts of the Lake is perforated by many circular holes  about 0.5  cm




 in diameter which extend as burrows  into sediments as deep as 16 cm.




 Densities of perforations reached as high as 800 M"2.  The burrows,




 surrounded by about a  2 cm diameter  halo of oxidized sediment were  often




 inhabited by Hexagenia larvae.   In regions of the lake with  high clay




 content, the upper 10-20 cm of sediments consisted of brownish  gray silty
                                    294

-------
material with some fine sand and occasional layers of shell fragments.  In a




few of the cores a transition layer was encountered between 10 and 20 cm




consisting of clay with a high water content.  An admixture of clay and sand




extended below this layer becoming consolidated and possessing increasing




sand content.  In nearly all cores, at sufficient depths, the sediment




composition shifted abruptly to fine post-glacial clay.  Sediments above




this layer may be considered to be of recent origin.  The thickness of the




recent sediments, shown in Fig. lb., corresponds roughly with lake depth and




reaches a maximum of over 30 cm in a narrowly defined area.  Cores collected




in shallower parts of the lake with low clay content occasionally possessed




macrophyte fragments and remnants of plants rooted in the sediments




collected.  A large live clam (Unionid) was found in the region of 2-6 cm




depth in a sediment core (discarded) collected at site 39.









Sources of Radiocesium in Lake St. Clair









     Cesium-137 in the Great Lakes originates almost exclusively from




atmospheric nuclear testing.  There is no natural source of the radionuclide




and contributions from reactors situated along the shores are many orders of




magnitude less than direct fallout loadings.  More than two decades of




routine monitoring of monthly fallout by U.S. and Canadian authorities have




provided the basis for developing estimates of time varying loadings which




greatly surpass, both in quantity and quality, our knowledge of any other




contaminant entering the lake system.  The most detailed records are for Sr-




90 which is co-produced with radiocesium in nuclear detonations as a fission




product. Cesium-137 and Sr-90 have comparable half-lives (0.2 and 28.1
                                    295

-------
years  respectively)  and the ratio of the two isotopes in air and in fallout




is well-determined and largely time-invariant.  For this reason loadings of




Cs-137 may be reliably inferred from Sr-90 records as well as from direct




measurement.  Data for both Sr-90 and Cs-137 from all monitoring sites in the




Great  Lakes region have been combined with empirical models relating




atmospheric concentrations to over-lake and drainage basin deposition




(Robbins,  1985a).  This regional source function is used to predict loadings




to Lake  St. Clair using a lake system response model.









     The response of the Great Lakes to fallout radionuclide loadings was




first examined by Lerman and Taniguchi (1972) for Sr-90 and subsequently by




others for the two long-lived fallout isotopes which are much more strongly




associated with particulate matter, Cs-137 and plutonium.  Wahlgren et al.




(1980) related plutonium isotope loadings to measured aqueous concentrations




using a long term fate model which included removal to sediments through an




apparent settling velocity.  Thomann and DiToro (1983) enlarged the




treatment to include the effects of equilibrium partitioning of plutonium




between water and suspended matter and included an explicit term for the




resuspension of the isotope.  Tracy and Prandtl (1983) reported on new Sr-90




and Cs-137 concentration data for several of the Great lakes and illustrated




the need to consider resuspension in accounting for the long-term behavior




of cesium-137.  Recently, Robbins (1985b) has developed an equivalent




formalism to calculate the response of the lakes to Cs-137 loadings which




includes an improved source function and adds the contributions of the




isotope from land drainage.  This latter component in the model is based on




the semi-empirical relations developed by Menzel (1974) and was calibrated
                                    996

-------
for the  Great Lakes through periodic measurement of the concentration of the




isotope  in tributary rivers (cf.  Nelson et al.1984).  Details of the




calculation are given in Robbins  (1985b).









     The results of the calibrated model calculation are shown in Figure 2.




Contributions from the watershed comprise a small (5X) virtually constant




fraction of the loading.  Direct atmospheric contributions are initially




about 40% of the total load but decrease in recent years to less than 20X.




Inflow from Lake Huron contributes the largest portion of the loading: about




60X during the period of maximum fallout and increasing to more than 80X at




the present time.  This increase reflects the persistence of small amounts




of the isotope in Lake Huron water.  It is fortunate that lake Huron




provides the major input to Lake St. Clair as the most comprehensive set of




measured concentrations exist for this lake (Barry, 1973; Dasgupta, 1975;




Durham and Tamro, 1980; Alberts and Wahlgren, 1981).  As a result, the




estimate of Cs-137 loadings are only weakly dependent on the assumptions of




the long term response model.









     The validity of the loading estimate rests on  several assumptions.




First, it is assumed that measured concentrations (of total Cs-37)




adequately represent the concentration in water exiting from the lake Huron.




Reported values  (especially the pre-1970 data) are  provided by Barry  (1973)




and represent hypolimnetic waters.  The values reported by Alberts and




Wahlgren  (1981) are for the summer when the lake is stratified but no




significant difference exists between epi- and hypoliomnetic water.  This  is




important because concentrations of some elements  (e.g. plutonium) decrease
                                    297

-------
significantly in the epilimnion as stratification develops.  If this were




the case  for Cs-137, surface waters which presumably exit preferentially




from the  lake would contribute less to Lake St. Clair during the summer




months  than predicted by the model.  Additionally, it assumed that the river




system  conveying the radionuclide to the lake has no sinks.  This is




probably  true for the St. Clair river which has a coarse grained, scoured




bottom.   However the delta and marshy areas which receive inflowing waters




could more plausibly store cesium-137.  While we were unable to collect




suitable  cores in these areas there are reasons why the effect should be




small.   Deposits in the delta are coarse, sand-sized materials which




undoubtedly contain negligible amounts of the isotope.  Finer grained




materials which carry the radionuclide accumulate in the open lake beyond




the delta area (Thomas et al., 1975).  The marshy areas may trap fines




efficiently but the proportion of water which flows through them is very




small in comparison with open channel flow.









Vertical Distributions of Cesium-137









     Representative profiles of Cesium-137 are shown in Figure 3 for sites




along a transect between stations 65 and 28 (see Fig. !.)•  Concentrations




are generally low in comparison with surface sediments in the main lakes




where in Lake Huron values are typically around 20 dpm/g and 5-10 dmp/g in




Lake Erie.   In most cases there is a peak in the activity which occurs 2 to




10 cm below the sediment surface.  The extent of penetration of cesium-137




is generally shallow ranging from 2-4 cm toward the ends of the transect and




reaching a maximum  of about 14 cm around the transect midpoint.  The
                                    298

-------
occurrence of subsurface maxima suggests that some information about the




history of deposition may be preserved in these deposits although the




correspondence between loading history and profile shape is generally weak




at best.









     Profiles of cesium-137 at sites 69 and 38 exhibit the best defined




maxima of any from this Lake.  As shown in Figure 4, both cores possess high




values of cesium-137 at about 10 cm with secondary maxima at about 25 cm  and




15 at sites 69 and 38 respectively.  If cesium-137 reached these sites by




direct transfer without sediment mixing or other integrative processes the




expected profile would appear as the dashed line in each case.




Sedimentation rates have been chosen to place the theoretical profile at  the




position of the maximum activity.  Inferred rates are 0.54 cm/yr and 0.43




cm/yr respectively for site 69 and 38.  In both cases the calculated




distribution agrees very poorly with observation.  The deeper maximum is  not




reproduced nor are concentrations predicted successfully elsewhere away  from




the primary peak region.









      These profiles are apparently artifacts  of stratigraphic




 inhomogeneities as can be  seen in the  companion plots in Figure 4.   In  each




 case  the  valley in the Cs-137 distribution  is associated with  (1)  an




 decrease  in  the sediment  solids  content  (FDW-fraction dry weight)  as




 verified  both by field observation and measurement,  (2) an  increase  in  the




 amount of material which  dissolves on  treatment with 10% HCl(FSOL=fraction




 soluble),  (3)  a marked  increase  in the K-40 content over  an interval of a




 few centimeters.  Potassium-40  is  a  direct  measure of the  total potassium
                                    299

-------
content of sediments and indicates the amount of clay sized minerals




present.   These results in combination with field data indicate that such




discontinuities as seen in the Cs-137 profiles probably result from




interlayering of old fine-grained materials of relatively high water




content.   As the two sites in question are relatively near the shipping




channel (Fig. 1.) it is plausible the such layers as well as the material




comprising the secondary peaks could originate from dredging operations.




The only other core showing double Cs-137 peaks is 71 (cf Fig. 5f) which is




also adjacent to the channel.  Adjacent cores 65,68 and 70 do not have




double peaks but the activity of cesium-137 in these cores penetrates only a




few centimeters and may be subject to mixing of near surface sediments which




might obliterate such  structure.









Vertical Distributions of Lead-210









     Distributions  of  excess  lead-210  are shown  for  eleven cores  in Figure  5




 a-f.   Excess lead-210  is computed  by subtracting the activity  of  supported




 lead-210  from each  value.   Supported lead-210  is taken  as  the  average  of the




 lowest values of  lead-210  in  each  core,  1.5 +/-  °-5  dpm/g-   In estimating




 the supported level,  activities  of the isotope within glacial  clay  were  not




 included  inasmuch as  the  supported lead-210  content  of  the clay may not  be




 representative of background levels  in recent sediments.   In these  cores the




 K-40 content is essentially uniform down to  the  gray clay interface




 indicating that distributions of lead-210 as  well as other contaminants  are




 probably not artifacts of grain-size dependent sorting processes.  In every




 case distributions of excess lead-210 are characterized by a zone,  extending
                                     300

-------
down from the sediment water interface to varying depths, possessing a




constant activity.   Below this depth the activity decreases virtually




exponentially.  Exceptions to this pattern occur in cores 38, and 69 which




exhibit disturbances in the vicinity of the stratigraphic anomalies




discussed above.









     The remaining distributions have the form seen elsewhere in the Great




Lakes,  both in the open lake environment (Robbins and Edgington, 1975;




Robbins et al., 1978;  Robbins, 1980) and in the silty, shallower recent




deposits of lower Saginaw Bay.  In previous discussions such profiles have




been generally treated in terms of the rapid steady state mixing model




(RSSM)  in which instantaneous mixing of sediment and lead-210 is postulated




to occur within a discreet zone at the sediment water interface.  Sediments




below this zone are free of mixing or other disturbance and accumulate at a




constant rate.  Details of the mathematical model are provided elsewhere




(Robbins et al.,  1977).  The solid lines shown in Figures 5 a-f are based on




the RSSM model, in which values for the depth of mixing (expressed in g/cm^)




and sedimentation rate (expressed in g/cm^/yr) are chosen to yield the best




weighted least-squares fit.  Inspection of the figures shows that the




agreement between observation and the model is excellent except for the




previously discussed cores 38 and 69.









     Apart from cores 38 and 69 (and 65) mixed depths have a relatively




narrow range, averaging about 3.5 (+/-1.3) cm.  Sedimentation rates are also




narrowly confined,  0.19 (+/-0.06) cm/yr.  If the process is correctly




characterized by the RSSM model then the ratio of the mixed depth( g/cm^) to
                                    301

-------
the  sedimentation rate (g/cm2/yr) is a measure of the time resolution for




reconstructing time histories from the sediment record.   The mean ratio




(excluding 38,65  and 69)  is 19 (+/- 12) yr.









     Since the model results refer to sediment processes and not lead-210




per  se,  a test of the self-consistency of the RSSM model may be developed by




applying it to the Cs-137 profiles using the mixed depth and sedimentation




rate derived from lead-210.  The results of this exercise are shown in the




companion plots in Figures 5 a-f.  Cs-137 distributions are plotted along




with the solid curve which is the unadulterated model fit.  In general (with




the  exception of  38 and 69 as usual) the agreement is good.  Dashed curves




also shown result from allowing the mixed depth and sedimentation rate to




vary so  as to produce an optimized fit.  In general the predicted




distributions are not much better.  The effect of unconstraining the




solution is to increase the model mixed depth so as to fit the region where




the activity of the isotope falls off rapidly.  This occurs with a loss of




structure in the  peak region.  In the case of core 65, considerable mixing




is implied by the Cs-137 data and no mixing by the lead-210 data.  Such an




inconsistency would occur if some recent event removed enough surface




material either by scouring or by disturbance during collection of the core.









     The idea of  surface sediment loss in core 65 is further substantiated




by the total Cs-137 to total excess lead-210 ratio provided for each core in




Table 1.  The three cores (38,69 and 71) with stratigraphic discontinuities




have ratios of 1.0-1.6.  The remaining cores have ratios ranging from 0.6 to




0.8 while core 65 has a ratio of 0.48.  Loss of surface sediment would
                                    302

-------
reduce  total cesium-137 in preference to excess lead-210 and suppress this




ratio.   The  mean ratio of the seven remaining cores is 0.69+/-0.08.









     While  the RSSM model provides a self-consistent account of




sedimentation processes,  it is not necessarily an accurate one.  Several




factors argue against sedimentation as the main control on profile shapes




below the mixed zone.   First, the lake is nondepositional in the sense that




at most 30  cm of material has accumulated in the post-glacial period.  Thus




the long-term net sediment accumulation rate must be considerably less than




the mean measured value of about 0.2 cm/yr.   It could be argued that recent




sedimentation rates are higher than historical values perhaps because of




increased erosion of  watershed soils etc.  However,  horizontal patterns of




deposition might be considered to remain the same even with increased




sedimentation,  in which case RSSM rates ought to at least correlate with the




thickness of recent sediments.  There is no significant correlation (r-0.45,




N-9).  Third,  organisms and their burrows extend to at least 15 cm depth and




their long-term effect should be partial mixing of sediments to considerable




depths.  In  open lake  sediments Robbins (1982) showed that the RSSM mixed




depth correlated well  with the range of occurrence of Oligochaete worms,  but




those deposits  were devoid of organisms such as Hexagenia and the Unionid




clams whose  scale of  sediment mixing is considerably greater.









     If the  profiles  are  chiefly the result  of sediment mixing, the RSSM




results can  be  reinterpreted to yield mixing rate constants.   It is well




known (Robbins,  1978)  that exponential excess lead-210 profiles can also




result  from  uniform eddy  diffusive mixing of sediments.   Robbins used this
                                   303

-------
approach (Robbins et al.  1978) to reconcile distributions of cesium-137,




lead-210 and pollen species in sediment cores from the shallow perturbed




sediments of western Lake Erie.  If the logarithmic decrease in lead-210




activity is  due to the combination of sedimentation and radioactive decay




then the slope of the activity with depth is given by 1/w, where 1 is the




radioactive  decay constant (0.03114 yr'l).  If it is due solely to the




combination  of eddy diffusive mixing and radioactive decay then it is given




by sqrt(l/Kb)  where Kb is the eddy diffusive mixing coefficient in cm2/yr.




Since  the "  best" value of w is already determined, the equivalent value of




Kb is  given  by w2/!.   These values are reported in Table 1 as the deep




sediment mixing model (DSMM) values. According to this view, sediments near




the surface  experience intense mixing characterized by eddy diffusion




coefficients sufficiently large to produce complete homogenization.  Below




the mixed zone,  values decrease rapidly but continuously to values given by




the DSMM model.   Such a depth-dependent eddy diffusion coefficient will also




produce  self-consistent results when applied to vertical distributions of




Cs-137.









Horizontal Distributions  of Cs-137 Based Quantities









    The extent of penetration of CS-137 into sediments is a model




independent  quantity which shows how far particle associated contaminants




introduced in the mid 1960s are found more than 20 years later.   The contour




maps of  the  depth in cm (Fig.  6a) and in g/cm2 (Fig.  6b) show that the




penetration  depth follows the thickness of the recent sediments quite




closely  (r-0.80,  N-31).   This is also true for the relation between the
                                   304

-------
thickness  of recent sediments and the surface (0-1 cm) concentration of Cs-




137  (Fig.  la.,  r-.66,  N-33) and the total accumulation (vertically




integrated)  Cs-137 (Fig 7b., r-0.82, n-35).   The central area in the middle




of Fig.  7b corresponds to the region of the two cores with stratigraphic




anomalies  (38 and 69)  which possess unusually high amounts of Cs-137 (58 and




117  dpm/cm2  respectively).   The total accumulation of Cs-137 in sediments of




the  Lake as  of 1985 is 37 Ci corresponding to an average of 7.4 dpm/cm2 of




sediment.   This may be compared with a total loading (decay corrected to




1985)  of 470 Ci.   Thus, roughly 8X of all the Cs-137 which has passed




through  the  lake remains in sediments today.









    The concentration of Cs-137 in surface (0-2 cm) sediments 9 years




earlier  is shown in Fig. 8a.  These samples were collected by Shipek grab




(Thomas  et al., 1977)  and represent the upper 2 cm of sediment if the sample




is undisturbed.  In order to compare these results with 1985 data the effect




of grab  sampling was simulated by vertically integrating measured




distributions of Cs-137 from 0 to 2 cm.  The resulting contour is shown in




Fig. 8b.  Also simulated was the possible effect of losses of surface




material in  grab sampling.   A 1 or 2 cm loss had only a small effect on the




estimate of  accumulation (20X or less) partly because the higher water




content  of surface materials reduces the amount which those levels




contribute to the total accumulation.   Levels of Cs-137 are considerably




higher in  1976 than in 1985 partly because of radioactive decay which




accounts for 19Z of the difference.
                                    305

-------
     Since  there has been virtually no new input to the system during this 9




year  period,  the ratio of 1985 surface Cs-137 to 1976 values  (decay-




corrected to  1985)  is a direct measure of the removal of the  isotope from




surface  layers .









     The ratio,  contoured in Fig 9, indicates that at selected sites 50Z or




more  of  the activity has gone in 9 years.  Elsewhere within the main body of




the deposits  the ratio is between 50 and 25Z removed while toward the




margins  the ratio approaches 100Z retained.  Sediments at the margins of the




deposit  consist  of coarse material (sand) with occluded fine  grained




sediments containing Cs-137.  Evidently, contaminants that manage to get




incorporated  into coarse sedimentary materials possess a longer residence




time  in  the system.









Sediment Reservoir Residence Times









     The changes in activity of surface materials over time,  as well as the




retention of  the isotope in sediments, are measures of the residence time of




contaminants  in  the sedimentary reservoir.  Since the sediments are




essentially nonaccumulating, the reservoir can be treated as  a system in




which the rate of mass input is exactly balanced by removal.  The flux of




tracer out  of the reservoir F' (dpm/cm^/yr) is given by
                         (F-F')/Tp -IF'                               (1)
                                    306

-------
where  F  is  the flux into the pool and Tp is the pool residence time  (yr)




The amount  of tracer in the pool at any time is given by









                 S(dpm/cm2)-F'*Tp,
and the  concentration of tracer anywhere in the pool (considered to be




uniformly mixed)  is









                 C(dpm/g)-F'/rs,
where  rs  is  the  mass  flux (which equals the resuspension flux in g/cm2/yr)




Under  steady state  conditions which may be approximated for lead-210,




dF'/dt-O  so  that the  amount in the pool is given by
                 S(lead-210)-F'*Tp=F/(l+lp)
where lp-1/Tp.   Since  in a closed system, the amount of lead-210 stored




would be  F/l,  the  fraction of lead-210 retained in a system with a residence



time of Tp  is




                fr-lTP/(l+lTp)
For Cs-137  the  fraction retained is similarly derived but must be evaluated




by numerical  solution of Eq.  1.   The resulting relations between fr and Tp




for both Cs-137  and  lead-210  are shown in Fig. lOa.   For Cs-137, fr, is of




course time -dependent and has been calculated for 1985.   Also shown in this




figure is the value  of fr based on total storage of Cs-137 in the Lake (8%).
                                   307

-------
This value  corresponds to a pool residence time of 6 years.  Lead-210




retention is  also about 8Z, but this value is based on a very limited number




of cores  and  may not be representative.  Additionally the estimate of




loading of  lead-210 to the lake is based on model estimates of the




concentration of the radionuclide in Lake Huron water.  In the calculation




it is  assumed that the Kd of lead-210 in water is 106.  With these




assumptions the pool residence time based on lead-210 is about 3 years.









     It can be seen from the above equations that if the pool contains a




given amount  of contaminant and receives no further inputs, the




concentration of nonradioactive (and nondegrading) contaminant in the pool




will  decrease according to the relation
                                                                      (6)
so that the  pool  residence time may be calculated as
                         Tp-t/ln(c0/c) .                                (7)









Thus the changes  in mean mercury concentrations (Thomas et al., 1977) and




chlorinated organic concentrations (Frank et al.,1977) in the four year




period between 1970 and 1974 lead to additional estimates of the pool




residence times.   These values are given together with those for Cs-137 and




lead-210 in Table 2.   In general values  are very consistent and imply a mean




pool residence time of about 4 +/- 1 year.
                                    308

-------
     The  pool  residence time may also be calculated (from Eq. 1-5) by




comparing changes  in Cs-137 in surface sediments between 1976 and 1985.  The




ratio as  a function of pool residence time, shown in Figure lOb, rises to a




value for very short residence times which reflects loading numbers




integrated over a  very short time span.  The measured mean ratio, 0.6,




predicts  a pool residence time of 16 years as indicated by the dashed line




in the figure.   This value differs appreciably from those calculated above.




Whether the difference is significant or not is uncertain.  However this




ratio reflects changes taking place long after the initial pulse of Cs-137




passed through the lake.   In contrast, the comparison of Cs-137 storage with




the total loading  emphasizes the processes of removal occurring during the




pulse period of the mid 60s.









     Since considerably less Cs-137 was lost from surface sediments between




1976 and 1985  than expected on the basis of a 4 year mean residence time,




perhaps the residence time of the tracer increases with pool contact time.




Certainly this is  consistent with the observation that little or no change




took place in concentrations of the tracer toward the margins of the




deposit.   It also  can be shown that if depth-dependent mixing controls the




downward transport of cesium-137, the residence time will increase with age




because material reaching deep sediments will be kinetically impeded from




returning to the surface where it can be exported through resuspension.




Menzel (1974)  has  described a somewhat analogous situation in which the




residence time (reflecting losses through runoff) of Sr-90 stored on




watersheds increases over the years as the isotope migrates deeper into




layers not accessible to erosion.
                                    309

-------
     A further estimate of pool residence time can be be obtained  from the




ratio of total Cs-137 to total excess lead-210 in individual cores.  This




method does  not rely on horizontal integration methods.  Apart from the few




anomalies discussed above, the ratio shows little variability as seen in




Table 1.   The theoretical ratio, shown in Figure 11 versus pool residence




time, is defined as an envelope to reflect the uncertainty in the  estimate




of lead-210  loading from Lake Huron.  Values of the distribution coefficient




used to calculate the expected concentration in Lake Huron water are assumed




to be comparable to that in Lake Michigan (cf Eadie and Robbins 1987). The




measured ratio (0.69+/--08) predicts a pool residence time of 11 +/- 5




years.  With direct measurements of the lead-210 loading based on  the 210pb




content of Lake Huron water much uncertainty in this estimate would be




removed.









Cs-137 in Sediment Traps









     The activity of Cs-137 in trap samples (Fig. 12) contrasts markedly




between stations.  At station 10 (near coring station 24 in Fig. la.) the




activity is  about 2 dpm/g during the early period and rises to a maximum of




3.7 dpm/g by the end of July.  By October activities are down again




approaching  levels of 2 dpm/g.  In contrast with site 10, little change




occurs in the activity of the radionuclide in trap material at site 13 (near




core station 28).  There is no peak and activities decline steadily from




1.45 dpm/g in early June to 0.8 dpm/g by mid-September, thereafter rising




toward the previous early June value.  In contrast to the behavior of Cs-
                                    310

-------
137,  K-40  is  essentially invariant with time of year and with location.




Values  shown  in Table 3 indicate no significant differences in K-40 content.









     Without  additional study the cause for these variations remains




unclear.   However it is likely that sediment traps collect two distinct




populations of particles: those resuspended from sediments and those




entering the  lake for the first time primarily by inflow from Lake Huron.




The Cs-137 activity on particles in Lake Huron itself may be inferred  from




measured total concentrations of the isotope in water as of 1985  (0.06




dpm/L,  Robbins, 1985b), the well-known partition coefficient for  Cs-137  in




open waters of the Great Lakes (3xl05 ml/g, Alberts et al. 1981)  and the




total suspended matter in the lake (TSM-0.5 ppm).  These values combine  to




give a Cs-137 concentration on particles of about 20 dpm/g.  This value




agrees well with concentrations of the isotope in surface sediments of the




lake (Robbins et al., 1977 and Robbins 1980).  Because TSM probably




increases and water composition is altered in the St. Clair River,




partitioning  of Cs-137 between particle and solution phases could change.




However it is likely that the activity of cesium-137 on particles entering




Lake St. Clair from the river greatly exceeds (by about a order of




magnitude) activities of the isotope on particles resuspended from the




bottom of the lake.  Such a large difference can be exploited to  infer the




admixture of  particles originating from the two sources which end up in




sediment traps.  The data are consistent with this interpretation.









     During the summer months when the lake is less mixed physically,




resuspension plays a smaller role in loading traps, especially those located
                                    311

-------
along  the  direct course of water flow between the St.  Clair and Detroit




River. With  less efficient horizontal averaging during the summer period a




site such  as 10 which is closer to the streamlines will record a higher




proportion of particles from river inflow.  A site such as 13 which is




further  from the streamlines will be less sensitive to river contributions




and collect  a greater proportion of resuspended materials.  This idea is




also supported by a comparison of the Cs-137 concentration of the trap




samples  with concentrations of the radionuclide in surface sediments




underlying the traps.  Site 10 is situated in the middle of a broad region




where  the  surface activity (0-1) cm is 2.0 dpm/g.  Thus, in early June and




again  after  mid-September, levels in traps at this site exactly match




surface  sediment concentrations.  In contrast Site 13 is situated over




sediments  with an activity of 0.2 dpm/g.  During the early part of the




observation  period, values are close to those at Site 10 and in fact reflect




the mean concentration of Cs-137 in surface sediments over a wide area.  As




the season progresses values decrease, presumably as a result of a decrease




in  the extent of horizontal mixing, and approach a value of 0.8. However the




decrease in  activity does not reach the value of 0.2 which would represent




concentrations directly under the trap.  This occurs presumably because at




least some material is derived from inflow and because even under relatively




quiescent conditions, regions of the lake bottom with higher surface Cs-137




concentrations contribute to the trap as well.









     While the results of Cs-137 measurements in trap materials from this




lake are only preliminary, it is clear that the radionuclide will be of




considerable use in characterizing the seasonal and perhaps shorter term
                                    312

-------
processes  of  transport of contaminants through the system.  Of particular




interest  is the  use  of the isotope to sort out the relative contributions




that new  (inflow)  and old (resuspension) sources of particle-associated




contaminants  make  to concentrations measured in water at various places  in




the Lake.









Chlorinated Organic  Compounds in Surficial Sediments









     Typical  contaminant distribution maps for Lake St. Clair surficial




sediments (0-1 cm) are shown in Figure 13.  The highest contaminant




concentrations are found near the center of the lake in the region of




greatest  water depth.  The sediment cores show the greatest accumulation of




fine-grained  sediments and the thickest layer of recent sediments over




glacial clay  in this vicinity as well.  Some minor contaminated sediment




accumulation  also  occurs in Anchor Bay at the northern end of the lake.  For




the most part the  sediments in the rest of the lake are sandy and, like  Cs-




137, contain  very  low concentrations of organic contaminants.  Table 4 shows




the range and mean values for several halogenated organic compounds.  These




concentrations are in good agreement with other previously published




surficial sediment data for the Lake (Frank et al., 1977; Pugsley et al.,




1985; Oliver  and Bourbonniere, 1985).  Although the mean contaminant




concentrations,  with the possible exception of HCB, are not particularly




high compared to other areas in the Great Lakes Basin, the maximum




concentrations reach significant levels for many of the Sarnia-source




contaminants.
                                    313

-------
Trends  from  Sediment Cores









     Fig.  14 shows the concentration of HCB (hexachlorobenzene), DCS (octa-




chlorostyrene),  PCBs (polychlorinated biphenyls),  HCB (Hexachlorobutadiene),




QCB (pentachlorobenzene) and DDT (total DDT-sum of DDT and degradation




products)  in sediment cores from the lake.  The absolute concentrations of




the contaminants in the cores varies considerably with location.  For




example the  maximum HCB concentration is 180 ppb in core 38 and only 20 ppb




in core 84.  Two  of the shallower cores, 64 and 71 (ca 10 cm mud over glacial




clay) had similar contaminant profiles. Chemical concentrations were fairly




constant in  the  upper 6-7 cm for most compounds, presumably reflecting the




mixing  processes evidenced in the radionuclide profiles.  Below the constant




concentration zone, levels gradually decrease and approach near zero




(undetectable levels) at the silt/clay interface.   Core 84 exhibited




peculiar behavior with several high concentration spikes for various




compounds appearing deep in the core.  This unusual behavior might be due to




translocation of active surface sediments to discreet depths through the




reworking behavior and burrow infilling by Hexagenia larvae found in this




and other cores.









     Cores at sites 38 and 69, possessing the greatest thickness of recent




sediments, provide some consistent information on the loading history of




certain compounds.  The loading pattern of DDT to the Great Lakes is well-




known (Rappaport et. al., 1985) and is closely related to U.S. usage data.




DDT production began in 1944, peaked in 1959 and stopped in 1972.  For core




38, a major subsurface maximum for total DDT is observed at 13 cm and the
                                    314

-------
peak profile  crudely follows production history.  However an anomalous early




peak at  19  cm for total DDT and for some of the other chemicals may indicate




that some disturbance in the sedimentation process has occurred.  The time




of this  disturbance (likely due to dredging in the area) seems to be before




the peak DDT  input to the lake or of the order of 25-30 years prior to 1985.




Despite  this  problem, the top 6 cm of the core still seems to provide a




useful indication of contamination trends in the lake.  The PCS maximum  in




this core occurs at a depth of about 8 cm.  Peak PCB production in the




United States occurred in 1970 (Peakall, 1975).  Contamination history can




be roughly  inferred by using total DDT and PCB peaks as markers, 1959 (13




cm) and 1970  (8 cm) respectively, assuming production/usage scenarios can be




used to infer loadings to Lake St. Clair.  OCS in core 38 peaks at 8 cm  and




shows  a significant decline in recent years.  In contrast, HCB, HCBD and QCB




exhibit a  fairly uniform concentration in the upper 6-8 cm of this core.









     In core  69, very high concentrations of many contaminants are found




below 18 cm.   This appears to be due to site disturbance at a time when  DDT




input to the  lake was high (probably around 1960).  Normal contaminant




distributions are found above 16 cm.  Again, PCB and OCS peaks occur at




about 8 cm depth with dramatic decreases in concentration towards the core




surface.  HCBD and QCB concentrations are high and relatively constant in




the 0-8 cm range in the core.  HCB concentrations seem to increase steadily




toward the core surface with the highest concentration at the surface.









      Both DDT and PCBs entered the lake mainly  from diffuse non-point




 sources including the tributaries.  The other four chemicals originate
                                    315

-------
mainly  from  industrial activity in Sarnia.  Like DDT and PCBs, OCS and QCB




exhibit higher  concentrations deeper in the sediment.  Reduced surface




concentrations  apparently reflect the decrease in loading of the chemicals




which has  likely occurred in recent years.  Both HCB and HCBD concentrations




are increasing  or are staying fairly constant near the top of the cores.




Loadings of  these chemicals to the lake are evidently not dropping




significantly and may even be increasing.









     Studies of sediments from the St. Clair River have shown that the




ratios  of  HCB to OCS and QCB are useful in tracking the source of




contaminants in the river (Oliver and Bourbonniere, 1985.  The HCB/OCS and




HCB/QCB ratios  are 1.3 and 4.0, respectively, for sediments near the  Scott




Road Landfill,  a site which contains waste byproducts from Dow's early




production of chlorine and chlorinated solvents.  The HCB/OCS and HCB/QCB




ratios just below Dow's outfall and where non-aqueous wastes have leaked




into the river  are 16 and 23 respectively.  In 1980 Dow began treating the




Scott Road leachate by carbon filtration  to reduce loadings from the  site.




In cores 38 and 69 the HCB/OCS ratio changes from  2 lower in the core to 9




at the surface.  Similarly the HCB/QCB ratio increased from 4 near the




bottom to 20 near the surface.  Thus, the trends in the Lake St. Clair cores




 are  consistent with a diminishing contribution from the Scott Road Landfill




 and  contributions from DOW's outfall which have  increased since the




 beginning of the decade.









     A quantitative treatment  of  these  trends requires further measurement




 of upstream loading and development  of better models  for characterizing  the
                                    316

-------
nature  of  transport processes in these sediments.  While vertical transport




is probably  dominated by a combination of physical and biological mixing




with little  net sedimentation, changes in the loading to sediments can




nevertheless leave behind a record which to some extent corresponds to the




loading history as discussed above for the radionuclides and organics.




Interpretation of sedimentary profiles in this energetic, shallow water




environment  must be approached with caution because of important processes




which can  affect and perhaps override that of continuous mixing.  These




include (1)  intense resuspension which operates with varying intensity over




the lake bottom, (2) discontinuous (event-related) processes of




sedimentation and resuspension, (3) spatial heterogeneity in the biological




and physical mixing process,  (4) horizontal reintegration of sediment loads,




(5) particle sorting effects, (6) and the effects of human activities such




as fishing,  anchoring, boating, shipping operations and dredging.  Sediment




core profiles are the cumulative end result of these many potential effects




and simple models, even self consistent ones, may well be inadequate.









Contaminant Storage and Loading









     Depth-integrated samples (interval composites) were prepared and




analyzed for each core to estimate the mass of contaminants stored in the




sediments.  Horizontal distributions in total storage have patterns which




are essentially congruent with the thickness of  recent sediments and form




 the basis for estimating total storage in the Lake  (contour integration).




 For the sandy non-accumulating areas, where cores were not collected, a




value  of 5 ng/cm^ was used  for PCBs and HCB and  a value of 0.5 ng/cm^ was
                                    317

-------
used for OCS.  These  approximations are not critical since the sandy areas




contributed  less  than 5X  of the contaminant mass for these chemicals.  Lake




St.  Clair  sediments presently contain about 960 Kg of HCB, 870 Kg of PCBs




and 210 Kg of  OCS.









     These values are much higher than the contaminant masses found by




Oliver and Pugsley (1986) for the St. Clair River sediments  (69 Kg HCB, lOKg




OCS) indicating  that  Lake St. Clair is a more significant repository for




chemicals  than the river itself.  Recent loading estimates for HCB and OCS




in the combined  dissolved and particulate fraction at Port Lambton in the




St. Clair  River  were  180 Kg/yr for HCB and 11 Kg/yr for OCS  (Chan et al.




1986).  At these rates Lake St. Clair sediments contain about a 5 year




supply of  HCB  and a 20 year supply of OCS.  Thus, the sediments retain




significant  fractions of these chemicals and, given the uncertainties in the




calculation, accumulation is consistent with sediment reservoir residence




times derived from historical studies of metal and organic chemicals in the




system and from  the response of sediments to particle-associated




radionuclides.









      A crude estimate of the total historical loading of  particle-associated




 organics can be  calculated  if  it is  assumed that their behavior  is similar




 to that of 13?cs.  The following values were estimated by multiplying  the




 average ratio of organics  (ng/cm2) to  137cs  (dpm/cm^) for cores  64,  71,  84,




 38 and 69, by the total mass of l^^Cs  which was loaded  into  the  system  (470




 ci converted  to dpm): HCB  15 metric  tons  (MT); QCB  2.8 MT; HCBD  3.3  MT;  OCS




 4.3 MT; PCBS  20 MT;  and  total  DDT  5.4  MT.  These calculations show that
                                    318

-------
considerable  quantities  of chemicals  have been discharged into the system.




The total masses  are  probably on the  low side for constituents which possess




a greater solubility  or  have a longer history of loading than radiocesium.
                             ACKNOWLEDGEMENTS









     The authors  wish to acknowledge the assistance of K. Hill and Technical




Operations  staff  of The Canada Centre for Inland Waters for help in the




collection  of cores and to Mrs.  Alena Mudroch for assistance in the field




sectioning  of sediment cores.  Thanks are due M. Charlton for the loan of




trap sample material for cesium-137 analysis.  The help of R. Rossmann and




E. Meriweather in carrying out laboratory analysis of sediments for lead-210




and of K. Nichol  for assistance in analysis of organic constituents is




gratefully  acknowledged.
                                    319

-------
                              LITERATURE CITED









Alberts, J. J. and M.  W.  Wahlgren.  1981.  Concentrations of 239/240^f  137CS|




     and 90sr  in  waters of the Laurentian Great Lakes:  comparison of 1973




     and 1976  values.  Environ. Sci.  Technol.  15: 94-98.









Anon.  1986.  St. Clair  River Pollution Investigation,  Environment Canada and




     Ontario Ministry  of the Environment









Anon. 1986.  St. Clair  River Pollution, Water Pollution  Res. J. Can. (J.




     Lawrence, Ed.)  21: 283-459.









Barry, P.  J.  1973.  Estimating dose commitments to populations from




     radioactive  waste disposals to large lakes. In:  Environmental Behavior




     of Radionuclides Released in the Nuclear Industry. International Atomic




     Energy Agency,  Vienna, pp. 499-505.









 Chan, C. H., Y L. Lau, and B. G. Oliver. 1986. Measured and modelled




     chlorinated contaminant  distributions in St. Clair River water.  Water




     Poll. Res. J.  Can.  21: 332-343.









 Charlton, M. N. and B. G. Oliver. 1986.  Chlorinated organic contaminants  on




      suspended sediment  in Lake St. Clair. Water Poll. Res. J.  Can. 21:




      380-389.
                                    320

-------
Dasgupta, A. K. 1975. Radioactivity in Lake  Huron water,  summary of 1963 to




     1973 data, and  radioactivity in Lake Ontario water  1971-1973.  In:




     Radio-activity  in  the  Great Lakes, a summary of radionuclide monitoring




     and nuclear  plant  discharge data available Jan.  1975.  Prepared by the




     Radio-activity  work group of the International Joint Commission, Great




     Lakes  Water  Quality Board by M. S. Olijnyk and R. W.Durham.









Durham,  R.  W.  and G.H.  Jamro.  1981. Great Lakes radiological surveillance




     1980.   Environment Canada, National Water Research  Institute,




     Burlington,  Ontario, Canada. July, 1981. 6 pp.









Eadie, B.  J. and  J.  A.  Robbins. 1987. The role of paryiculate matter in the




     movement of  contaminants in the Great Lakes. In Sources and Fates of




     Aquatic Pollutants  (R. A. Hites and S. J. Eisenreich, Eds.) Advances  in




     Chemistry Series 216, American Chemical Society, Washington, D. C., pp.




     319-364.









 Frank,  R. , M. Holdrinet, H. E. Braun,  R. L. Thomas, A. L. W. Kemp AND J.-M.




     Jaquet. 1977. Organochlorine  insecticides and PCBs  in sediments of Lake




      St. Clair (1970 and 1974) and Lake  Erie  (1971). Sci. Tot.  Environ.




      8:205-227.









 Lerman, A.  and  H. Taniguchi.  1972,  Strontium-90 in  the  Great Lakes:




      concentration- time model.  J.  Geophys  Res.  77:  3256-3264.
                                    321

-------
Menzel, R. G.  1974.  Land surface erosion and rainfall as sources of




     strontium-90  in streams.  J. Environ. Qual.  3:  219-223.









Nelson, D. M., D.  N. Metta and J. 0. Kartunnnen. 1984. 239/240pu> 137cs, and




     90 Sr  in Tributaries of Lake Michigan. Environmental Research Division




     Annual  Report,  Jan.-Dec.  1983, Argonne National Laboratory, Argonne,




     II.,  Dec. 1984, pp.45-54.









Oliver,  B.  G. and R. A. Bourbonniere. 1985. Chlorinated contaminants in




     surficial sediments of Lakes Huron, St. Clair and Erie: implications




     regarding sources  along  the St. Clair and  Detroit Rivers. J. Great




     Lakes Res.  11: 366-372.









 Oliver, B. G. and C. W. Puglsey. 1986.  Chlorinated Contaminants  in St.  Clair




     River sediments.  Water Poll.  Res.  J.  Can.  21: 368-379.









 Peakall, D.  B. 1975. PCBs  and their environmental effects.  CRC  Crit. Rev.




     Envrion. Cont.  5:  469-508.









 Pugsley, C.  W., P.  D.  N.  Hebert, G. W.  Wood,  G. Brotea AND  T. W. Obal.  1985.




     Distribution of contaminants in clams and sediments from Huron-Erie




      corridor.  I-PCBs and octachlorostyrene.  J. Great Lakes Res. 11:




      275-289.








 Rappaport,  R. A., N.  R. Urban, P. D. Capel, J. E.  Baker, B. B.  Looney,  S. J.




      Eisenreich AND E. Gorham. 1985. Chemosphere 14; 1167-
                                     322

-------
Robbins, J. A.  1986.  Sediments of Saginaw Bay,  Lake Huron:  Elemental




     Composition and accumulation rates.  Special Report 102, Great Lakes




     Research Division,  University of Michigan, Ann Arbor,  Michigan. 102 pp.




     (Appendix,  278 pp.)









Robbins, J. A.  1985a. Great Lakes regional fallout source functions, Great




     Lakes Environmental Research Laboratory Technical Memo, ERL-GLERL-56,




     Feb.  1985,  22 pp.









Robbins, J. A.  1985b. The coupled lakes model for estimating the long-term




     response of the Great Lakes to time-dependent loadings of particle




     associated contaminants. Great Lakes Environmental Research Laboratory




     Technical Memo, ERL GLERL-57, Apr. 1985,41 pp.









Robbins,  J.  A.  1982. Stratigraphic and dynamic effects of sediment reworking




     by Great Lakes zoobenthos.  In:  Developments in Hydrobiology 9,




     Sediment/Freshwater Interaction, Ed. P. G. Sly.  Proc. of the 2nd




     International Symposium on Sediment-Water Interactions, Kingston, Ont.




     Hydrobiologia 92 (1982) 611-622.









 Robbins,  J. A. 1980. Sediments of southern Lake Huron: elemental composition




     and accumulation rates.  U.S. Environmental Protection Agency,




     Ecological Research Series, EPA-600/3-80-080. August,  1980. 309 pp




     (Appendix, 198 pp.)
                                    323

-------
Robbins, J.  A.  1978.  Geochemical and geophysical applications of radioactive




     lead.  In:  Biogeochemistry of lead in the environment,  Part A.  (J.  0.




     Nriagu,  Ed.),  Elsevier Scientific Publishers,  Amsterdam, Netherlands,




     Vol.  1A:  285-303.









Robbins, J.  A.  and D. N.  Edgington.  1975. Determination of recent




     sedimentation rates  in LAke Michigan using Pb-210 and Cs-137.  Geochim.




     Cosmochim. Acta. 39: 285-304.









Robbins, J.  A., D.  N. Edgington AND A. L. W. Kemp.  1978. Comparative lead-




     210,  Cs-137,  and pollen geochronologies of sediments from LAkes Ontario




     and Erie.  Quatern. Res. 10: 256-278.









Robbins,  J.  A., J.  R. Krezoski and S. C. Mozley. 1977. Radioactivity in




     sediments of the Great Lakes: postdepositional redistribution by




     depositfeeding organisms. Earth Planet. Sci. Lett. 36: 325-333.









Thomann,  R.  V.  and D. M.  DiToro. 1983. Physico-chemical model of toxic




     substances in the Great Lakes.  J. Great Lakes Res. 9: 474-496.









Thomas, R. L. ,  J.-M.  Jaquet, and A. Mudroch. 1977.  Sedimentation processes




     and associated changes in surface sediment trace metal concentrations




     in Lake St. Clair,  1970-1974, Proc. Int. Conf. on Heavy Metals in the




     Environment,  Toronto,  1975, pp.  691-708.
                                    324

-------
Tracy, B. L. and  F. A.  Prantl.  1983.  Twenty-five years of fission product




     input  to  Lakes Superior and Huron.  Water,  Air and Soil Poll. 19:  15-27









Wahlgren, M. A.,  J. A.  Robbins  and D  N.  Edgington.  1980.  Plutonium in the




     Great  Lakes.  In  Transuranics in  the Environment^ W.  C. Hanson (Ed.),




     Report No. DOE/TIC-22800,  Technical Information Center, U.  S.  Dept.  of




     Energy, Washington D.C.  pp 659-683.
                                   325

-------
Table !•  Summary  of  lead-210 and cesium-137  mixing and sedimentation model  results.
RSSM Mixed depth
2
ite (g/cm ) (cm)
17 1.7 2.8
18 4.3 4.8
20 3.0 3.0
21 5.1 5.1
23 4.1 5.0
24 2.0 2.8
38. 10. 9.
64 2.8 3.1
65 0.2 0.2
69 ? ?
71 1.2 1.2
RSSM Sedimentation
2 Rate
(g/cm /yr) (cm/yr)
0.29
0.19
0.14
0.12
0.14
0.20
0.16
0.21
0.29
1.6
0.14
0.34
0.18
0.10
0.11
0.13
0.20
0.14
0.21
0.20
1.3
0.09
RSSM Time DMM
Resolution Rate
(yr) (cm /yr)
6.
22.
21.
44.
30.
10.
60.
13.
<1
?
9.
3.7
1.0
0.4
0.4
0.57
1.2
0.65
1.4
1.3
54.
0.23
Total
(dpm/cm )
29.6
27.3
22.7
30.7
34.3
36.5
49.5
36.7
18.7
75.6
12.1
Total Total 
-------
Table 2.  Sediment reservoir residence times inferred
from radionuclide storage as of 1985 and changes in mean
contaminant levels from 1970-1974
   Constituent      Residence time( yr)       Reference
Cesium-137
Excess lead- 210
Mercury
DDE
IDE
DDT
Total PCBs
6.0
3.0
4.0
3.6
4.6
2.9
6.2
This work
ibid.
Thomas et al.(1974)
Frank et al . (1977)
ibid.
ibid.
ibid.
   Mean
                           327

-------
Table -3  Cs-137 and K-40 in trap samples  collected from Lake
St. Clair in 1985*.
Location-
Sample number
10-401
10-402
10-403
10-404
10-405
10-406B
10-408
10-410
Collection
Period
5/30-6/5
6/5-6/26
6/26-7/10
7/10-7/31
7/31-8/21
9/13-9/25
9/25-10/17
10/17-11/17
Cesium-137
(dpm/g)
1.93 ± 0.20
1.90 ± 0.15
2.73 ± 0.30
3.00 ± 0.30
3.70 ± 0.50
2.71 ± 0.34
2.10 ± 0.15
2.19 ± 0.13
Potassium-40
(dpm/g)
30 ± 4
34 ± 3
34 ± 10
22 ± 8
26 ± 10
57 ± 10
27 ± 3
37 ± 3
   Mean                        2.53 ± 0.62         32 ± 9
13-401
13-402
13-403
13-404
13-405
13-406
13-406B
13-408
13-410
5/30-6/5
6/5-6/26
6/26-7/10
7/10-7/31
7/31-8/21
8/21-9/13
9/13-9/25
9/25-10/17
10/17-11/17
1.45 ± 0.17
1.17 ± 0.06
1.33 ± 0.30
1.19 ± 0.16
1.01 ± 0.18
1.18 ± 0.19
0.80 ± 0.10
0.94 ± 0.10
1.33 ± 0.11
34 ± 4
33 ± 2
30 ± 9
33 ± 4
33 ± 6
33 ± 5
32 ± 3
30 ± 2
35 ± 3
   Mean                       1.16 ± 0.21         33 ± 2
* Trap collections by Murray Charleton,  Canada Centre for Inland
  Waters.
                                   328

-------
Table 4. Chlorinated organic compounds in surficial (0-1 cm)
sediments of Lake St. Clair in 1985 (ng/g).
      Compound                      Range           Mean


Hexachlorobenzene  (HCB)            0.4-170         32

Octachlorostyrene  (OCS)            n.d. - 21          4.8

PCBs                               n.d. - 60         19

Hexachlorobutadiene (HCBD)         n.d. - 32          5.4

Pentachlororbenzene (QCB)         n.d. - 8.7         3.2

Total Trichlorobenzene (TCB)       n.d. - 28          4.3

Total Tetrachlorobenzene (TeCB)    n.d. - 20          3.7

Total DDT and metabolites (SDDT)   n.d. - 12          3.8
                            329

-------
                              LIST OF FIGURES









Fig. 1.  (a) Locations of the 1985 coring sites in  Lake St. Glair. Stations




         correspond to those occupied previously in the U. Windsor study




         (Pugsley et al., 1985).




         (b) Thickness (cm) of recent sediments.  Net post-glacial sediment




         accumulation is exceedingly small.









Fig. 2.  Loadings of Cs-137 to Lake St. Clair from the three principal




         sources:  inflow from Lake Huron, direct fallout and transfer from




         the drainage basin. Loading is dominated by inflow from Lake Huron




         which is a persistent low-level source.









Fig. 3.  Vertical distributions of cesium-137 along an east-west transect




         (Fig. 1).  Activities are generally very low compared with open lake




         depositional sites and distributions correspond only weakly to




         loading history.









Fig. 4.  (a) Cs-137 distributions in cores with distinctive stratigraphic




         anomalies. The dashed line is the profile expected from direct




         transfer of cesium to undisturbed sediments with peak positions




         arbitrarily set.




         (b) Distributions of quantities (see text) indicating bulk sediment




         composition. A layer of old clay (dashed line) produces a valley in




         the distribution of cesium-137.
                                 330

-------
Fig. 5.  (a-f). Distributions of excess lead-210 (top, log scale) and cesium-




         137 (bottom) for 11 cores. Distributions of lead-210 are usually very




         well characterized by the RSSM model (solid lines) showing intense




         near surface mixing (ca 3 cm) and either sedimentation or uniform




         mixing below.  Distributions of Cs-137 are generally predicted well




         by the RSSM model using mixing and sedimentation parameters set by




         lead-210 (solid lines).  When mixing and sedimentation parameters are




         allowed to vary for Cs-137 the dashed line results.  The anomalous




         cores 38 and 69 give poor fits.









Fig. 6.  Extent of penetration of Cs-137 (a, cm; b, g/cm2) into sediments.




         Associated with fine-grained materials, Cs-137 has reached a maximum




         depth of about 20 cm in roughly 25 years.  The penetration depth




         correlates with the thickness of recent sediments and bathymmetry.









Fig. 7.  (a) Concentration of cesium-137 in surface sediments (0-1 cm) and (b)




         total accumulation (dpm/cm**2).  Anomalously high total Cs-137




         (shaded region with >50 dpm/cm2 Cs-137) occurs in cores with marked




         stratigraphic inhomogenieties.









Fig. 8.  (a) Concentration of Cs-137 in surface sediments (0-2 cm) collected




         by Shipex grab sampling in 1976. (b) Concentration in the 0-2 cm




         interval calculated from profiles in cores collected in 1985. Only




         19% of the decrease can be attributed to radioactive decay.
                                    331

-------
Fig. 9.  Ratio of Cs-137 activity in surface sediments (1985) to the activity




         in 1976, decay-corrected to 1985. The ratio approaches unity toward




         deposit margins where the net loss of the radionuclide is thus least.









Fig. 10. First order loss rate model calculations for a sediment contaminant




         pool.   (a) Fraction of Cs-137 and excess lead-210 retained in the




         pool vs pool residence time. For both lead-210 and cesium-137 the




         fraction retained is about 8% indicating pool residence times of 6




         and 3 years respectively. Uncertainties in total excess lead




         inventories are large.  (b) Ratio of 1976 to 1985 concentrations of




         Cs-137  in surface sediments vs pool residence time. The measured




         value of 0.60 predicts a residence time of 16 years.









Fig. 11. Ratio of total Cs-137 (1985) to total excess lead-210 vs pool




         residence time.  The envelope reflects uncertainties in the value of




         the partition coefficient for lead-210 used to calculate loadings of




         the radionuclide from Lake Huron. The measured mean ratio of 0.69




         predicts a pool residence time of 12 years.









Fig. 12. Seasonal variation in the concentration of Cs-137 in trap materials




         collected at two sites in Lake St. Clair. Differences between sites




         are  ascribed to varying proportions of Cs-137 labeled particles from




         two  principal  sources: resuspension of low activity sediments already




         in the  Lake and inflow of new, high activity material from Lake




         Huron.
                                     332

-------
Fig. 13. Distribution of chlorinated organic compounds in selected cores.




         Significant concentrations of contaminants are confined to about the




         upper 10 cm.









Fig. 14. Distribution of chlorinated organic compounds in cores for which




         companion cores possess stratigraphic inhomogenieties.









Fig. 15. Concentrations of chlorinated organic compounds in surface (0-1 cm)




         sediments. Concentrations are well correlated with the thickness of




         recent sediments and associated with fine grained materials traced by




         Cs-137.
                                   333

-------
       _AKE  ST. CLAIR
Sediment Coring Sites (1985)
            o \ -e

-------
          LAKE ST. CLAIR
                            82°30'
Thickness of Recent Sediments (cm)

               335

-------
            Cs Loading to Lake St. Clair
                Watershed
                Watershed + Atmosphere
                Watershed +Atmosphere
                +Inflow from Lake Huron
1960          1970
       Year
1980


-------
                      137.
       Distribution of   Cs on an East/West Transect
                 in Lake St. Clair (1985)
    4  8  12  0
4   8  12   0  4   8  12   0
      DEPTH (CM)
             4   8   12
     LSC-85 25
 LSC-85 26
LSC-85 27
LSC-85 28
0   4  8   12  0   4  8   12   0  4   8  12   0  4   8  12
                        DEPTH (CM)
                     pi c\o^e-   t) .

-------
u>
U>
00
'0    10   20    30

     Depth (cm)
                                                   '0      10      20
                                                       Depth (cm)
                                                                          Q.
                                        \c\o^e,

-------
    6

 _ 5



 t 4
 Q.

 3 3


 J


2°- z
CVJ
^  3


^


 Q.

3  2
               LSC-85 171
               LSC-85 171
   6


   5



   4




   3





   2
j LSC-85 181
                  10
15   0
                      Depth (cm)
              jLSC-85 181
    10
                                                  15
                             5".
                         339

-------
Depth (cm)
 340

-------
0
10     15 W0     5


   Depth (cm)
                                 10
15
                        r
                      o C

-------
                    20
Depth (cm)
  342

-------
5-
               dLSC-8565
O.
-o
.a
Q.
   0.5
               fLSC-85 65
 a.
 (O
                         2



                         I



                       0.5




                       Q2


                         8
                                *•  J LSC-86 69
                    10
                      15  0
20    30
                       Depth (cm)
                 t~ i 'v, o -f e.
                        343

-------
      Depth (cm)
r t c\ o
         344

-------
LAKE ST. CLAIR
LAKE ST. CLAIR
                                                                     ,
                    '   1 4 Ov/e    6  <*., b

-------
           LAKE ST. CLAIR
Cs Surface (0-1 cm) Activity in I985(dpm/g)
            \^o<«. 1
                  346

-------
           _AKE ST. CLAIR
Total   Cs Storage as of 1985 (dpm/cm )

             < a -o r
               .

-------
          LAKE  ST. CLAIR
l37Cs in Surface Grab Samples (0-2cm)
         as of 1976 (dpm/g)
          r i
                348

-------
          _AKE  ST. CLAIR
l37Cs in Surface Grab Samples (0-2cm)
        as of 1985 (dpm/g)
          •: \ Q v
                349

-------
           _AKE  ST. CLAIR
137
  Cs Surface Activity Ratio: 1985/1976
                      -
                .
                 350

-------
   0.6
2 0.4
IT
c
o
I 0.2
III

   0.0
 o 0.6
   0.2
 o
O
   0.0
rCs
      0       10     20      30
          Pool Residence Time (yr)
                        0
                  351

-------
O .0
        Pool Residence Time
Uul
                                  a;
          F
                  352

-------
      137
        Cs in Lake St. Clair Trap Samples (1985)
_  3
 a.
                              SITE 10
                SITE 13
      May  Jun.   Jul.   Aug.  Sept.  Oct.   Nov.
                        Month
                * -
                i-
                      353

-------
LAKE ST. CLAIR      _  HCBD
LAKE ST. CLAIR          HCB
LAKE ST. CLAIR
                                         LAKE ST. CLAIR          ZDDT
LAKE ST. CLAIR
                                           ,   /3

-------
       Core 64
     Core 71
                    Core 84
c
o

o

"c

o
c
o
o
             10    '0   5

                Depth in Core (cm)
n
<\
                            '  »
                            355

-------
            Core 38
Core 69
g

"o


o>
o
c.
o
o
                       TOTAL
                       DDT
                     Depth in Core (cm)
                                   15"
                               356

-------
     TOXICOKINETICS  OF  SELECTED XENOBIOTICS IN HEXAGENIA LIMBATA:




                LABORATORY STUDIES AND  SIMULATION MODEL




                   Peter F. Landrum  and Ronald Poore
                                ABSTRACT
    Understanding the role of benthos  in  the  fate and transport of




toxicants requires understanding  the  toxicokinetics of those organisms for




both water borne and sediment associated compounds.  This effort focused on




the toxicokinetics of Hexaeenia  limbata as an important component of the




benthic  community of the  upper Great  Lakes connecting channels.  The




accumulation and loss of  selected polycyclic aromatic hydrocarbon congeners




and a hexachlorobiphenyl  congener were followed over the course of a season




inH. limbata collected from Lake St. Clair.  Both the uptake  and




elimination rate  constants increased with increasing temperature through  the




spring and summer.   The elimination constant was  relatively large  compared




 to other Great Lakes benthos.  The uptake constant for sediment associated




 compounds was essentially constant the  two  times  it was  measured.  The




 limits  for the parameters to a deterministic  simulation  model  were derived




 from the laboratory kinetics constants.   The  model suggests that toxicant




 concentrations in the organisms  should decline  as the temperature  increases,




 as a result  of a greater increase in the  elimination constant, and rise




 again as the temperature declines.   Based on the best estimates of




 environmental concentration of  the toxicants studied in both sediment and
                                     357

-------
water, the model suggests  that H.  limbata should obtain greater than 90% of




its contaminant body burden from the sediment associated pollutants.
                               INTRODUCTION
     Hexagenia limbata are an important food source  for  fish in the




 interconnecting waterways of the Great Lakes (Hunt,  1958).   Further,




 g. limbata are extremely sensitive  to pollution  resulting in their




 disappearance from some locations  in bays  and the  connecting channels (Carr




 and Hiltunen, 1965, Howmiller and  Beeton,  1971,  Schneider et al. , 1969,




 Hiltunen and Schlosser, 1983).  While  the  disappearance has been attributed




 to low dissolved oxygen as  a  result of eutrophication, oil inputs into the




 St. Mary's River below Sault  Ste.  Marie,  MI, have eliminated the mayfly




 larvae from  stretches of  the  river that they previously inhabited (Hiltunen




 and Schlosser,  1983). The mayfly larvae are beginning to repopulate areas




 of the Great Lakes (Thornley, 1985).  Since H.  limbata lives and feeds in




 the sediment,  the mayfly larvae may mobilize contaminants  from the sediments




 up through the food chain.  Because of their sensitivity  to pollution and




  their importance in the food web,  I undertook  to  define  the toxicokinetics




  of these organisms for selected polycyclic aromatic hydrocarbons (PAH) and




  polychlorinated biphenyl congeners (PCB) .  The studies  were run over the




  course of a field season to  determine the seasonal  variability that  occurs




  in the toxicokinetics resulting  from  environmental  variables such as




  temperature and physiological  variables  such as changes in llpld content.
                                      358

-------
                          MATERIALS AND METHODS









    H_._ limbata were collected over the course of a field season beginning




in early May 1986 and continuing until November 1986 on approximately




monthly intervals.  The  collection site was near the middle of Lake St.




Clair, at Loran coordinates 49934 and 31128.  The depth was approximately




6m. The animals were collected by ponar grab, gently screened from the




sediment and placed  in  a container of lake water and sediment for transport




back to the laboratory  (generally about 2 h).  The bottom temperature was




taken at the same  time  the animals were collected.  The animals were




maintained in  the  laboratory in aerated (50  L) aquaria containing




approximately  3  cm lake sediment and  10 cm  lake water at the  temperature  of




collection.








     The compounds studied were  3H-benzo(a)pyrene (BaP)(specific activity




 23.8 Ci/mmol,  Amersham),  l^C-phenanthrene (Phe)(specific  activity 14




 mCi/mmol, Pathfinder Laboratories) and 14C-2,4,5,2',4',5'-hexachlorobiphenyl




  (HCB)(Specific activity 14.06 mCi/mmol,  Pathfinder Laboratories).  All




  compounds were determined to be at least 98% radiopure prior to use by thin




  layer chromatography  (TLC) and liquid scintillation counting (LSC) .  The




  purity check  was  performed on silica gel plates (E Merck, 250 /rn) using




  hexanerbenzene  (8:2 V:V) as the solvent system.  The plate was scraped at




  the same  Rf as  a standard and several sections lower.  These scrapings were




  placed in scintillation vials and counted.   The counts below the Rf of the




  standard were assumed to be degradation products.  All analytical procedures
                                     359

-------
were performed under gold fluorescent lights (X >  500 ran) to minimize  the




degradation  of the PAH congeners.









    Toxicokinetic studies were performed within one week of the  time  of




organism collection.  The accumulation from water  was performed as a




constant infusion, flow through, experiment.  The  temperature for each




experiment was the same as the environmental temperature measured at the




time of the  collection (Table 1).   Water, collected from Lake St. Clair at




the same time  as  the animals, was  filtered prior to use through glass  fiber




filters (Gelman AE) . The water was dosed in bulk (5 L) with the radiolabeled




compounds using a methanol carrier and allowed to  equlibrate for  1 h prior




to starting  the flow and introducing the organisms.  Methanol as  a carrier




is not expected to influence the kinetics (Landrum, 1983).  The animals were




provided with  artificial burrows to minimize the potential thigmotactic




stress (Henry  et  al. ,  1986).  The  artificial burrows used in these studies




were made from stainless steel screen instead of glass tubes.  Preliminary




studies with glass tubes indicated that depletion  of compound concentration




within the tubes  was occurring resulting in a very large variation in  the




accunulation of compound by the organisms (Landrum, unpublished data) .




Organisms were removed after 1,2,4,  and 6 h exposure,  blotted dry, weighed




and placed in  scintillation cocktail for radioanalysis.   The remaining




organisms in the  exposure chamber  were removed and placed in uncontaminated




sediment for depuration.   These animals were removed at approximately  1, 3,




5, 7 and 14  d, rinsed of sediment,  blotted dry,  weighted and placed in




scintillation  cocktail for radioanalysis.  Actual  times were used for




kinetic determinations.   The uptake and elimination rate constants were
                                   360

-------
determined by  fitting the data to a simple two compartment model using




initial  rate assumptions for uptake and a simple first order decay model for




elimination  (Landrum et al. ,  1985) .  The terms clearance constant and rate




constant to  describe the uptake constants determined will be used




interchangeably throughout this paper.   This uptake kinetics  constant  for




uptake actually describes the rate of clearance  of the compound from water.









     The accumulation from sediment was measured by sorbing  the radiolabeled




compounds onto sediments in an aqueous slurry overnight.  The  sediment  was




transferred to a large beaker and allowed to settle at 4°C for 24 h. The




overlying water was decanted and  the sediment mixed and  distributed to




individual beakers (approximately  75 g per beaker).  The actual weights of




the sediment in each beaker were  determined.  Subsamples of  the sediment




were  taken for analysis of the concentration of  radiolabeled compound and




 for wet to dry weight of the sediment.  Care was taken  to insure that the




 samples for the initial sediment  concentration  and for wet to dry weight




 were  taken at the beginning, middle and end of  the distribution of  material




 for exposing the organisms to insure  that  there  was no  apparent bias in the




 exposure  concentrations.  The beakers were placed in small aquaria,  six per




 aquaria,  and overlying, Lake St.  Glair, water  added.   The aquaria were




 allowed to stand for an additional 24 h before  adding the organisms. One iL.




 limbata was added per beaker and screening was  placed over  the top  of the




 beaker until  the organism had burrowed  into  the sediment. The burrowing was




 almost immediate.  Three organisms were removed for uptake,  one from each of




 three aquaria,  at approximately  1, 3,  5,  7,  and 14 d (actual times  were used




 for determination of kinetics).   The  animals  were rinsed of sediment,
                                     361

-------
blotted dry, weighed and placed in scintillation cocktail for radioanalysis.




Samples of sediment  were taken for analysis of toxicant concentration and




for wet to dry weight of sediment. All sediment concentrations were based on




the dry weight of sediment.   Accumulation kinetics were fit to a two




compartment model with a source function that could show exponential decay




to account. for changes in the concentration of the toxicant in the sediment




and for potential changes in the biological availability.  This model had




previously been  shown to be most appropriate for sediment accumulation of




PAH by Pontoporeia hoyi (Landrum, unpublished data).
       - KsCse-At - KdCa                                           (1)
 Where:




 Ca - concentration in the organism  (ng g~^)




 Cs - Concentration in the sediment  (ng g~^)




 Ks - Uptake rate constant from sediment  (g sediment  g"l  animal  h'l)




  A - Rate constant for reduction in the bioavailability  of  the  contaminant




     in the sediment




 K
-------
were obtained by  allowing the samples to sit for 2 to 3 weeks after mixing




with the ethylacetate  at  the time of sampling and separating the solvent by




filtration.  Subsequent soxhlet extraction after filtration did not yield




additional removal  of  radiolabeled material.  Concentrations of the




toxicants were determined by radioanalysis of a known fraction of the




extracting solvent.  The  bulk of the extract was reduced in volume by a




combination of rotary  evaporation and evaporation under a stream of nitrogen




to a final volume of approximately 500 pL.  A portion of the extract was




chromatographed by  TLC using hexane:benzene 8:2.  The extent of degradation




was determined as described above for the pre-use purity check.









    Organisms exposed through water were analyzed for biodegradation




products by extraction in ethylacetate:acetone 4:1 (2 X 20 mL) followed by




extraction with cyclohexane (1 X 20 mL) .   The extraction was accomplished by




nacerating the animals in a tissue grinder with the extracting solvent. The




extracting solvents were  combined and filtered from the residue of the




carcass.  The solvent  was reduced in volume and the biotransformation




determined as was done for the sediment  extracts.  The carcass residue was




analyzed for radioactivity as described  for whole animals.   The  extent of




metabolism was determined as the amount  of non-parent compound found by TLC




and the amount activity remaining in the  residue.









    Radiometric analyses  were performed  on an LKB 1217  scintillation




counter using the external  standards ratio method.  Additional measures made




over the course of the season included lipid content  and wet to  dry weight




aeasurements.   The lipid content was determined with  a microgravimetric
                                   363

-------
method  (Gardner e_t_al. ,  1985a) .   The wet to dry to ash free dry weight




measures were made by weighing animals wet within a few days after




collection.  These animals were desiccated to a constant dry weight and




subsequently ashed at 500°C for one h and the ash weight determined.  The




ashfree dry weight is the  difference between the ash weight and the dry




weight.









    Oxygen consumption  and respiration rates were determined at the same




temperature as the kinetics experiments.  One animal was placed in each of




five to ten 300 mL BOD bottles ,  the bottles capped and the oxygen




concentration determined after removing the animal.  The duration of the




experiment was 24 h.  The concentration in the water was compared to a




control containing the same water but no organisms.  The oxygen consumption




normalized to biomass was  determined by the Winkler titration using 0.005 M




thiosulfate (Grasshoff et  al., 1976).  Clearance constants for oxygen,




equivalent to the uptake clearance constants of the toxicants, were




calculated from oxygen consumption and the concentration of oxygen in the




control bottles.









    A  deterministic  simulation model of the toxicokinetics for H. limbata




was developed to examine the predicted seasonality of body burdens.  The




nodel used the general differential equation:
             + KSCS  - KdCa                                       (2)
                                   364

-------
Where 
-------
425 and  325 ng "1 respectively as determined in Lake Erie  sediments  (Eadie




et al.,  1982).
                          RESULTS AND DISCUSSION
     The dry to wet weight and  ashfree  to  dry weight determinations were




 made three times over the course  of the season.   The dry to wet weight ratio




 was 0.18 ± 0.02 (mean ± sd, n-3)  while  the ash free dry to dry weight was




 0.75 ± 0.10 ( mean ± sd, n-3).  Each determination employed 5 to 10




 organisms.  The lipid content appeared  to  peak at the beginning of June




 (Table 1) and ranged from about 4 to 6  % of dry weight from about August




 through the fall  (Table 1).  The   lipid content of H. limbata is low




 compared  to other Great Lakes  invertebrates.  The lipid content was most




 comparable with the oligochaetes.  The  chironomid larvae, another insect




 larvae  in the Great Lakes, contain nearly  twice the lipid content of JL.




 limbata (Gardner  et al.,  1985b) .   Oxygen consumption, respiration rates,




 also appeared to  peak  in  the  summer and declined toward the fall (Table 1) .




 Respiration rates in  the  spring and fall were significantly lower than rates




 in the  summer while  respiration rates  in the  fall were significantly  lower




 than spring  (p < 0.05,   Students T test).  The increase  in oxygen




 consumption in the summer,  over  those  measurements  in the spring and  fall,




 was probably due in part to increases  in  temperature.  However, the low




  oxygen consumption in October  was  at the highest environmental  temperature




  exaained.  This low October value  was  probably  due  in part to the  long hold




  time in  the laboratory before  the  measure was made  (approximately  60  d)
                                     366

-------
should not necessarily be considered representative for the time of year.




If the low value is incorrect then the oxygen consumption increases with




temperature as expected.  These measures of oxygen consumption are similar




to those found in the literature for H. limbata (Zimmerman et al., 1975).









     The toxicokinetics for water were measured on a monthly basis over  the




course of  the field season for 1986  (Table 2).  For BaP at the same




temperature, the accumulation clearance constants from water for H.  Limbata




were lower than those  for M. relicta.  a Great Lakes invertebrate of  similar




size  (Frez and Landrum,  1986).   The  elimination rate constants were  greater




than  those for either  M. relicta or  P. hovi  although the  study  temperatures




for H.  limbata were also greater.  The relatively  low  lipid content  of iL_




limbata probably accounts,  in part,  for  the  higher  elimination  rate




constants.  The uptake from water  generally  seemed to  rise  in the  summer and




decline again in the  fall  for all  the compounds.   The  rise  in uptake rate




 constant for the HCB  seemed to  lag that  of the  PAH.  The  uptake rate




 constants for HCB  generally had larger variances  than  those  for the  PAH.




 The reasons for this  are not clear.   The variances of  all the accumulation




 kinetics constants were generally greater than had been previously observed




 for other Great Lakes invertebrates (Frez and Landrum,  1986,  Frank et al.,




 1986).  The depuration rate constants also showed the  same trend of an




 increase  in the summer and a decrease in the fall.  These changes generally




 tracked the changes observed in the temperature of collection,  the




 temperature at which  the experiments were performed.   Bioconcentration




 factors  (BCF) calculated from the kinetics  in water showed a general trend




 of higher BCF  in  the  spring and summer for  PAH while the BCF for HCB was
                                     367

-------
essentially constant over the course of the season (Fig. 1).  No




biotransformation was attributable to the organisms; although, some overall




degradation was observed, 4-5% of the total.  The degradation did not




increase with time as is expected for PAH and has been observed with other




organisms  (Leversee et al.. 1982).  The degradation was not observed for




Phe.








      The accumulation of neutral organic xenobiotics from water  is  assumed




to  occur as passive diffusion across the respiratory membrane of  the aquatic




organisms.  A clearance rate for oxygen was determined  from the  oxygen




consumption experiments  (Table 1) and can be compared with the uptake




clearance  constants for the toxicants (Table 2).  In most cases,  the




toxicant was cleared from  the water with a  greater efficiency than  the




oxygen.  This increased efficiency suggests that  the integument  as  well as




the respiratory membrane is a route for some of the uptake.









      The sediments used  in the studies had  an  organic carbon content of




6.98%.  The size  fractionation of the sediment indicated that 31.9% of the




sediment mass was <74 ^m diameter.  This fine  grained material had  an




organic carbon content of  8.3%.  The determination of the uptake  rate




constant from sediment required the use of  a model that could account  for




changes in both  the chemical and bioavailable  concentration of the  toxicant




 in the sediment.  The extractable concentration of the  toxicants  often




 showed slight declines with time over the course  of the studies.  These




 changes are  accounted for  with the A value  in  the kinetics model.   The




 accumulation from sediments was essentially constant the two times  it  was
                                    368

-------
measured, (Table 2).  This measurement was made only twice due to the time




required for performance of the experiments and the analysis of the samples




generated.









     The assimilation efficiency from sediments can be estimated from




literature values of feeding rates  (Zimmerman and Wissing, 1978) and the Ks




values.  Our organisms were generally 20 mm in length or longer and Ks was




determined at 15 and 20°C.  The feeding rates for the larger nymphs were




0.21 g sediment g'1 organisms h'1 at 15°C and 0.31 g g'1 h'1 at 20°C after




converting the feeding rates based  on dry weight nymphs to a wet weight




basis and converting to hourly averages (Zimmerman and Wissing, 1978) .




These  feeding rates have the same units as the Ks values.  Thus, it is




possible  to compare the feeding rates directly with the Ks values.  Since




the Ks values are  less than the feeding rate the implication is that all of




the contaminant is  not being removed from the sediment as it passes through




the gut  of the organisms.  Therefore a ratio of Ks to feeding rates should




give the  fraction  of material assimilated or removed.  With these feeding




rates  the efficiency for assimilation for BaP ranged from 11.3 - 21%, Phe




from 13.6 - 31.3%  and HCB  from 14.4 - 29.1%.  This estimated assimilation




efficiency is in the same  range as  that determined for obligochaetes for HCB




in an  elegant dual  labeled study (Klump et al., 1987).









     Using the same estimates of water and sediment concentrations as used




for the  simulation model,  calculation of the amount of compound accumulated




from water versus  sediment at steady state was determined by the following




equations
                                    369

-------
Css - (KwCw + KSCS)/ Kd
Cw -
Cas - KsCs/Kd








Where Css  is the  concentration of  the  toxicant  in the  organism  from both  a




water and  sediment source at steady  state,  Caw  is the  concentration in  the




animal  at  steady  state from water  (ng  g'1),  and Cas  is the  concentration  in




the  animal at  steady state from sediment (ng g'1).   From these  equations  the




fraction from  water can be computed  from Caw/Css.  Similarly,  the fraction




from sediment  would be computed from Cas/Css.








      For the two times that the HCB accumulation from sediment was measured,




 the fraction of the body burden from the water was estimated to be 0.1 and




 0.03.   Thus the fraction  from sediment was 0.9 and 0.97.  The route of HCB




 accumulation for this organism is apparently via the  sediment.  Similar




 comparisons for  the PAH  yielded ratios  for BaP of 0.9 and 0.96 and for Phe




 0.95 for both  determinations  as the fraction of  toxicant obtained from the




 sediments.  This  determination is very  dependent on the ratio  of the




 products  of KWCW and  KSCS; therefore, changes  in environmental




 concentrations without  any change in  rate  constants would alter the fraction




 obtained  from a  particular source.








       Comparing the estimated role of  sediment  as a  source with other




  organisms, H.  limbata obtains a greater fraction of its body burden from the
                                     370

-------
sediment.  Using the same water and sediment concentrations as for the




projections for H. limbata and the kinetics constants for the oligochaete,




S. heringianus. (Frank et al., 1986) and P. hoyi (Landrum et al., 1985), the




oligochaetes would obtain 34 to 67% of their BaP body burden from sediment




while P.  hoyi would only obtain 39% from sediments.  This suggests that the




role of sediments as a source will depend on the organism as as well as the




sediment characteristics.









     Using the environmental data available, the deterministic simulation




model suggests that the highest concentration of the toxicants studied will




occur in the winter and early spring.  There is a consistent peak in the




body burden that occurs in the spring for all the compounds (Figs. 2-4) and




is driven primarily by a reduction in K^ resulting in a reduced flux out of




the organism (Fig. 5). The organism concentrations then decline during the




summer (Figs.  2-4).   In general the flux of compound into the organism is




primarily from the product of KSCS and under the conditions used in the




simulation for HCB would be 2.7 ng g~l h"l.  This flux from the sediment is




augmented by the  flux in from the water, KWCW, and is reduced by the flux




out of the organism, K
-------
impact the seasonal changes in body burden but there was insufficient data




to justify a change with season.  The bioaccumulation factor (BAF) (organism




concentration/sediment concentration) was based on sediment concentration




because it was the predominant source for the organisms.  The BAF ranged




from about 4.5 to 15.5 for HCB and reflects the generally low elimination




rate constant while PAH showed a much lower BAF, Phe 0.9 - 2.2 and BaP 1.5 -




3.8.  The change with season of the BAF values tracks the change in organism




concentration.  The BAF predicted for H. limbata are higher than those found




for oligochaetes for HCB in the field (Smith et al., 1985) and BaP (Eadie et




al., 1982) but were about the same as oligochaetes for Phe (Eadie et al.,




1982.  Comparing H. limbata to P. hovi the range of BAF's are about the same




for the two PAH studied (Eadie et al., 1985).









     In conclusion, H. limbata exhibits seasonal changes in the




toxicokinetics and these changes are expected to result in changes in the




BCF and BAF.  The organism appears to obtain the preponderance of its body




burden for the toxicants examined from the sediment based on the




calculations  from the toxicokinetics and sediment and water data from the




literature.
                                    372

-------
                             ACKNOWLEDGEMENTS









     This work was jointly supported by the Great Lakes Environmental




Research Laboratory, NOAA and by the U. S. Environmental Protection Agency




through interagency agreement No. DW13931213-01-01.  I wish to thank Tom




Fontaine for his critical discussions and suggestions on the kinetics




modeling.  I also wish to thank Brian Eadie, Tom Nalepa, Mike Quigley and




Guy Stehly for their critical review of this manuscript.
                                    373

-------
     Table 1.  Oxygen consumption and lipid content for Hexagenia in 1986


Date     Oxygen Consumption  Oxygen Clearance1   %Lipid    Temperature
                     -l'1       mL  '1 h'1                     °C
                62 mg-lh'1       mL g'  h'
May        0.327 ± 0.111        23.3 ± 7.9       7.8 ± 1.9     10
               n - 5                               n = 7

June    •       Lost                             15.1 ± 2.6     15
                                                   n = 5

July       0.667 ± 0.296        41.4 ± 18.3      9.1 ± 3.4     15
               n = 5                               ti = 8
                                                 4.3 ± 1.82

August     0.435 ± 0.10         61.9 ± 13.5      3.6 ± 1.0     20
               n _ 5                               n = 7

September    0.25 ± 0.06         44.5 ± 12        6.0 ± 2.4     20
               n = 10                              n = 6
                                                 6.9 ± 1.92

October3     0.16 ± 0.10         18.6 ± 11.8      3.7 ± 1.2     20
               n = 4                               n = 7
                                                 3.3 ± 0.92

November         ND                              6.0 ± 1.4     10
                                                   n = 4
 1.  The n for the clearance determination is the same  as  the  oxygen consumption
    determination.
 2.  Samples collected in 1985.
 3.  Sample actually collected on 30 September 1986,  Oxygen consumption was run
    60 d after collection.
 ND - not determined
                                     374

-------
           Table 2.  Seasonal Uptake and Elimination Rate Constants
                             for Hexaeenia  limbata
Month
May Ku*
    Kd'
**
June Ku
     Kd
     Ks
 ***
          Benzo(a)pyrene    Phenanthrene    Hexachlorobiphenyl  Temp.
 68.5 ± 11.2
0.011 ± 0.003

 67.0 ± 28.0
0.006 ± 0.002
0.043 ± 0.005
0.025 ± 0.0042
 131.1 ± 46.8
 0.032 ± 0.004

  43.3 ± 12.0
0.0076 ± 0.0016
 0.065 ± 0.016
 47.5 ± 23.9
0.007 ± 0.001

 44.2 ± 8.0
0.005 ± 0.002
0.030 ± 0.01
                                                                 15
July Ku
Kd
Aug Ku
Kd
Ks
Sept Ku
Kd
Sept3Ku
Kd
Nov Ku
Kd
101.9 ± 32.6
0.013 ± 0.002
65.1 ± 29.1
lost
0.035 ± 0.005
149.5 ± 29.0
0.016 ± 0.003
76.3 ± 41.0
0.028 ± 0.001
40.9 ± 30.6
0.010 ± 0.001
57.5 ± 5.0
0.029 ± 0.002
11.9 ± 4.0
lost
0.042 ± 0.008
56.3 ± 6.8
0.032 ± 0.004
33.0 ± 8.0
0.067 ± 0.008
34.2 ± 7.2
0.026 ± 0.002
40.8 ± 37.3
0.005 ± 0.001
40.8 ± 37.3
0.007 ± 0.001
0.09 ± 0.02
128.7 ± 20.3
0.015 ± 0.003
95.0 ± 17.3
0.017 ± 0.002
45.5 ± 16.1
0.004 ± 0.0006
15
20
20
20
10
* Ku has been corrected  for sorption to dissolved organic carbon and has
   units of mL g"*- h~l.
** Kd has units of h'l.
*** Ks has units of g dry  sediment g'l animal h'l
1. Temperature is in degrees centigrade.
2. Uptake from sediment  was measured twice for BaP
3. This collection was actually made on September 30, 1986.
                                    375

-------
                               LITERATURE CITED






Carr, J. F. and J. K.  Hiltunen.  1965.  Changes in the bottom fauna of western




    Lake Erie from  1930 to 1961.  Limnol. Oceanogr.  10:551-569.









Eadie, B. J.. W. Faust,  W.  S.  Gardner and T.  Nalepa. 1982.  Polycyclic




    aromatic hydrocarbons  in sediments and associated benthos in Lake Erie.




    Chemosphere 11:185-191.









Eadie, B. J., W. R.  Faust,  P.  F. Landrum, N.  R. Morehead, W. S.  Gardner and




    T.  Nalepa.  1983.  Bioconcentrations of PAH by some benthic organisms of




    the Great Lakes.   In:  Polynuclear Aromatic Hydrocarbons: Seventh




    International Symposium on Formation. Metabolism and Measurement.  M.




    W.  Cooke and  A. J.  Dennis, Eds. Battelle Press, Columbus, OH. pp. 437-




    449.









Eadie,  B. J., W. R.  Faust,  P. F. Landrum and N. R. Morehead. 1985. Factors




     affecting bioconcentration of PAH by the dominant benthic organisms of




     the Great Lakes.  In: Polynuclear Aromatic Hydrocarbons: Eighth




     International Symposium on Mechanisms. Methods and Metabolism. M. W.




     Cooke  and A.  J. Dennis, Eds. Battelle Press, Columbus, OH.  pp. 363-377.









Frank,  A. P., P.  F.  Landrum and B. J. Eadie.  1986. Polycyclic aromatic




     hydrocarbon rates of uptake, depuration, and biotransformation by Lake




     Michigan Stylodrilius heringianus. Chemosphere 15:317-330.
                                    376

-------
Frez, W. A. and P. F. Landrum. 1986. Species-dependent  uptake  of PAH  in




    Great Lakes  invertebrates.   In: Polvnuclear aromatic hydrocarbons:




    Ninth International  Symposium  on  Chemistry.  Characterization and




    Carcinoeenesis.  M. W.  Cooke  and A.  J.  Dennis,  Eds.  Battelle Press,




    Columbus, OH. pp.  291-304.









Gardner, W. S., W. A. Frez,  E. A. Cichocki  and C. C.  Parish.  1985a.




    Micromethod  for lipid analysis in aquatic invertebrates.  Limnol.




    Oceanogr.  30:1099-1105.









Gardner, W. S., T.  F. Nalepa,  W.  A. Frez, E.  A. Cichocki and P.  F. Landrum.




     1985b.  Seasonal patterns in lipid content of Lake Michigan




     macroinvertebrates.  Can. J.  Fish. Aquat. Sci. 42:1827-1832.









Grasshoff,  K.,  M. Ehrhardt and K. Kremling. 1976. Methods of Seavater




     Analysis.  Second Edition, Verlag Chemie, Federal Republic of Germany,




     pp. 419.









Henry, M.  G., D.  N. Chester and W.  L. Mauck. 1986. Role of artificial




     burrows in Hexagenia toxicity tests: Recommendations for protocol




     development. Environm. Toxicol. Chenu 5:553-559.









 Hiltunen, J. K. and  D. W. Schlosser. 1983. The occurrence of oil  and the




     distribution of Hexagenia nymphs in the St. Mary's River, Michigan and




     Ontario. Freshwat.  Invertebr. Biol. 2:199-203.
                                    377

-------
Howmiller, R.  P.  and A.  M.  Beeton.  1971. Biological evaluation of




     environmental  quality,  Green Bay, Lake Michigan. J. Wat. Pollut. Cont.




     Fed.  123-133.









Hunt, B. P.  1958. The life  history and economic importance of a burrowing




     mayfly, Hexagenia limbata.  in southern Michigan lakes. Bulletin of the




     Institute  of Fisheries  Research.  No.  4.  Michigan Department of




     Conservation,  Lansing,  MI pp 151.









Klump, J. V., J.  R.  Krezoski, M.  E.  Smith  and J.  L.  Kaster.  1987.  Dual




     tracer  studies  of the assimilation of an organic contaminant  from




     sediments by deposit feeding oligochaetes.  Can.  J.  Fish.  Aquat.  Sri




     44:(In press).









Landrum, P. F. 1983.  The effect of co-contaminants on the  bioavailability of




    polycyclic aromatic hydrocarbons  to Pontoporeia  hoyi.   Polvnuclear




    Aromatic Hydrocarbons:   Seventh  International Symposium on Formation.




    Metabolism and Measurement.  M.  W.  Cooke  and A. J. Dennis  eds. Battelle




    Press, Columbus, OH. pp. 731-743.









Landrum, P. F. , B. J. Eadie, W. R. Faust, N.  R. Morehead and M. J.




    McCormick. 1985. Role of sediment  in the bioaccumulation  of




    benzo(a)pyrene by the amphipod, Pontoporeia hovi.  Polvnuclear Aromatic




    Hydrocarbons: Eighth International  Symposium on Mechanisms. Methods and




    Metabolism. M.  W. Cooke and A. J. Dennis  Eds. Battelle  Press, Columbus,




    OH. pp.  799-812.
                                   378

-------
Leversee, G. J..  J.  P.  Giesy,  P.  F.  Landrum, S. Gerould,  J.  W. Bowling,  T.




     E.  Fannin, J.  D.  Haddock and S. M.  Bartell. 1982. Kinetics and




     biotransformation of benzo(a)pyrene in Chironomus riparius.  Arch.




     Environ.  Contain.  Toxicol.  11:25-31.









Schneider,  J.  C.,  F.  F.  Hooper and A. M. Beeton. 1969. The distribution and




     abundance  of benthic fauna in Saginaw Bay, Lake Huron.  Proc. 12th Conf.




     Great  Lakes  Res.  International Assoc. Great Lakes Res,  pp. 80-90.









Smith,  E. V., J.  M.  Spurr,  J.  C.  Filkins and J. J.  Jones. 1985.




     Organochlorine contaminants  of wintering ducks foraging on Detroit




     River  sediments.   J. Great Lakes Res. 11:247-265.









Thornley, S. 1985.  Macrozoobenthos of the Detroit and St. Clair rivers  with




     comparisons  to neighboring waters.  J. Great Lakes Res.  11:290-296.









Zimmerman,  M. C.,  T.  E.  Wissing and R. P. Rutter. 1975.  Bioenergetics of the




     burrowing  mayfly,  Hexagenia  limbata. in a pond ecosystem. Verh.




     Internat.  Verein.  Limnol.  19:3039-3049.









Zimmerman,  M.  C.  and T.  E.  Wissing.  1978. Effects of temperature  on gut-




     loading and  gut clearing times of the burrowing may fly,  Hexagenia




     limbata.  Freshwat.  Biol.  8:269-277.
                                    379

-------
                             LIST OF FIGURES









Fig. 1.    Calculated bioconcentration factors  in Hexagenia limbata over the




          1986 field season calculated  from the ratio of
Fig. 2.    Simulation of the concentration of HCB  in H.  limbata  through one




          season based on the concentrations of HCB in  sediments and water




          found in the Detroit River..









Fig. 3.    Simulation of the concentration of BaP  in H.  limbata  through one




          season based on the concentrations of BaP in  Lake Erie sediments




          and water concentrations  found in the Great Lakes .









Fig. 4.    Simulation of the concentration of Phe  in H.  limbata  through one




          season based on the concentrations of BaP in  Lake Erie sediments




          and water concentrations  found in the Great Lakes .









Fig. 5.    Simulated flux of HCB  into  the organism from  water and out of the




          organism through elimination.
                                   380

-------
                                                            r'*
OJ
00
*_
+-»
o
03
LL
centration
c
o
O
O
in

•

H
^
^


\c. \J\J\J

10 000-
r
2T500-

• ^
IBenzo(a)nvrene**^
!*•/ vi i^uv/yd/|^Jr i wl l W|
Hexochlorobiphenylj ^
* .A A
*i* • Mft
^Jfr ^F
PhenanthrerieK •
^^^ ' •
^i i^ 5fe
ii n1' or1 -in'1 o ' -i ' nr ' -*-• > r« I '-±"1 ' 1 ^
Jl Fl Ml Al MI J1 IJ 'A 'S 'D TN fD



•»
••
••

                              Time (months)

-------
               1000:
Lo
03
KJ
             O)
             O)
              O
              c
              O
             O
             m
             o
800-


6QQL


400^


200^
                   0!
          —I	I	,	,	,	
          601     120     160    i24Q]    300    £60
                Time (Julian days)

-------
CD
              2800;
                 0}
                  Q!
 1?_Q    180    I24Q
Time (Julian days)
300
1360

-------
               2400
to
oo
•C-
                                 12Q1   1SQ:    I240
                                 Time Gulian days)
1300
!360

-------
U)
oo
Ol
O)

O)


CD
O
I
H—
O

X

E
    1-


    OH
                =2r
                ~3r
                   0!
                        Water Uptake
                         Depuration

            60i    r120i__ J80	|240   300
                  Time (jtiilian days!

-------
    MODELING THE FATE AND  TRANSPORT OF CONTAMINANTS IN LAKE ST.  CLAIR
               Gregory A.  Lang and Thomas D.  Fontaine,  III
                               INTRODUCTION









     When looking at the Great Lakes System as a whole, two aspects of Lake




St.  Clair stand out: 1) its average depth is only about 3 m (the next most




shallow lake is Lake Erie at 19 m) and 2) its theoretical hydraulic




retention time is only about 9 days (again, the next closest is Lake Erie at




about 3 years).  Lake St. Clair has been described as simply a "wide part of




the  river" that extends from mouth of Lake Huron to the head of Lake Erie,




and  past Great Lakes budget calculations have generally overlooked the




dynamics of nutrient and contaminant transport into and through Lake St.




Clair.   Sediment studies of Frank et. al. (1977), Pugsley et. al. (1985),




and  Oliver and Bourbonniere (1985) document the levels of organic




contaminants, including polychlorinated biphenyls and octachlorostyrene, in




the  surficial sediments of Lake St. Clair and the St. Clair and Detroit




Rivers.  The concentrated chemical plumes located near the center of Lake




St.  Clair, between the South Channel and the Detroit River, indicates the




ability the lake's sediments to trap particle-bound contaminants.









     Recent attention to the connecting channels of the upper Great Lakes




and the recognized harmful effects to biota and the extreme environmental
                                    386

-------
persistence of many of these hydrophobic  chemicals  have served  as  the




impetus for studying  the  dynamics  of  contaminant transport within  Lake  St.




Clair.  The objectives of this  study  were two fold: 1) to develop  a multi-




segment mass balance  model to  simulate contaminant  fate and transport  in




Lake St. Clair, based on  the Environmental Protection Agency's  model




TOXIWASP; and  2)  to  calibrate,  verify, and apply the model using available




contaminant and tracer  data.   This report presents  the model, briefly




describes the  model  processes  affecting chemical and solids concentrations,




details  the physical lake characteristics and model segmentation,  describes




 the  tracer and contaminant input data, and presents the simulation results.
                              MODELING APPROACH
      The mathematical model used in this analysis was based on the




 Environmental Protection Agency's Chemical Transport and Fate Model TOXIWASP




 (Ambrose et.  al.  1983).  TOXIWASP combines the kinetic  formulations in the




 Exposure Analysis Modeling Simulation,  EXAMS  (Burns et. al. 1982), and the




 transport processes in the Water Analysis Simulation Program, WASP (DiToro




 et. al. 1983).  The mechanics of TOXIWASP were left relatively unchanged.




 However, in an attempt to streamline  the model,  improve its execution time,




 and to make it more specific  to Lake  St. Glair's physical  and contaminant




 data, some modifications were made.   Numerous programming  errors  found in




  the source code  (in particular, subroutine  SETTLE) were corrected.  The  EPA-




  Athens modeling  group  was  informed  of all corrections  needed.
                                      387

-------
     Two state variables are included  in TOXIWASP, total organic chemical




and total sediment.  Sediment concentrations are affected by advection,




dispersion,  mass loading, settling,  and resuspension.  Chemical




concentrations are affected by  these same  processes, plus degradation,




sediment-water diffusion, and biological mixing  to deep  sediments.  Chemical




degradation is due to hydrolysis,  biolysis, photolysis,  oxidation,  and




volatilization.  Sorption onto  sediments  and biota  is  calculated via




equilibrium kinetics using  a chemical-specific partition coefficient  and




spatially-varying  environmental organic carbon fractions.   These transport




and  transformation processes are detailed fully in the EPA TOXIWASP Manual




 (Ambrose  et.  al.  1983),  EPA EXAMS Manual  (Burns et.  al 1982),  and the EPA




WASP Manual  (DiToro  et.  al.  1983).









 Physical Lake Characteristics









      The volume and surface area  of Lake  St.  Clair were taken to be 3.75 km3




 and 1065.7  km2, respectively.  The  study  area was segmented into 126 well-




 mixed segments: 42 water segments,  42 active  sediment layer segments, and 42




 deep sediment layer segments (Fig.  1).  The segment numbers in  Figure 1




 correspond to water column  segments;  active layer and deep layer segment




 numbers are determined  by adding one  and  two, respectively, to  the water




  column numbers.   The average segment size was about  5 km on a side.  The




  segmentation  scheme was based  on the results  of two  contributing projects.




  One was  an extensive cluster analysis performed on the  available chemical,




  physical, biological,  and  contaminant data in the  water column  and sediments




  of Lake St.  Clair (Rybczyk 1986).  The other  was the 1.2  km grid Lake  St.
                                      388

-------
Clair Rigid Lid Hydrodynamic Model  (Schwab and Liu 1987) which generated the




wind-driven flow fields used in the present  study to  transport solids and




chemical throughout the water column.   The 126-segment  grid  and the 1.2 km




grid were superimposed such that  segment interfaces were  shared.  The




hydrodynamic model flows were summed  along the  larger TOXIWASP segment




interfaces yielding 75 net  interface  flows.









     The volumetric inflow  and  outflow rate  was constant at 5700  cms.   The




depth of the water column segments ranged from 2 m to 5.2 m (areal  average  =




 3.5  m) .  The depth of the active sediment layer was uniform at 2  or 10 cm,




 depending  on the  simulation.   The depth of the deep sediment layer was equal




 to the  total depth of the  "recent sediments" minus the active layer depth.




 The "recent sediments" are  defined by Robbins and Oliver (1987)  as those




 sediments  deposited  on top of the post-glacial till and range from 3-30 cm.




 The rate of horizontal dispersion was  assumed to be IxlO5 cm2/s (Great Lakes




 Institute  1986).   Pore water diffusion was  set to IxlO'5 cm2/s.
  Solids
      Suspended solids  from tributary sources were loaded at a  constant rate




  of 2700 MT/day into 8  water column segments.  The St.  Clair River  load




  entered segments 16, 25,  37,  49,  and 55 (Fig. 1), the  Sydenham River  load




  entered segment 58; the Clinton River load entered segment 19,  and the




  Thames River  load  entered segment 124.  The loads were based on average




  tributary concentrations and flows reported by the Great Lakes Institute




  (1986) .   The  settling  velocity of solids was set equal to 3 m/d,  based on
                                      389

-------
results of an empirical modeling procedure  for Lake St. Clair (Simons and




Schertzer 1986).  The resuspension velocity in each segment was calculated




such that the particulate  settling and  resuspension fluxes across each




segment's sediment/water interface were equal  (i.e.,  the net sedimentation




rate was assumed zero) .  The  assumption of  zero  net sedimentation is based




on the relatively shallow  depth of the  recent, post-glacial sediments




(Robbins and Oliver  1987).









     Under constant  wind  (and thus constant flow)  conditions, the solids




concentration  in each  water column segment  could be calculated  from the




advective flux of solids  to and from each segment,  the dispersive solids




flux between segments,  the rate of solids loading  into each segment,  and  the




assumption of  zero net sedimentation.  The  suspended  solids concentrations




resulting from a steady 6  m/s wind  from the southwest ranged  from 4.7-10.5




mg/1  (volumetric average  - 6.33 mg/1).   These values  are consistent with




 those  measured by Bukata et.  al.  (1987) along ship transects  during three




 separate cruises  in  Lake St.  Clair  in September, 1985.  They  reported a




 range  of 2.5-16 mg/1,  with a mean of approximately 5  mg/1.  The




 concentration  of  solids in the active and deep layer model sediments was  set




 to 1.2xl06 mg/1, based on the average of 19 10-cm sediment cores from 1985




 (Robbins 1986, Great Lakes Environmental Research Laboratory, personal




 communication).  The calculated resuspension rates ranged from  0.42-0.95





 cm/yr.








      The model does not distinguish  between different types of sediment,  a




 particular concern when simultaneously  modeling nearshore and offshore
                                     390

-------
concentrations of organic, hydrophobic contaminants.  However, it does




provide a parameter for defining the spatially-varying  organic carbon




content of the sediment.  It was assumed that  the  organic  carbon content in




each segment would not vary over time; i.e., particulate matter resuspended




from the active layer would immediately  be  available  for transport to




another segment, but the  average organic carbon content of each segment




would be maintained over  time.









     TOXIWASP requires  the user  to  quantify certain model  parameters and




constants; parameters vary over  space  but not  time, constants do not vary.




Ranges and values of Lake St.  Clair parameters and constants used  in the




present study are presented in Table 1.   Constants associated with




particular contaminants  are presented in separate  tables.









Chemical









     The  chemical being modeled  was loaded into Lake  St.  Clair from




 atmospheric,  land runoff, and/or tributary sources.  A whole-lake  areal




 atmospheric  load was  calculated  from available flux measurements  and  then




 segmented according to  each  segment's surface area.  The tributary chemical




 loads  entered the same  segments  as the solids loads (segments 16,  19,  25,




 37, 49,  55,  58,  and 124); although, not necessarily all eight.  For example,




 octachlorostyrene,  thought to originate mainly from the Sarnia area,  entered




 the lake  through segments 37  and 49 only.  The total load from land runoff




 sources was  partitioned to all water segments bordering the shore according




 to each segment's  drainage area.
                                     391

-------
     The model assumed a local equilibrium between the dissolved, sorbed,




and bio-sorbed chemical as defined by the organic carbon content of




particles and octanol-water partition coefficients, KQC and KOW.  KOW, held




constant throughout time and space, was multiplied by the varying organic




carbon content of the sediment,  resulting in  a  spatial distribution of the




solid/water partition coefficient, Kp,  that corresponded closely to the




distribution of fine-grained,  organic-rich sediment.  KOW was multiplied by




the organic carbon content  of  the biomass to  yield  an overall biota/water




partition coefficient, KD.









     Vertical  distributions of excess lead-210, measured from a set of




 diver-collected sediment cores from Lake St.  Clair  in 1985,  possessed a zone




 of constant  activity extending down to about 3 cm with exponential  fall off




 below  (Robbins and Oliver 1987).  Analyses  of these distributions plus  those




 of cesium-137  indicate sedimentation rates  on the order of 0.1-0.2  cm/yr.




 Robbins and  Oliver state that because of the evidence of extensive




 biological  activity down to at least 15 cm,  biological mixing,  and  not  net




 particulate  settling, is favored as the mechanism producing observed




 vertical radionuclide and contaminant sediment distributions.  Because  there




 is no mechanism for biological mixing in TOXIWASP, the model was modified to




 account for constant, spatially-varying, burial of chemical from the active




 sediment layer to the deep sediments.  This  rate was set to 0.1 cm/yr





 throughout the lake.
                                     392

-------
                                SIMULATIONS









     A series of simulations were performed during this  study  for purposes




of model calibration, verification, and application.   These  include




simulations of the conservative  ion chloride,  the  radionuclide cesium-137,




and the organic contaminants octachlorostyrene and polychlorinated




biphenyls.   Descriptions of simulation conditions,  available chemical




observations, loading functions, and  chemical  constants  required for these




simulations are presented here;  the results are presented in a later




section.
 Chloride









     Chloride, a conservative ion,  was chosen to test the efficacy of the




 transport mechanisms  in  TOXIWASP.   Bell (1980) documents chloride  and




 meteorological data collected during a series of cruises in Lake St.  Clair




 during the summer of  1974.   Cruises 3 (19-29 June), 5 (15-24 July),  6 (5-15




 August), and 8 (16-25 September)  were the most complete and were thus




 selected for this study.   At least 30 stations were monitored during each of




 the 4 cruises.  An average chloride load to the lake was calculated for each




 cruise as the product of tributary flows and chloride concentrations.  The




 flow fields during each  cruise were generated with the hydrodynamic model




 using the cruise-averaged wind conditions (speed and direction).  The model




 flows were multiplied by a factor of 0.93 to account for the difference in




 total inflow between  the 1974 value (5300 cms) and that used in the
                                     393

-------
hydrodynamic model (5700 cms).  The horizontal  dispersion coefficient was




set to 1x10^ cm^/s for all  segment  interfaces.   Chloride is  a dissolved




conservative substance; thus  partitioning,  settling,  resuspension,  and




degradation were not included.









     Four steady-state  simulations  were performed, one corresponding to  each




cruise; each with a different chloride load and wind-induced flow field.




The average wind conditions for cruises 3, 5, 6, and 8 were 4 m/s from the




northeast,  5 m/s from  the  north,  5  m/s from the east, and 6 m/s from the




west,  respectively.  The  total chloride loads to Lake St. Glair during the




 four  cruise periods were  3.7xl06 kg/day, 3.4xl06 kg/day, 3.2xl06 kg/day, and




 3.3x10^ kg/day,  respectively.  Eight additional simulations were generated




 for each  cruise;  one  for each wind direction other than the observed




 direction and one  for  the no wind case.









 Cesium-137








      Cesium-137, a surrogate of  many  hydrophobic,  organic contaminants, was




 chosen to calibrate and verify the processes associated with sediment-bound




 contaminant movement.  A  35-year history (1950-1985) of Cs-137 loading  to




 Lake St. Clair (Figure 2)  was extracted from Robbins and Oliver (1987).




  During this period, the largest  portion of loading has been inflow from Lake




  Huron.  A  nearly equal contribution (about 40% of the total load)  originated




  from the atmosphere during the 1950s  and early 1960s.  Lake Huron presently




  contributes about 75%  of  the total load, while the atmosphere contributes




  less than  20%.  The remainder of the load (about 5%) originates from land
                                      394

-------
runoff sources.   Observed cesium concentrations  in  the upper  2 cm of bottom




sediment were available for 1976 and  1985  (Robbins  and Oliver 1987).  In




both years,  concentrations were closely  related  to  the thickness of the




recent sediments, with the highest  concentrations  located in  the corridor




between the mouth of the South Channel of  the St.  Clair  River and the head




of the Detroit River.  In addition, Robbins and  Oliver estimated the total




accumulation of  cesium-137 over  the entire depth of sediments in 1985.









     Cesium-137  has  a half life  of 30.2  years and a partition coefficient,




Kp, of about  29000 lw/kg.  The partition coefficient was calculated from the




 lake-averaged suspended  solids concentration in Lake St. Clair and  the




 fraction  of  dissolved  contaminant (0.847), assumed constant for all of  the




 Great Lakes  (Robbins 1985).   With an average organic carbon content of  the




 open-lake sediments  equal to about 2.5%, Koc is calculated to be 1.16x!06




 lw/kg-  The  active  sediment  depth was set  to 2 cm  to allow direct comparison




 with  the  available  1976 and 1985 data.









      The  model  was  run for 9700 days, corresponding  to  the period 1950 to




 mid 1976, with  a constant flow field generated by  6  m/s wind from the




 southwest.   Wind frequency data from meteorological  sampling stations in the




 St. Clair Region during the period 1951-1980 show  a  predominantly moderate




 wind speed (5-8 m/s) with a southwesterly flow  in  all seasons  (Great Lakes




 Institute 1986).  Initial 1950 conditions of cesium-137 were set to zero in




 all segments.   The organic carbon  content of each  segment was  adjusted




 within the range of O.OX to 5% until the  model  results  matched the 1976




  cesium data.  This  calibration  exercise was verified by running the model
                                      395

-------
for an additional 3300 days  (9 years) and comparing the  results  to the




observed 1985 data.









Octachlorostyrene (PCS)









     DCS was first reported  in the lower Great Lakes in 1980 (Kuel  et.  al.




1980); however,  information  regarding its sources and environmental  effects




is limited.  The only  documented and recognized regional source  of  OCS,  a




by-product of  several  industrial processes including chlorine and solvent




production, appears  to be chemical companies in the Sarnia area  (Great  Lakes




Institute  1986,  Oliver and Bourbonniere 1985).  OCS concentrations  in




sediments  from the  St. Clair River indicate that the source is located  along




the eastern shore  of the St. Clair River, possibly from drainage of the




Scott  Road landfill  and  Dow Chemical's First Street Sewer discharge (Oliver




 1987).   Little historical data exists, however, and the magnitude of total




 loading to Lake St.  Clair is unknown.  Limno-Tech, Inc. (1985) speculated




 that  OCS was  introduced to the lower Great Lakes beginning in the 1970s as




 industries manufacturing chlorine converted from a process using mercury to




 one resulting in OCS as a by-product.  Concern exists because OCS tends to




 bioconcentrate and is a chlorinated hydrocarbon.  The simulation of OCS in




 Lake St. Clair is intended as an  application  of the model.









      Pugsley et. al.  (1985) presented  the distribution  of OCS levels in




 surficial (0-10 cm),  diver-collected sediments for  1983.  The values ranged




 from non-detectable to  26.2 pgAg-  The  distribution  of OCS  showed a




 concentrated plume extending  from the  mouth of the  South Channel towards the
                                     396

-------
the head of the Detroit River, very  similar  to  the  pattern of  cesium-137 in




the surficial sediments.  Great Lakes  Institute (1986)  estimated  the total




load of OCS to Lake St. Clair  to be  about 1.9  Ibs/day,  based solely on model




results in which the  load was  adjusted until predicted sediment




concentrations matched  the  data.   In the present study, the load  was assumed




to enter the lake  solely  through  the South Channel outflow because the




source of OCS  is believed to be  located along  the eastern shore of the St.




Clair River.









     The model was run for 5000 days with a constant flow field




corresponding  to  a 6  m/s wind from the southwest.  Initial conditions  of OCS




were set to  zero  in all segments based on the assumption that background




 levels  of  OCS  in  Lake St. Clair prior to the 1970s were negligible.  The




 active  layer depth was set to 10 cm for  direct comparison with the data.




 The chemical constants used in the OCS  simulation, taken from Ibrahim




 (1986),  are  presented in Table 2  (e.g.,  KOW, Koc,  molecular weight, etc.).










 Polychlorinated Biphenvls  (PCBs)









      As a final application,  the  model  was  used to simulate the  fate and




 distribution of total  PCBs  in Lake  St.  Clair.   PCBs  were  first prepared in




 1881 and had been manufactured and  extensively used since 1930 for




 industrial purposes  where  extreme thermodynamic conditions exist such as




 dielectric  fluids, heat  transfer agents, and  flame retardants (Limno-Tech,




  Inc. 1985).   They are  formed  by  the chlorination of biphenyl  in  the presence




  of an  iron  catalyst.  PCBs are relatively nonvolatile, insoluble in water,
                                      397

-------
soluble  in organic compounds, have high  dielectric constants, are relatively




inert towards acids,  alkalies, and other corrosive chemicals, and are stable




towards  oxidation (Roberts et. al. 1978).  They  are  of  particular concern




because  of their known toxicity to biota,  low  solubility  in water,




bioaccumulation, and extreme environmental persistence.









     Frank et. al. (1977) measured PCB concentrations of  surficial  sediments




 (0-2 cm) collected from Lake St.  Clair in 1970 and 1974.   They  noted that




 the distribution of PCB in 1970 showed a concentrated plume  entering from




 the St.  Clair and Thames Rivers.  Great Lakes  Institute (1986)  estimated the




 annual load of total PCB to  the lake to be 1860 Ibs  (5.1  Ibs/day),  based on




 model simulations in which the  load  was adjusted to  reach the best




 comparison between model predictions and measured values  in 1974.   This




 value is comparable to  1-4 Ibs/day,  a Lake St. Clair PCB  load estimate




 calculated from Lake Huron outflow  concentrations and average atmospheric




 and tributary loading  rates  for Lakes Erie and Huron during the late 1970s




 (Thomann  and Mueller 1983).   It was  assumed that 18Z of the total load




 entered through Thames  River,  16% through the Clinton River,  and 66% through




 the St. Clair River.   The  atmospheric input of PCB was assumed negligible




 based on  an  atmospheric loading rate of l.OxlO"10 Ibs/m2/day which is the




 average value for Lakes Erie and Huron  (Thomann and Mueller 1983).









      The  model  was run for 4 years with a constant  flow field corresponding




 to a 6  m/s wind from the southwest.   Initial  conditions of PCB were set




 equal to  the 1970 values (Figure 3) .  The active layer was set to 2 cm for




 direct  comparison with the data.   The chemical  constants used in the PCB




 simulation,  taken from Mabey et.  al.  (1982),  are presented in Table 3.




                                     398

-------
                          RESULTS AND DISCUSSION









Chloride









     The magnitude and  distribution of  simulated chloride  concentrations in




the water column of Lake  St.  Clair  agree very well with the  corresponding




data for cruises 3, 6,  and  8.   Figures  4 and 5 compare the observed




concentrations  (left  frame)  and the predicted concentrations (right  frame)




for all four cruises.   The  model did especially well in predicting the




chloride concentrations in  Anchor Bay and in locating the  7, 7.5, and 8 mg/1




contour lines.  Although it did quite well overall, the simulation




corresponding  to cruise 3 (Figure 4a) tended to overpredict  the




concentrations along  the northeast and east shores of the  main lake  (from




Figure  1,  segments 55,  58,  61,  79,  82,  and 103).  The simulation




corresponding  to  cruise 5 (Figure 4b) overpredicted the entire eastern  half




 of the  main lake.   This condition will be addressed in the next paragraph.




 The simulation corresponding to cruise 6  (Figure  5a)  slightly underpredicted




 the values along the  eastern shore between Anchor Bay and the main lake




 (segments  25,  34,  37, 46, and 49).  The  simulation corresponding to cruise 8




 (Figure 5b) was probably the best  of the  four,  although it  slightly




 underpredicted the concentrations  along  the  northwestern  shore of the main





 lake (segments 28, 40, and 64).









      Examining the simulations  generated using wind directions other than




 the observed directions  and  the no wind case revealed that  simulations using




 the observed winds for cruises  3,  6, and 8  were unique solutions.  They more
                                     399

-------
closely matched the data  than  simulations  using  any  other wind direction and




the no wind case.  The  simulated  chloride  concentrations for  cruise 5 did




not agree well with the data using  the  observed  average wind  speed of 5 m/s




from the north (Figure  4b) .  However, simulations  generated using winds from




the NE, E, and SE agree much better with the data.   Looking more closely at




the observed wind directions from cruise 5;  13 stations measured winds from




the north, 7 from the NE, 3  from  the E, 4 from the SE,  1  from the S, zero




from the SW, 2 from the W,  and 3  from the NW.  Therefore,  it  is not




inconceivable that the  NW,  E,  and SE simulations agree well with  the data




since  14 of the  33 stations reported winds from these three directions.




Cruise 5 results may  point out the  importance of easterly winds  in  flushing




dissolved  compounds from  the relatively low flowing eastern portion of Lake




St. Clair.









      In general,  considering that the wind conditions were averaged over




 10-11  days and  over  at least 30 stations, the distribution of simulated




 chloride concentrations  in Lake St. Clair were quite close the observed




 data.   The results  of  the chloride  simulations tend to confirm the accuracy




 of the wind-induced transport mechanisms used in the Lake St. Clair




 Contaminant Fate and Transport Model.
 Cesium-137
      Model-simulated 1976  cesium-137  concentrations  in  surficial  (0-2 cm)




 sediments compare well with  observed  1976 values  for Lake St. Clair  (Figure
                                     400

-------
6a).   Through calibration of the organic carbon  content  of  the sediments,




the  model was able to match the magnitude  and  distribution  sediment-bound




contaminant in the active layer throughout the lake.   The calibrated organic




carbon content values closely matched the  observed distribution  of fine-




grained surficial  (0-10 cm) sediments in 1983  and 1984 and  the observed




distribution of organic carbon  content of surficial (0-10 cm), SCUBA diver




collected sediments  in  1983  (Great Lakes Institute 1986).   However, the




magnitude of the  calibrated values ranged from 0.14X to 5X  (areal mean -




1.27X), while  the observed values ranged 0.06X to 1.9Z (mean - 0.58X).




Kaiser et.  al.  (1985)  and Maguire et. al. (1985) reported values of percent




organic  carbon in Detroit River sediments which were also consistently




higher than those reported in the Great Lakes Institute report.   Ibrahim




 (1986) and  Great Lakes Institute  (1986) used values of 5X for their model





 simulations.









      The calibrated organic carbon contents,  combined with the  organic




 carbon partition coefficient, Koc, and  the  suspended  solids concentrations,




 yielded water column dissolved Cs-137  fractions of 0.77-0.99; the lower




 values located in the  zones of highest  deposition (i.e. fined-grained




 sediment, rich in organic carbon).   The average value for  the open-lake




 segments was 0.84  (n-10).  Robbins  (1985) reported a  constant value of  the




 fraction of dissolved  Cs-137 equal  to 0.847 for the open-waters of each of





 the Great Lakes.








      Without  any further calibration, the model was run for an  additional




  3300 days,  corresponding to  1985.  The predicted magnitude and  distribution
                                      401

-------
of sediment-bound Cs-137 in  the  active  layer  compare well with observed 1985




values (Figure 6b).  Cs-137  was  concentrated  in the sediments along the




South Channel-Detroit River  corridor  in 1976  and 1985,  closely matching the




distribution of fine-grained,  organic-rich sediments  in the  depositional




zones of Lake St. Clair.   Modeled and observed maximum cesium-137




concentrations in the active sediment layer declined  during  this  time  period




from about 5 dpm/g  to about  2 dpm/g due to decreased  loading, radioactive




decay, particle resuspension, and burial to the deep  sediments.   In




addition,  the model predicted the total lake-wide accumulation  of Cs-137  in




Lake St. Clair sediments to  be 41 Ci in 1985 (corresponding to  an average of




8.6 dpm/cm2), which agrees reasonably well with the measured 1985 value  of




37 Ci  (Robbins and Oliver 1987).








      During  the  35 years of simulation, the model predicted a total Cs-137




 loading  of 800 Ci.   Of the total load, 720 Ci or 90X, exited through the




 Detroit  River,  38.8 Ci or 5X, were lost due to radioactive decay, and 41.5




 Ci or 5X,  remained in the system.  The calibration and verification




 exercises  performed during  the  cesium-137  simulations provided valuable




 insight  to the  processes  associated  with  sediment-bound contaminant




 movement.   Cesium-137 proved  to be a unique  data set  in Lake St. Clair.   The




 loading function was well documented,  the initial conditions were known, and




 two sets of spatially-complete  observations  were available.  The knowledge




 gained from the Cs-137  simulations  (e.g.,  organic carbon contents) was




 applied to the following  OCS and PCB simulations.
                                     402

-------
Octachlorostyrene


     The simulation of DCS  in Lake  St.  Clair was intended as  an application
of the model, not necessarily as  a  further test of the model's  ability  to
predict an observed distribution.   A fundamental problem existed:  the actual
time-variable loading function was  unknown.  However, on the  basis of Great
Lakes Institute  (1986) estimates,  the load was assumed to be  constant at 1.9
Ibs/day.  The model was  run until the simulated DCS levels in the active
layer (0-10  cm)  agreed with the observed 1983 values.  This occurred at 4500
days, or just over  12 years (Figure 7), implying that the load was first
 introduced  in the  latter part of 1970.  This result is consistent with
 Limno-Tech,  Inc.'s  (1985) speculation that OCS was introduced to the lower

 Great Lakes  beginning in the 1970s.


      Using the  sediment organic carbon values  from this  study, the estimated
 OCS load,  and the constants  in Table 2, the  model was able to reproduce the
 observed OCS pattern in the  sediments; that  of a concentrated plume of OCS
 extending from the mouth of  the South Channel  to the head of the Detroit
 River.   However, the model-calculated plume  was wider and slightly longer
 than the measured plume.   The  model  predicted that the mean and maximum
 sediment concentrations  in 1983 were 3.8  and 23.1 MgAg, respectively.
 These compare with 2.7  and 26.2  ^gAg  reported by Pugsley et.  al.  (1985).
 The model predicted  1983 active  layer  bio-bound OCS  levels of  0-96 ,gAg dry
 wt., with a mean concentration of  20 ,g/kg dry wt.   Pugsley  et.  al.  (1985)
 reported levels of 2-154 MgAg dry wt.  (mean - 43 MgAg dry  wt.)  measured in
 whole clam  tissue  (lamUit ^ia^ jlUuaidtt)  collected  in Lake St.

  Clair during 1983.
                                    403

-------
     Even though 55X of the St. Clair River  flow  is  directed  through the




North and Middle Channels, the negligible  levels  of  OCS  in  the sediments of




Anchor Bay tend to verify  that the majority  of the OCS load to Lake St.




Clair entered through  the  South Channel  from its  likely source in Sarnia.




Model-calculated loss  rates ranged from  0.03-0.08 day"! in  the water column




and 0.34x10-^ - 0.37xlO'3  day1  in the sediments.









     During  the 4500-day simulation  period,  the model predicted  that  3.9 MT




of OCS entered Lake  St.  Clair.   Of this  total load,  2.6 MT or 68Z,  were




flushed  from the  system through the  Detroit River, 0.7 MT or 18X,  were lost




due  to biological  degradation and volatilization, and 0.5 MT or  13X,




 remained in  the  system.   That which remained  in  the lake was concentrated  in




 the  sediments between the South Channel and the  Detroit River.   It is worth




 noting once  again that the OCS simulation was performed without prior




 knowledge of the OCS load to Lake St. Clair.  Although we were able to




 adequately reproduce the  1983 OCS observations using  a constant load of 1.9




 Ibs/day for 4500 days, this solution is not necessarily unique.   There is  no




 reason to believe that the load remained  constant during the entire




 simulation period.  Any number of time-varying load magnitude and duration




 combinations could have produced similar  results.   However,  with known




 initial conditions  (zero  concentration  in all segments),  the model was able




 to predict how much OCS  (3.9 MT) had to be  loaded into the system  to  produce




 the 1983  observations.   It also provided  some insight on the origin of the





 OCS load  to  Lake  St.  Clair.
                                      404

-------
Polvchlorinated Biphenvls
     The simulation  of total PCB in Lake St.  Clair was also  intended as an




application of  the model.   However, as with the OCS simulation,  the loading




function of PCB was  unknown during the period 1970-74.  Great Lakes




Institute  (1986)  estimated the load to be about 5 Ibs/day, based on model




simulations.   This  load was used in the present study.  With initial




conditions set equal to conditions in 1970 (Figure 3), and using the




constants  presented in Table 3, the model was able to reproduce the  observed




 sediment  PCB  distribution in 1974  fairly well (Figure 8).  In general,  the




 model  accurately predicted  the  1974 PCB sediment concentrations in the




 Anchor Bay and the open-lake sediments.  However,  the model tended to




 overpredict the PCB values  along  the  eastern and western segments of the




 main lake, which may  indicate  additional or increased PCB sources in these
 areas.
      The data  indicate a decline in mean lake-wide sediment  (0-2 cm)




  concentration  of  total PCB from 19 MgAg 1970 to 10 MgAg in 1974 and in




  maximum sediment  PCB concentration from 40 MgAg in 1970  to  28 ,g/kg in 1974




  (Frank et.  al.  1977).  The model simulated a similar 4 year  decline in mean




  active layer sediment concentration from 19.8 MgAg to 8.7 Mg/kg -i* -




  maximum  sediment  concentration from 39.0 ,gAg to 26.0 MgAg- The model




  predicted 1974 active layer bio-bound total PCB levels of 31-212 ,gAg dry




  wt   with a mean concentration of 97 MgAg-  *, comparable literature




  estimates of bio-bound  PCB were available for the early 1970s.   However,




  Pugsley et. al.  (1985)  reported mean values of 90.6 and 44.2 ,gAg whole
                                      405

-------
clam tissue (L. radiata)  for Aroclors  1254  and  1260,  respectively, in Lake




St. Glair during 1983.









     The model-calculated volatilization rates  ranged from 0.11-0.18 m/d.




This range compares  well with the theoretical rates for Aroclor 1242, 0.21




m/d, and Aroclor  1260,  0.17 m/d (Richardson et. al. 1983).  The total loss




rate was calculated to be 0.04-0.11 day1 in the water column and 0.34x10'^




 -  0.45x10'3  day"1 in the sediments.









      From 1970 to 1974,  the model predicted  that the  total system mass of




 PCB decreased from 2.5 MT  to 1.5 MT.  During this  time period, 3.4 MT of PCB




 were loaded into the system, 2.3 MT were flushed from the lake through the




 Detroit River, and  2.1 MT  were  lost due to biological degradation and




 volatilization.  Again,  as with OCS,  the PCB simulation was  performed




 without prior knowledge  of the  PCB load to Lake St.  Clair.   Thus, the 1974




 solution, although  an adequate  reproduction of the data, is  not  necessarily




 unique.  Any  number of time-varying load magnitude and duration combinations




 could  have  produced similar results.   However, with known initial conditions




  (1970  values),  the model was able to predict how much PCB (3.4 MT)  had  to  be




  loaded into the  system  to produce the  1974  observations.  It also provided




  some  insight to the possibility of additional PCB sources along the eastern





  and western main-lake segments of Lake St.  Clair.
                                      406

-------
                             LITERATURE CITED









Ambrose, R.B., Hill, S.I., Mulkey, L.A. 1983. User's Manual for the Chemical




    Transport and Fate Model TOXIWASP. Version  1.  EPA Report No. EPA-




    600/3-83-005. Environmental  Research  Laboratory, Office of Research and




    Development, U.S. Environmental  Protection  Agency, Athens, Georgia. 178





    PP-








 Bell,  G.L.  1980. Lake St.  Clair  and St.  Glair and Detroit Rivers Chemical




    and Physical Characteristics Data for 1974.  NOAA Data Report  ERL GLERL-




    12.  Great Lakes Environmental Research Laboratory,  Ann Arbor,  Michigan.





    10 pp.









 Bukata, R.P., Jerome,  J.H., and Bruton, J.E. 1987. Remote and In Situ




     Optical Studies of Seston and Suspended Sediment Concentrations in Lake




     St^ Clair. Preliminary Report. Rivers Research Branch, National Water




     Research Institute, CCIW, Burlington, Ontario, Canada.









  Burns, L.A., Cline, D.M., andLassiter,  R.R. 1982. Exposure Analysis




     Mode line System (EXAMS^) : Use^ Manual gnd System Documentation. EPA




     Report  No. EPA-600/3-82-023.  Environmental Research  Laboratory, Office




     of Research and Development,  U.S.  Environmental Protection Agency,





     Athens,  Georgia. 145  pp.
                                      407

-------
DiToro, D.M., Fitzpatrick, J.J., Thomann, R.V. 1983. Documentation for Water




    Quality Analysis  Simulation  Program (WASP) and Model Verification




    Program  (MVP).  EPA Report No.  EPA-600/3-81-044.  Environmental Research




    Laboratory, Office of Research and Development,  U.S. Environmental




    Protection Agency, Duluth, Minnesota.  145  pp.









Frank, R. , Holdrinet, M., Braun, H.E.,  Thomas, R.L., Kemp, A.L.W., and




    Jaquet,  J.-M.  1977.  Organochlorine insecticides  and PCBs  in  sediments of




    Lake  St.  Clair (1970 and 1974) and Lake Erie (1971).  Sci.  Tot. Environ.




    8:205-227.









 Great Lakes  Institute. 1986. A Case Study of Selected Toxic  Contaminants  in




    the  Essex Region. Volume I:  Physical Sciences.  Parts One and Two.  Final




    Report.  University of Windsor, Windsor, Ontario, Canada.









 Ibrahim,  K.A. 1986.  Simulation  of Pollutant Transport Responses  to Loading




     and Weather Variations  in Lake St. Clair and the Connecting  Channels.




     PhD Dissertation, Department  of Civil Engineering, University of




     Windsor, Windsor, Ontario,  Canada. 436 pp.









 Kaiser, K.L.E., Comba, M.E.,  Hunter,H., Maguire, R.J., Tkaca, R.J.,  and




      Platford, R.F.   1985.   Trace  organic  contaminants  in the Detroit River.





      J. r.r**r T^kes Res.   11(3) : 386-399 .
                                     408

-------
Kuel, D.W., Leonard, E.N. , Welch, K.J., and Veith, C.D. 1980. Identification




    of hazardous organic  chemicals  in  fish from  the Ashtabula River, Ohio




    and Wabash River,  Indiana. Association of  Official Analytical Chemists




    Journal . 63:1238-1244.









Limno-Tech, Inc. 1985.  Summary of the  Existing Status of  the Upper Great




     Lakes  Connecting Channels Data. Limno-Tech,  Inc. , Ann Arbor, Michigan.




     157 pp.









Mabey, W.R. ,  Smith,  J.H., Podoll , R.T., Johnson, H.L. ,  Mill, T. , Chou,




     T.-W., Gates.  J., Waight Partridge, I.,  Jaber, H. ,  and Vanderberg, D.




     1982.  Aquatic  Fate Process Data for Organic Priority Pollutants.  EPA




     Report No.  EPA- 440/4 -81 -014. Monitoring and Data Support Division,




     Office of Water Regulations  and Standards, Washington, DC.  145 pp.









 Maguire,  R.J.,  Tkaca, R.J.,  and  Sartor, D.L.  1985.  Butyltin species and




     inorganic tin in water and sediment of the  Detroit and St.  Clair Rivers.





     T  fir«at lakes  Res.  11(3) :320-327 .
  Oliver, B.C. 1987.  ^   n«1r Riv-r  sediments. A  level II report for the




     Upper Great Lakes  Connecting Channels Study.  ELI Eco Laboratories Inc.,





     Rockwood,  Ontario,  Canada.
                                      409

-------
Oliver, B.C. and Bourbonniere , R.A.  1985.  Chlorinated contaminants in




    surficial sediments  of Lakes  Huron,  St.  Clair,  and  Erie: implications




    regarding sources  along the  St.  Clair and Detroit Rivers. J. Great Lakes




    Res. ll(3):366-372.









Pugsley, C.W.,  Hebert,  P.D.N., Wood, G.W. , Brotea,  G. ,  andObal, T.W. 1985.




    Distribution  of contaminants in clams and sediments from the Huron-Erie




    corridor.  I-PCBs and octachlorostyrene. J. Great Lakes Res.




    ll(3):275-289.









 Richardson,  W.L.,  Smith, V.E., and Wethington, R. 1983. Dynamic mass balance




    of PCS and suspended solids  in  Saginaw Bay- a case study. In Physical




    Behavior of PCBs in the Great Lakes. Edited by Mackay, D. ,  Paterson, S.,




    Eisenreich, S.J., and  Simmons,  M.S. Ann Arbor Science, Ann Arbor,





    Michigan.  442 pp.









 Robbins, J.A.  and Oliver,  E.G. 1987. Accumulation of fallout cesium-137 and




     chlorinated organic contaminants in recent sediments of Lake St.  Clair.




     Can. J. FJ«h- and Ag^tic Sci.  Submitted.
 Robbins  J A   1985.  Th^  CoupleH  T^kes Mndel  for Estimating the Long-Term




     i,,,^-  nf  t-.h»  Great JLakes_1-o Tii^-nPT^dent  Loadings of Particle-




     AMOCi.ted Contaminants.  NOAA Technical  Memorandum  ERL GLERL-57. Great




     Lakes  Environmental  Research Laboratory, Ann Arbor, Michigan. 41 pp.
                                      410

-------
Roberts, J.R., Rodgers,  D.W.,  Bailey, J.R.,  andRorke, M.A. 1978.




    Polychlorinated  Biphenvls:  Biological Criteria for an Assessment of




    their Effects  on Environmental Quality.  Publication No. NRCC 16077.




    National  Research Council of Canada,  NCR Associate Committee on




    Scientific Criteria for Environmental Quality Ottawa, Canada. 172 pp.









Rybczyk, J.M.  1986.  Cluster Analysis of Physical, Biological,  and Chemical




    Data of Lake St. Clair. Unpublished Data. Great Lakes Environmental




    Research Laboratory, Ann Arbor, Michigan.









 Schwab, D.J.  and Liu, P.C. 1987. Development of a Shallow Water  Numerical




    Wave Model for Lake St. Clair. Upper Great Lakes Connecting  Channels




    Study  Final Report. Great Lakes Environmental Research Laboratory, Ann




    Arbor,  Michigan.









 Simons, T.J. and Schertzer, W.M.  1986. Modelling Wave-Induced Sediment




     R^usoension  in Lake  St.  Clair. Preliminary  Report.  Aquatic  Physics  and




     Systems Division,  National  Water Research  Institute, Burlington,





     Ontario, Canada.









 Thomann,  R.V. and Mueller,  J.A.  1983.  Steady state modeling of toxic




     chemicals-theory and application  to  PCBs in  the  Great Lakes and Saginaw




     Bay. m  m.y-^.1  R*h»vi~ ^ »™«  *" the Great Lakes. Edited by Mackay,




     D., Paterson,  S.,  Eisenreich, S.J.,  and Simmons, M.S. Ann Arbor Science,





     Ann Arbor,  Michigan. 442 pp.
                                      411

-------
Table 1.  Lake St. Clair parameters  and constants  required by the TOXIWASP
model.  Units are consistent with model specifications.  Reference in
parentheses.
Variable
Parameter
TEMP
DEPTH
VELOC
WIND
BACTO
BIOMS
OCS
PCTWA
PH
US
CMPET
Constant
OCB
CLOUDG
LATG
Description

Average segment temperature
Depth of segment
Average water velocity
Average wind velocity
Bacterial population density
Total biomass in segment
Sediment organic carbon content
Percent water in sediments
Hydrogen ion activity
Settling rate in water column
Resuspension rate in bed
Contaminant burial rate
Light extinction coefficient


Biomass organic carbon content
Average cloud cover
Geographic latitude
	 	 	 • 	 ' 	
Units

degrees C
feet
feet s"1
meters s~l
cells ml'l (water)
cells 100 g'1 (bed)
mg 1~1 (water)
g m"2 (bed)
dimensionless
dimensionless
pH units
m dayl
cm yr'l
cm yr'l
m-1

dimens ionless
Value

13(1)
6.6-17.1(2)
0.48(2)
6(3)
106(4)
107-108(4)
10(4)
1-50(4)
0.14-5.0X(5)
1.67-1.71(5)
8.1(1)
3.0(6)
0.45-0.95(5)
0.1(7)
2.0(1)

«(5)
tenths of full cover 4(3)
degrees and tenths

43.2

  ISTORET DATA
  2schwab (1987)
  3GLI (1986)
  ^Ibrahim (1986)
  5This study
  ^Simons and Schertzer (1986)
  Bobbins and Oliver (1987)
                                      412

-------
Table 2.  Octachlorostyrene constants required by the TOXIWASP model.   Units
are consistent with model specifications.  From Ibrahim (1986).


Constant  Description                            Units            Value


ROW       Octanol water partition coefficient    lw !Oct"^       2.48x10**

KOC       Organic carbon partition coefficient   lw kg"1         1.20x10^

MWT       Molecular weight                       g mole'1          300

HEN       Henry's Law  constant                   Atm m^ mole"1     10"^

VAP       Vapor  pressure                         torr             4x10"5

SOL       Aqueous solubility                     mg I'1            0.02
                                    413

-------
Table 3.   Total polychlorinated biphenyl constants required by the TOXIWASP
model.  Units are consistent with model specifications.  From Mabey et. al.
(1982).
Constant
KOW
KOC
MWT
HEN
VAP
SOL
Description
Octanol water partition coefficient
Organic carbon partition coefficient
Molecular weight
Henry's Law constant
Vapor pressure
Aqueous solubility
Units
lw loct
lw kg-1
g mole'l
Atm m^ mole'l
torr
mg l"1
Value1
4.14xl05
2xl05
300
3.9x10-3
5x10 '4
5xlO-2
 IWithin range  of values for Aroclors  1232,  1242,  1248,  1254,  and  1260.  They
 most  closely  resemble values for Aroclor 1248.
                                     414

-------
                               LIST OF FIGURES

Fig. 1.  Lake St. Clair numerical grid used in Lake St. Glair contaminant fate
         and transport model, based on EPA's  TOXIWASP.  Segment numbers
         correspond to water column segments.  Active layer and deep layer
         segment numbers are determined by adding one and two, respectively,
         to the water column numbers.

Fig. 2.  Loading of Cs-137  to  Lake  St. Clair  during the period 1950-1985 from
         three principal sources:  inflow  from Lake Huron, direct  atmospheric
         fallout,  and land  runoff  from the watershed.  Reprinted  from Robbins
         and Oliver  (1987).

Fig. 3.  Observed  distribution and model  initial conditions  of PCBs  (^g/kg)  in
         surface (0-2 cm)  sediments in Lake  St.  Clair,  1970.  Data from Frank
         et.  al.  (1977).

Fig 4.   (a)  Comparison of observed and  model-simulated concentrations  (mg/1)
         of chloride in the water column of Lake St.  Clair  during cruise 3,
          19-29 June 1974.   Wind speed: 4 m/s NE.  (b) Comparison of observed
          and model-simulated water column concentrations (mg/1)  of chloride
          during cruise 5,  15-24 July 1974.  Wind speed:  5 m/s N.   Data from
          Bell (1980).

 Fie  5   (a) Comparison of observed and model-simulated concentrations (mg/1)
          of chloride in the water column of  Lake St. Clair during cruise 6,
          5-15 August 1974.  Wind  speed: 5 m/s E.  (b) Comparison of observed
          and model-simulated water column concentrations (mg/1) of chloride
          during cruise 8,  16-25 September 1974.  Wind speed: 6 m/s W.  Data
          from Bell  (1980).

 Fie   6   (a) Comparison of observed  and model-simulated concentrations  (dpm/g)
       '  of Cs-137  in  surface (0-2  cm)  sediments of  Lake St. Clair  in  1976.
          Model  was  calibrated by  adjusting  the  spatially-varying sediment
          organic  carbon content until the model results matched  the data.   (b)
          Comparison of observed and model-simulated  0-2 cm  concentrations
           fdDm/R)  of Cs-137 in 1985.   The 1985  simulation verified the  1976
          model  results.   Data from Robbins  and Oliver (1987).

  FiE   7   Comparison of observed  and model-simulated concentrations    of
    g'      octachlorostyrene (DCS)  in surface (0-10  cm)  sediments  of Lake St.
           Clair  in 1983   The simulation was run until the  model  results agreed
           wlS the 1983'data,  which occurred at day 4500;  implying that the
           load was first introduced during the latter part  of 1970.   Data from
           Pugsley et.  al.  (1985).

  Fig  8   Comparison of observed and model-simulated concentrations (^g/kg) of
           total PCBs in surface (0-2 cm) sediments of Lake St  Clair in 1974
           Simulation was run for four years  -ing initial conditions presented
           in Figure 3.   Data from Frank et.  al. (1977).
                                      415

-------
                                                              Scale in Kilometers

                                                               I  I  I  I   I  I   I
                                                              048
i
' •
 •
                                                                                                     82°30'

-------
.
• •
                                                Cs Loading to Lake St. Clair
                                                   Watershed
  Watershed * Atmosphere
                                                  j Watershed + Atmosphere
                                                  1 -i-Inflow from Lake Huron
                                   1960
197O
1980
                                          Year

-------
                            Scale in Kilometers

                             I  I  I  I  I
                             0    4
i
i •
0
PCB Concentration 1970
Surficial (0-2 cm) Sediments
Q  Not Sampled

D  <
n?n  5-10
    10-20 /L/g/kg
    20-30 /ug/kg
    >30 /43/kg

Initial Conditions

-------
                           Scale in Kilometers

                           I  I  I  I  I   I  I
                           0   4
i
i •
   Cruise 3
19-29 June 1974
Wind - 4 m/s NE

-------
!

I '
                                                               Cruise 5
                                                            15-24 July 1974
                                                            Wind = 5 m/s N

-------
I
I 1
                                                                   Cruise 6
                                                               5-15 August 1974

                                                                Wind = 5 m/s E
                                                          82°30'
                                                          _JL_

-------
                             Scale In Kilometers
                              II  I  I  I  I  I
i
i i
      Cruise 8
16-25 September 1974
   Wind = 6 m/s W
                                                              82°30'

-------
 6/iudpg<


 6/Ludp Q-


 6/tudp £-


 6/u)dp 2-1.


6/tudp 1.-9 0


6/iudp g o>
sjuauijpas (wo 2-0)

   9Z61
                                                                                                                '• i
                                                                                                                ' i
                                                                                                                 i

-------
                              Scale in Kilometers
                               I  I I  I   I I  I
                              0   4   8   12
I
i
;
137Cs Concentration 1985
Surficial (0-2 cm) Sediments
                                         <0.5 dpm/g
                                         0.5-1 dpm/g
                                         1-2 dpm/g
                                         2-3 dpm/g

                                        Simulation

-------
M  ,
•    I
   r

-------
I I
i
                                                            PCS Concentration 1974

                                                            Surficial (0-2 cm) Sediments
Not Sampled
                                                                                                      5-10 j/g/kg

                                                                                                      10-20 ,ug/kg

                                                                                                      20-30 /jg/kg

                                                                                                    Simulation

-------