Effect of channel changes in the St. Clair River since 1900

[From the U.S. Government Printing Office, www.gpo.gov]

#0 tAT OF C041

NOAA Technical Memorandum ERL GLERL-40

."p4rEs Of

EFFECT OF CHANNEL CHANGES

IN THE ST. CLAIR RIVER SINCE 190

Jan A. Derecki

Great Lakes Environmental Research Laboratory
Ann Arbor, Michigan
February 1982

QH
541.S
.M3
D47
1982

NATIONAL OCEANIC AND Environmental Research
noaa ATMOSPHERIC ADMINISTRATION Laboratories

NOAA Technical Memorandum ERL GLERL-40

EFFECT OF CHANNEL CHANGES
IN THE ST. CLAIR RIVER SINCE 1900

Jan A. Derecki

Great Lakes Environmental Research Laboratory
Ann Arbor, Michigan
February 1982

q.q'q06q2 qAqTqMqOqSqPq"qE UNITED STATES NATIONAL OCEANIC AND Environmental Research
DEPARTMENT OF COMMERCE ATMOSPHERIC ADMINISTRATION Laboratories
LD
Malcolm Baldridge, John V. Byrne, George H. Ludwig
Secretary Administrator Director
ENT Of

NOTICE

Mention of a commercial company or product does not constitute
an endorsement by NOAA Environmental Research Laboratories.
Use for publicity or advertising purposes of information from
this publication concerning proprietary products or the tests
of such products is not authorized.

CONTENTS

Page

Abstract 1

1. INTRODUCTION 1

2. METHOD 4

3. RESULTS 15

4. CONCLUSIONS 17

5. ACKNOWLEDGMENT 19

6. REFERENCES 19

2qi2qi2qi

FIGURES,

Page

1. St. Clair-Detroit River system with location of water
level gages. 5

2. Comparison of 1900 and present areas at the Mouth of
Black River gage section. 9

3. Manning's roughness coefficients for FG-MBR. reach. 11

4. Manning's roughness coefficients for FG-DD reach. 12

5. Manning's roughness coefficients for MBR-SC reach. 13

6. Manning's roughness coefficients for DD-SC reach. 14

7. Upper St. Clair River water surface profile for present
and 1900 channel conditions. 18

TABLES

Page

1. St. Clair River hydraulic parameters for the present channel. 6

2. St. Clair River hydraulic parameters for the 1900 channel. 8

3. St. Clair River Manning's roughness coefficients for present
and 1900 channel conditions. 10

4. Results of computations for upper St. Clair River profile
with present and 1900 channel conditions during average
water levels (1970). 15

5. Results of computations for upper St. Clair River profile
with present and 1900 channel conditions during high water
levels (1973). 16

EFFECT OF CHANNEL CHANGES IN THE ST. CLAIR RIVER SINCE 19001

Jan A. Derecki

Periodic man-made changes in the outlet of Lake Huron
through the St. Clair River date back to the middle of the last
century. These artificial channel changes are well documented
during the present century and consist of dredging for commercial
gravel removal in the upper river during 1908-25 and uncompen-
sated navigation improvements for the 25-ft and 27-ft projects
completed in 1933 and 1962, respectively. The total effect of
these changes on the levels of Lakes Michigan and Huron
(hydraulically one lake) and on the upper St. Clair River pro-
file was determined with dynamic flow models. The ultimate
effect of the above dredging was a permanent lowering of the lake
levels 0.27 m, which represents a tremendous loss of fresh water
resource (32 km3). This total lowering of lake levels is 0.09 m
higher than previous estimates, but present determinations repre-
sent a more sophisticated approach.
It is unfortunate that appropriate hydraulic data for the
previous century are not available for the application of this
method. In the late 1880's the levels of Lake Michigan-Huron
dropped by 0.8 m. In contrast to the other Great Lakes, the
levels of these two lakes remained depressed afterwards,
producing controversy about the reasons for this unexplained
drop, namely, reduced precipitation and/or dredging. Existing
estimates for the lake level drop due to unspecified dredging
prior to 1900 vary from under 0.1 m to over 0.4 m, an unaccep-
tably large difference. Reduction of lake levels due to precipi-
tation or dredging implies different lake level behavior in the
future, which complicates water resource studies based on the
analysis of lake levels and related outflows.

1. INTRODUCTION

The Great Lakes represent a tremendous fresh water resource, and infor-
mation on their levels and outflows is becoming progressively more important
for water resource planning purposes and the operation of the presently
regulated lakes (Superior and Ontario). Currently, this information in-
cludes over 120 years, 1860 to date. Because of their size, the Great Lakes
possess a large self-regulating capacity and the natural fluctuation of
their levels is relatively small. During the 1885-1900 period, the levels
of the Great Lakes below Lake Superior dropped sharply, with the most acute
drops for Lakes Michigan and Huron (0.8 m), which hydrologically are con-
sidered to be one lake. This drop in lake levels was accompanied by a

1GLERL Contribution No. 307.

corresponding reduction of basin precipitation and eventually the levels of
lakes other than Lake Michigan-Huron rebounded to their normal state. The
levels of Lake Michigan-Huron remained depressed below previously estab-
lished normals for reasons that have remained unexplained for several
decades.
The reasons could be either natural, namely, basin precipitation per-
sistently below previous levels; artificial, caused by man-made changes in
the lake outlet or drainage basin conditions; or a combination of both
natural and artificial changes. Artificial changes in the Lake Huron outlet
through the St. Clair River for navigation improvements date back to 1856,
when a channel was cut across sand bars in the St. Clair Flats area of the
lower river to provide a 9-ft draft (International Joint Commission, 1976).
Changes in land use affecting runoff characteristics from the drainage basin
also date back to the mid-1800's. The outflows from the Great Lakes are
based on stage-flow relationships derived from periodic flow measurements
made after 1900. Since navigation improvements and commercial gravel
dredging in critical locations could make the channel more efficient, the
application of the post-1900 St. Clair River stage-flow relationships to the
previous period, as was actually done, would make the 1860-1900 flows arti-
ficially high. On the other hand, present regulation plans for Lakes
Superior and Ontario, based on lake outflows determined for the post-1900
period, would fail under more severe natural conditions.

There is some controversy regarding the reasons for, and corresponding
magnitudes of, the unrecovered drop in the lake levels during the 1880-1900
period. Brunk (1961, 1963, 1968) conducted investigations to determine the
drop of Lake Huron levels using ratios of Great Lakes levels, St. Clair-
Detroit and Niagara River flows, and basin precipitation records. He felt
that previously published regional precipitation (Day, 1926) for Lake Huron
was too high and modified these'precipitation values. Brunk concluded that
most of the drop in Lake Huron levels prior to 1900 was caused by dredging
of the lake outlet in the St. Clair River and that published flows for the
St. Clair-Detroit River prior to the drop (1860-1900) are excessive. The
magnitude of dredging-related lowering of Lake Huron levels prior to 1900
determined by Brunk (1968) is 0.43 m (1.4 ft). Lawhead (1961), in a dis-
cussion of Brunk's (1961) results, concluded that most of the decrease in
Lake Michigan-Huron levels during this period was due to precipitation
changes and that only 0.09 m (0.3 ft) could be attributed to the St. Clair
River channel dredging. Additional lowering of the Lake Michigan-Huron
levels due to dredging after 1900 was not contested in the above studies.
The additional drop in the lake's levels for the post-1900 period was esti-
mated by the International Great Lakes Levels Board (1973) to total 0.18 m
(0-59 ft). This total amount consists of uncompensated dredging for water-
way improvements for the 25-ft and 27-ft navigation channels completed in
1933 and 1962) respectively, and commercial gravel dredging in the upper St.
Clair River in the vicinity of Point Edward, Ont., between 1908 and 1925.
The effect of gravel dredging is estimated by the Levels Board to be about
0.09 m (0.3 ft), which leaves about half (0.09 m or 0.29 ft) of the total
lowering during this period to be attributed to uncompensated navigation
dredging in the St. Clair River. Navigation dredging in the Detroit River
was compensated by dikes.

The unresolved Brunk-Lawhead controversy over the pre-1900 drop of Lake
Michigan-Huron levels complicates water resource studies involving lake
level analysis. If the late 1800's levels were high because of above-normal
precipitation, such high levels may occur again in the future; if the pre-
vious high levels resulted from a less efficient outflow channel, it is
unlikely they will be repeated. This problem was addressed by Quinn and
Croley (1981) in a study based on precipitation climatology, which used more
recently determined precipitation data. Quinn and Croley concluded that
dredging prior to 1900 (1893-99) may be responsible for about 25 percent
(0.2 m) of the total lake level drop of 0.8 m during the late 1800's (1886-
92) and that published St. Clair River flows for the 1860-1900 period are
excessive. Most of the drop (0.6 m) was apparently caused by a long term
change in precipitation. As in previous studies, their determinations
involved assumptions and some rather weak input parameters, such as avail-
able rainfall/runoff regressions with low correlation coefficients (0.26-
0.85). The accuracy of all these lake-level-drop determinations may be
classified as relative, rather than absolute, and require further verifica-
tion.

The present study was initiated to provide verification for both the
unexplained drop in Lake Michigan-Huron levels prior to 1900 and for the
uncompensated dredging effects after 1900, using the St. Clair River dynamic
flow models. Although there is generally no controversy regarding these
later uncompensated dredging effects, with sufficient documentation (Inter-
national Joint Commission, 1976; International Great Lakes Levels Board,
1973; U.S. Senate, 1955; Joint Board of Engineers, 1927), the use of flow
models for this purpose represents a more sophisticated approach than those
employed in previous estimates. The requirements for such a verification
study with the flow models are the channel cross-sectional areas and channel
roughness coefficients for the appropriate periods. Determination of chan-
nel roughness coefficients requires river flow measurements. Historic
hydrographic surveys and flow measurements for the Great Lakes were made by
the U.S. Lake Survey, a former Corps of Engineers District. Archive records
for the hydrographic surveys are presently maintained by the National Ocean
Survey, NOAA, while those for flow measurements are stored by the Detroit
District, Corps of Engineers. The earliest hydrographic survey of the St.
Clair River was conducted during 1867, providing cross-sectional areas prior
to any significant channel changes, while a second survey in 1900 provided
cross-sectional areas after the unexplained drop in Lake Michigan-Huron
levels and before the known uncompensated dredging during this century (25-
and 27-ft channels and gravel removal). Channel roughness for the second
period was determined from flow measurements conducted during 1908-10. Flow
measurements for the preceding period are not available. Although it was
possible to estimate channel roughness coefficients that appear reasonable
for this period, this was immaterial since analysis of the 1867 cross-
sectional areas showed them to be grossly inadequate. Either the measure-
ments were crude and inaccurate or the available small-scale field sheets
did not permit reproduction of the areas with sufficient accuracy. In
either case, this eliminated the first objective of the study concerning the
channel changes during the previous century. Consequently, only the uncom-
pensated channel changes during the present century are evaluated.

2. METHOD

The Great Lakes Environmental Research Laboratory (GLERL) dynamic flow
models for the St. Clair River are described by Derecki and Kelley (1981).
These models are one-dimensional transient flow models based on equations of
continuity and momentum, with option for the surface wind stress effects.
Disregarding wind stress effects, not used in this study, the equations of
continuity and momentum are expressed in terms of flow and stage

N + -1 bQ = 0
Tt T 6X

and
I 6Q - 2 21 6Z + @Z + gn2 Q/Q/ oil (2)
2 bt _3
A bt A (g A3 ax 2.208 R47

where Q = flow rate,
Z = stage above fixed datum,
X = distance in the positive flow direction,
t = time,
A = channel cross-sectional area,
T = top width of channel,
g = acceleration due to gravity,
R = hydraulic radius, and
n = Manning's roughness coefficient.

The model solution uses an implicit finite-difference method with
Newton-Raphson iterative algorithms for initiating the computations, which
can be operated with variable time steps. Several versions of the models
incorporating different river reaches are all confined to the upper one-
third of the river, between a gage at Fort Gratiot, Mich., at the head of
the river and a gage at St. Clair, Mich., 23 km downstream (fig. 1). This
portion of the river contains most of the river slope and is usually free of
ice concentrations during winter. The model programs are written in a
generalized manner and can be easily modified or adapted to other rivers by
appropriate substitution of physical characteristics (cross-sectional areas,
channel widths, roughness coefficients, etc.) and boundary conditions (down-
stream and upstream controls). Model computations incorporate detailed
channel definition to indicate the actual river channel (table I).

For the purpose of this study, it was necessary to use a model that
would cover the entire upper river reach between Fort Gratiot and St. Clair,
two of the oldest river gages in the system. Since none of the existing
operational models covered this reach, they were modified to obtain two
desired models, each comprising upper and lower reaches with "mid-point"
locations at the Mouth of Black River and Dry Dock gages. For future

LAKE HURON

Ft. Gratiot
Dunn Paper Pt. Edward
Mouth of Black River..
Dry Dock..
Marysville

St. Clair
Michigan

SCALE IN MILES
Roberts
0 10 20 anding
KILOMETERS Port Lambton
1!0@ 1---l Igon ac
0 10 20 30

St. Clair Shores LAKE
Grosse Point
ST CLAIR
Windmill Point

'R @cbmse
Ft. Wayne
Be iver.

La Salle
Wyandotte':
Ontario
FZ
LU

Amherstburg
Gibralter

LAKE ERIE

FIGURE 1.--St. Clair-Detroit River system with location of water Level gages.

TABLE l.--St. Clair River hydraulic parameters for the present channel

Gage location Station Width Length Azimuth Reference Base area
(ft) (ft) (ft) (0) elevation (ft2)
IGLD (1955)*

Fort Gratiot 207 31970 1,800 -- 30 576.7 57,500
207,640 1,320 330 30 576.5 459800
Dunn Paper 2079090 1,000 550 30 576.4 40,800
206.%790 1,000 300 30 576.4 33,100
206,350 1,000 440 30 576.3 35,000
2063,030 920 320 30 576.3 349700
205,320 880 710 30 576.1 28,800
2053%030 940 290 3 576.1 32,100
204,600 1,000 430 3 576.1 33$400
204$280 1,220 320 3 576.1 44,000
203.%970 1,360 310 3 576.1 492700
202.%920 10480 1,050 3 576.1 55,200
202J.570 1,520 350 161 576.1 65,600
202,140 1,480 430 161 576.1 649900
2009840 19400 1,300 161 576.0 48,200
2003,530 1,320 310 161 576.0 47,300
1999520 1,360 1,010 143 575.9 53,100
1999240 1,360 280 143 575.9 50,200
1979790 1,620 1,450 143 575.8 492300
Mouth of Black 196,410 2,590 1,380 14 575.8 67,800
River 1959410 2,630 1,000 14 575.8 76,000
193)480 2,500 1,930 14 575.7 76,000
190*400 1,840 3,080 31 575.7 502700
Dry Dock 182,480 2,180 7,920 44 575.4 58,800
1709920 1,890 11,560 14 575.1 57,100
Marysville 1662980 23,250 3,940 14 574.9 68,400
1669480 23,400 500 18 574.9 683%300
165,930 2)630 550 18 574.9 64,400
1632380 3,490 2,550 18 574.9 70$700
162,810 3,290 570 18 574.9 71,600
161$350 2)660 1,460 18 574.8 64,300
155V470 2jo640 5,880 177 574.7 62,600
1511480 3,120 3,990 177 574.7 75,600
148,430 2,420 3,050 10 574.5 65p9OO
145,980 1,840 2,450 10 574.4 54,600
144,970 1,960 1,010 10 574.4 61,500
135)330 3)080 9,640 8 574.2 77,800
134,290 22760 1,040 8 574.1 65,600
St. Clair 132,270 23,280 2,020 8 574.1 66,300

*IGLD--International Great Lakes Datum. Data in this table are listed in
English units since all computations are done in English units and the
final results listed in either the English or the SI system.

reference, these models are identified by a three-gage system as FG-MBR-SC
and FG-DD-SC.

The procedure employed in determining channel changes involved com-
puting river flows with the present channel configuration, using current
water level gage data, then matching these flows with the previous channel
configuration modified by appropriate cross-sectional areas and channel
roughness coefficients. The difference in water levels for the same flow
with present and previous channel configurations represents the effect of
channel changes due to dredging. The ultimate effect of channel changes on
the water levels should be nearly the same during periods of low or high
water supply. This is demonstrated by computing the channel-change effects
for 1970, a year of mid-range or average water levels, and for 1973, a year
of high water levels. To eliminate possible ice effects, the computations
were limited to the open-water season and, furthermore, were restricted to
the June-August period, which represents annual peak water levels.

The channel cross-sectional areas prior to the uncompensated channel
changes during this century were determined from the 1900 hydrographic sur-
vey and used to evaluate dredging effects by both models. The 1867 cross-
sectional areas were similarly determined, but as mentioned previously, they
were found to be inadequate, producing illogical results, and so were dis-
carded. The upper St. Clair River hydraulic parameters for the 1900 channel
are given in table 2. Basic input data are listed in English units since
all computations are done in English units and the final results listed in
either English or SI units. Comparison of the 1900 and present areas at the
section corresponding to the Mouth of Black River gage location is shown in
figure 2. As indicated, the present channel at this section has a nearly
uniform depth as a result of substantial dredging over most of the width.
It is apparent that at least final stages of this dredging were connected
with the 27-ft (8.2-m) navigation project completed in 1962. The present
navigation channel at this location covers slightly over half of the river
on the United States or western side. It is a common practice to provide
approximately 2-ft (0.6-m) overdraft in deepening the navigation channels.
During both the 25-ft and the 27-ft projects, dredged material was deposited
in river areas where it would not interfere with navigation to partially
offset some of the effects on upstream water levels. This explains the
filling of the deeper portion of the river along the eastern bank. Thus,
present river depth in this location is approximately 8.8 m (29 ft), with
the exception of a reduction in depth to 7.9 m (26 ft) along the eastern
boundary of the navigation channel and the overbanks. The assumption of a
2-ft overdraft is verified by the average depth of the present channel at
this section, which is also about 8.8 m (29 ft). This compares with a value
under 7.6 m (25 ft) in 1900, giving an 18-percent increase in the cross-
sectional area for the present period.

Calibration of the models for both periods consisted of computing
roughness coefficients for each reach of the river bounded by water level
gages. The roughness coefficients were determined from Manning's formula

TABLE 2.--st. Clair River hydraulic parameters for the 1900 channel

Gage location Station Width Reference Base area
(ft) (ft) elevation (ft2)
IGLD (1955)*

Fort Gratiot 207,970 2,100 575.52 65,720
207,640 1,600 575.52 55,680
Dunn Paper 207,090 1*040 575.52 43,045
206,790 960 575.52 40,700
206,350 920 575.52 35,115
206,030 880 575.52 33,678
205,320 780 575.52 34,460
205,030 800 575.52 33,555
204,600 19000 575.52 37,965
204,280 1,150 575.52 38,395
203,970 19250 575.52 41,855
202,920 1,420 575.52 449163
202,570 1,480 575-52 45,155
202,140 1,420 575.52 45,805
200,840 1,200 575-52 41,100
200,530 1,300 575-52 45,850
199,520 1,100 575-52 42S915
199,240 1,150 575-52 42,970
197,790 1,580 575.52 49,245
Mouth of Black 196,410 2,600 575.52 67,430
River -195,410 2,600 575.52 63,720
193,480 2)400 575.52 57,430
190,400 1,800 575.52 51,525
Dry Dock 182,480 29000 575.52 57,665
170,920 12840 575-52 609155
Marysville 166,980 22200 575.52 68,800
166,480 2*400 575-52 68,720
165,930 23,650 575.52 -66,860
163,380 2,950 575-52 68,980
162,810 2,860 575.52 68,225
161,350 2,530 575-52 70,045
155,470 2$820 575.52 70,600
151,480 2JI700 575.52 74,555
148,430 1,950 575.52 63,630
145,980 13,920 575.52 66,530
144,970 2.%020 575.52 68,455
135,330 2,900 575-52 80,980
2,480 575-52 75,700
St. Clair 132,270 2,050 575.52 70,380

*IGLD-International Great Lakes Datum. Data in this table are listed in
English units since all computations are done in English units and the
final results listed in either the English or the SI system.

10 Shaded Area =- 63 720 ftz
0

00
20 o oo
r Navigation channel
30
Approximate present depth

40-
Comparison of Areas at 575.8 ft datum:
Present Area = 76,000 ft2
50- 1900 Area = 64,400 ft2
- Area Difference = 11,600 ft2
60- Percent IncreasO = 18%

70 1 1 1
0 500 1000 1500 2000 2500
Feet From Western Shore

FIGURE 2.--Comparison of 1900 and present areas at the Mouth of Black River
gage section.

1.486 A R2/3 Zu - zd + Q2AA 1/2
n Q L 32.2 L A3 (3)

where n-Manning's roughness coefficient,
A=mean channel area,
R=hydraulic radius,
Q=flow rate,
zu= water surface at upstream gage,
Zd= water surface at downstream gage,
AA= change in channel area between gages, and
L = length of channel reach between gages.

The roughness coefficients for the present channel are based on 14 sets
of flow measurements taken by the Corps of Engineers during 1959-77. For
the 1900 channel, seven sets of flow measurements made during 1908-10 were
used. Although commercial gravel dredging in the upper St. Clair River
started in 1908, there are no indications that channel changes in the first

Present
n = 0.003506(FG)
581- 73 73 -0.17218

580-

579 -
77 68 08 10 08
0 10 10 A
4-0 Discarded A
cu 578- 68
I- A 09
0 62 09
0 6;* 60
U- 577 - e59
4--j
Ca 66* 59
W
Cm
CU 63*
-&-a
U) 576 - 64,
1900
64 n 0.0003506(FG)
575 -0-16810

574' 1 1 1
0.028 0.030 0.032 0.034 0.036 0.038
Roughness Coefficient (n)

FIGURE 3---manning's roughness coefficients for FG-ABR reach.

581 - A...-Present
73 n = 0.0002037(FG)
73 -0.09053

580-

579-
77 68 10 08
0 0 0 A10 AN&08
Disregarded 68*10AA
20 578- 09
0 62 09
0@0- 0
0
LL 62
.6.4 577- 59 059
CU
W 6E@
0) a 0
CO V%j
576- 64o 1900
64 n 0.0002037(FG)
575- -0.08803

574 1 1 1 1 1
0.024 0.026 0.028 0.030 0.032 0.034
Roughness Coefficient (n)

FIGURE 4.-Manning's roughness coefficients for FG-DD reach.

579 -*.--Present
73 n = 0.0236
* 0
578- 73 APre-project
n = 0.0245

577-

77
68
576 - Disregarded 9. 10
10 68 & 10
A 08 09
U) Disregarded 60 09 08
;P62
575- 5962
CY)
Ca 66
0 59
574- 6 3@ 1900
064
*641 n 0.0263
573-

572,
0.022 0.024 0.026 0.028 0.030
Roughness Coefficient (n)
0

FIGURE 5--manning's roughness coefficients for ABR-SC reach.

579- ,.,.Present
73 n = 0.0240

578- 73
1.*@Pre-project
n 0.0252

.---.577-

77
0 68 1
576- Disregarded *6
68 08 10 A10
A A 0
Cl) 0
I-a
CO 60 62 09 08
575- 00
0 62
Cm 59r
CO
.&-I
CO 660 5
631
574- 0
t4 '4@1900
0 n = 0.0274
573- 64

5721 L--
0.022 0.024 0.026 0.028 0.030
Roughness Coefficient (n)
I

FIGURE 6.--Manning's roughness coefficients for DD-SC reach.

few years were significant and these measurements should provide satisfac-
tory indication of channel roughness conditions for the 1900 period. The
relationships between computed roughness coefficients for channel reaches
along the upper St. Clair River and the river stages at adjacent water level
gages for the FG-MBR, FG-DD, MBR-SC, and DD-SC reaches are shown in figures
3-6, respectively. Best-fit relationships were derived for each reach by
regression analysis (least squares) or graphic plots (means), as shown in
the figures. Some points were omitted in this derivation to eliminate
possible gage errors or questionable flow values. Since the 1908-10 flow
measurements were made during similar river stages, they give no indication
of the variation in roughness with depth. The 1900 best-fit lines for the
upstream reaches were estimated using mean values for the 1908-10 measure-
ments and slopes from the present channel. The relationships for downstream
reaches during the 1959-77 flow measurements were affected by regimen
changes associated with dredging for navigation improvements. For these
reaches, separate best-fit roughness coefficients were derived for each
regime, representing 1900, pre-project (through 1963), and present (starting
in 1964) conditions. The calibrated roughness coefficients for the four
reaches are summarized in table 3.

TABLE 3.--St. Clair River Manning's roughness coefficients for present and
1900 channel conditions

Reach Channel Flow Roughness coefficient (n)
measurements

FG-MBR Present 1959-77 n = 0.0003506 (FG) - 0.17218
1900 1908-10 n = 0.0003506 (FG) - 0.16810

FG-DD Present 1959-77 n = 0.0002037 (FG) - 0.09053
1900 1908-10 n = 0.0002037 (FG) - 0.08803

MBR-SC Present 1964-77 n = 0.0236 (starting 1964)
Pre-project 1959-63 n = 0.0245 (through 1963)
1900 1908-10 n = 0.0263

DD-SC Present 1964-77 n = 0.0240 (starting 1964)
Pre-project 1959-63 n = 0.0252 (through 1963)
1900 1908-10 n = 0.0274

Gages: FG = Fort Gratiot
MBR = Mouth of Black River
DD = Dry Dock
SC = St. Clair.

3. RESULTS

Results for the effects of channel changes in the upper and total St.
Clair River by dredging during the present century (since 1900) for commer-
cial gravel removal (1908-25) and navigation improvements (1933 and 1962)
are presented in tables 4 and 5. These tables show June-August average
values for an average water level year (1970) and a high water level year
(1973), respectively, given by two dynamic flow models developed specifi-
cally for this purpose. Results from the two models agree closely, with a
maximum variation of 0.01 m, which is well within limits of expected
accuracy. Flow measurement accuracy for the Great Lakes connecting channels
is generally considered to be 2 percent, which is about 100 m3 s-1 for the
normal St. Clair River range of flows and is equivalent to about a 0.03-m
difference in head or water levels. These values represent zero com-
putational errors and may be doubled for acceptable errors. The agreement
for the 2 years is also very good, with maximum deviations at Fort Gratiot
of 0.01 m, showing that the effect of channel changes on water levels is
nearly the same, regardless of water supply conditions. The effect of

TABLE 4.--Resutts of contputations for upper St. Clair River profile with
present and 1900 channel conditions during average water Levels (1970)

Elevation in meters Dredging effects (m)
Flow River Present 1900 Channel- Upper Total
Model M3 s-1 gages channel upper total river river

FG-MBR-SC 6014 FG 176.49 176.67 176.76 -0.18 -0.27
MBR 176.27 176.35 176.44 -0.08 -0.17
SC 175.75 175.75 175.87 0 -0.12

FG-DD-SC 6023 FG 176.49 176.66 175.75 -0.17 -0.26
DD 176.15 176.21 176.31 -0.06 -0.16
SC 175.75 175.75 175.87 0 -0.12

Combined 6019 FG 176.49 176.66 176.75 -0.17 -0.26
MBR 176.27 176.35 176.44 -0.08 -0.17
DD 176.15 176.21 176.31 -0.06 -0.16
SC 175.75 175.75 175.87 0 -0.12

Gages: FG = Fort Gratiot
MBR = Mouth of Black River
DD = Dry Dock
SC = St. Clair.

TABLE 5.--Resutts of computations for upper St. Clair River profile with
present and 1900 channel conditions during high water Levels (1973).

Elevation in meters Dredging effects (m)
Flow River Present 1900 Channel Upper Total
Model M3 s-1 gages channel 'upper total river river

FG-MBR-SC 6567 FG 176.99 177.18 177.27 -0.19 -0.28
MBR 176.75 176.83 176.93 -0.08 -0.18
SC 176.23 176.23 176.35 0 -0.12

FG-DD-SC 6586 FG 176.99 177.17 177.26 -0.18 -0.27
DD 176.63 176.69 176.79 -0.06 -0.16
SC 176.23 176.23 176.35 0 -0.12

Combined 6577 FG 176.99 177.17 177.26 -0.18 -0.27
MBR 176.75 176.83 176.93 -0.08 -0.18
DD 176.63 176.69 176.79 -0.06 -0.16
SC 176.23 176.23 176.35 0 -0.12

Gages: FG = Fort Gratiot
MBR = Mouth of Black River
DD = Dry Dock
SC = St. Clair.

dredging in the upper St. Clair River on the levels of Lake Huron, indicated
by the Fort Gratiot gage at the head of the river, is a lowering of lake
levels by 0.18 m. Computed effects by the FG-MBR-SC and FG-DD-SC models
vary, respectively, from 0.18 m to 0.17 m for 1970 and from 0.19 m to 0.18 m
for 1973. These effects at the Mouth of Black River and Dry Dock gages,
about 4-km and 8-km downstream, respectively, are reduced to a lowering of
river stages by 0.08 m and 0.06 m.

The above determinations for the upper river dredging effects on lake
Huron levels agree well with previous total estimates published by the
International Great Lakes Levels Board (1973). The Board lists the overall
effect for the total river as 0.59 ft (0.18 m), about half of which or 0.3
ft (0.09 m) is attributed to commercial gravel removal and 0.29 ft (0.09 m),
to uncompensated lowering of lake levels by the 25-ft and 27-ft navigation
projects. The uncompensated dredging in the lower St. Clair River, espe-
cially the construction of the Cutoff Channel in the St. Clair Flats area
for the 27-ft project, is bound to have some negative effect on the levels
of the upper river and Lake Huron. This is verified in a study conducted by
the U.S. Lake Survey (1961) which, although indicating lower overall effect,

with similar amounts for the two navigation projects, shows that while the
effects of the 25-ft project are restricted mostly to the upper river, those
associated with the 27-ft project occur mainly in the mouth of the river.
The total dredging effect published by the Levels Board appears, therefore,
to be substantially underestimated.

The lower St. Clair River is below the physical limits of the available
models and the dredging effects in this reach of the river have to be
supplied as a model input before the total effects can be computed with the
models. This value was determined from a gage relationship based on avail-
able data that shows that the water level at the St. Clair gage was about
0.12-m (0.4-ft) higher during 1900. With this input, the total dredging
effects were recomputed and show about 0.09-m (0.30-ft) additional drop in
Lake Huron levels due to dredging in the lower river. The ultimate effect
of dredging in the entire St. Clair River since 1900 for gravel removal and
the two navigation projects is a lowering of lake levels (Fort Gratiot) by
0.27 m (0.88 ft). The amount of lowering is reduced downstream to 0.18 m
and 0.16 m at the Mouth of Black River and Dry Dock gages, respectively.
Maximum deviations due to model accuracy or water supply conditions are 0.01
M, as indicated previously for the upper river dredging effects. The upper
St. Clair River water surface profile for the present and 1900 channel con-
ditions computed for both years are shown in figure 7. The profiles are
nearly identical despite large differences in flows and water levels.

It is regrettable that a more detailed 1867 hydrographic survey is not
available for more precise determination of cross-sectional areas of the St.
Clair River channel during that period. Present determinations for the
post-1900 period indicate that the flow model method employed would be very
useful in resolving the controversy about the causes and respective magni-
tudes of the Lake Huron drop in water levels before 1900. Existing esti-
mates for the dredging effects for that period vary from 0.43 m (Brunk,
1968) to 0.09 m (Lawhead, 1961), with the most recent estimate of 0.2 m by
Quinn and Croley (1981). If the last estimate is correct, the levels of
Lakes Michigan and Huron were lowered permanently by roughly similar amounts
during both the present and the previous centuries, with the total artifi-
cial lowering due to dredging amounting to nearly half a meter (about 0.47
m). This depth superimposed on the combined area of the lakes (117,400 km2)
represents a volume of 55 km3, a tremendous amount of permanently lost water
resource. The loss exceeds approximately 16 times the volume of Lake St.
Clair (3.4 km3), which is a large inland body of water by any standards but
those of the Great Lakes proper.

4. CONCLUSIONS

Artificial channel changes in the St. Clair River since 1900 include
dredging for commercial gravel removal between 1908 and 1925 and uncompen-
sated navigation improvements for the 25-ft and 27-ft projects completed in
1933 and 1962, respectively. These channel changes increased the efficiency
of the Lake Michigan-Huron outlet through the St. Clair River and caused
permanent lowering of the lake's levels. The total effect of these man-made
channel changes is the lowering of the levels of Lake Michigan-Huron by 0.27

177.4-

177.2

177.04-

Control Conditions
176.8 1973: Flow = 6,577 M3 S-1
St. Clair 176.35 m

176.6
Gq/
E
cl) 176.4-
(D
CD
CO f@ffect from
.&.a * lower river
Cr)
" 176.2 -
> 1970: Flow 6,019 M3 S-1
iz 176.0 - ? & N,,@ St. Clair = 175.87 rn

Effect from
175.8 - lower river
T(Q.12 m)

Ae
175.6 -
.0
)k C,
Od"', 00
,@O loci
175.41 if.I T
0 5 10 15 20 25 30
River Distance (km)

FIGURE 7---Upper St. Clair River water surface profile for present and 1900
channel conditions.

m. This depth superimposed on the combined area of Lakes Michigan and Huron
represents a permanent water loss of 32 km3, which is more than nine times
greater than the volume of Lake St. Clair.

5. ACKNOWLEDGMENT

The author thanks Dr. F. H. Quinn of GLERL for the suggestion to conduct
this study.

6. REFERENCES

Brunk, J. W. (1961): Changes in the levels of Lakes Michigan and Huron.
c7. Geophys. Res. 66(10):3329-3335.

Brunk, J. W. (1963): Additional evidence of lowering of Lake Michigan-
Huron, Pub. No. 10, pp. 191-203, University of Michigan, Great Lakes
Res. Div., Ann Arbor, Mich.

Brunk, J. W. (1968): Evaluation of channel changes in St. Clair and
Detroit River. Water Resour. Res. 4(6):1335-1346.

Day, P. C. (1926): Precipitation in the drainage area of the Great Lakes,
1875-1924. Mon. Wea. Rev. 54(3):85-106.

Derecki, J. A., and Kelley, R. N. (1981): Improved St. Clair River dynamic
flow models and comparison analysis, NOAA Tech. Memo. ERL GLERL-34,
National Technical Information Service, Springfield, Va. 22151. 36
pp.

International Great Lakes Levels Board (1973): Regulation of Great Lakes
water levels, Rept. to the International Joint Commission, pp. 43-47,
IGLLB, Ottawa, Ont.-Chicago, Ill.

International Joint Commission (1976): Further regulation of the Great
Lakes, IJC Rept. to the governments of Canada and United States, pp.
21-23, IJC, Windsor, Ont.

Joint Board of Engineers (1927): St. Lawrence Waterway, Rept. of JBE to
governments of Canada and United States, U.S. Government Printing
Office, Washington, D.C.

Lawhead, H. F. (1961): Discussion--Changes in the levels of Lakes Michigan
and Huron. c7. Geophys. Res. 66(12):4324-4329.

Quinn, F. H., and Croley, T. E., 11 (1981): The role of precipitation cli-
matology in hydrologic design and planning on the Laurentian Great
Lakes, in Fourth Conf. on Hydrometeorol., pp. 7-11, American
Meteorological Society, Boston, Mass.

United States Lake Survey (1961): Hydraulic design memorandum, Great Lakes
connecting channels, effect of and compensation for deepening of the
St. Clair River for 25-foot and 27-foot projects, Rept. File No.
3-3898, U.S. Army Corps of Engineers, Lake Survey District, Detroit,
Mich.

United States Senate (1955): Senate Document 71, 84th Congress, lst
Session, U.S. Government Printing Office, Washington, D.C.

j - ---,E

3 6668 14109 6539