3ifl$§lfifi$§£§ififlfifljfififli>SfflfiikiflflmiTHE
SSEWHEASfifififiE§§fi§iflFEJflGEE%CHM¥§i

magmas EN WATER LEVEL Am mmmm ’ '
m was LAKE wwzw am

Rfi$amnh5%5%W€%£fhsfikfiyessfi‘fiiiL
fifififlfiflw€83éTEififivafififiY
fiMfiiflfiflEfi.BfiCfilfiR

E973 '

IIIIIIIIIIIIIIIIIIIIIIIIIII

llllllUllLllflllNEIHHHHIIMUWWIIIHHIIIIHW

9. 10569 0469

 

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINE retum on or before date due.
MAY BE RECAU£D with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

moo museum

 

 

BLUFF EROSION
AT
SELECTED SITES
ALONG THE
SOUTHEASTERN SHORE
OF
LAKE MICHIGAN

By
William Rdeuckler

A RESEARCH PAPER
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF ARTS
Department of Geography

1973

ABSTRACT

Based on comparison of the original land survey data of the early
l800's as compared to resurveys in 1956 and I973 it is apparent that
bluff erosion rates are not uniform at selected sites along the south-
eastern shore of Lake Michigan nor can they be anticipated to be similar
at very similar sites during two or more distinct time periods. The
factors of bluff composition, height, slope, vegetative cover, beach
frontage, hydrological conditions, shoreline orientation and man-made
structures in the shore zone were observed at eight study sites in order
to determine if any of these variables might be related to erosion
rates. Beaches fronting the shoreline bluffs provide the best pro-
tection against erosion. Bluff composition, sl0pe and hydrological
conditions all are involved in modes of bluff slope retreat. Till
bluffs appear to recede through the processes of slumping, block gliding,
and fluvial activity whereas sandy bluffs appear to retreat through
the process of debris sliding. Major cyclonic disturbances, especially
in November and March, in conjunction with high levels provide poten-
tially favorable conditions for bluff erosion. Man-made shore zone
structures may cause an escalation in bluff recession instead of pre-
venting shoreline erosion. A case study revealed three severe erosion
periods since l9l6 in addition to illustrating the dynamic changes pos-

sible within a shore zone.

 

ACKNOWLEDGEMENT

The author wishes to acknowledge appreciation for the
instructive guidance of Dr. Harold A. Winters in completing
this paper. Appreciation is also extended to Mr. William J.
Gibbs, Jr. for his friendly cooperation in making his
records and photographs freely available. Special thanks
is extended to Mrs. Susan Buckler and Mrs. Claudia Lusch
for typing the various manuscripts and Mr. Richard L. Rieck
and Mr. David P. Lusch for needed assistance in the survey

work.

TABLE OF CONTENTS

Page
LIST OF TABLES ..... .... ............................... v
LIST OF FIGURES........... ...... ...... ...... .......... vi
INTRODUCTION ................ . ......................... 1
INVESTIGATIVE OBJECTIVES........ ........ . ............. 2
PREVIOUS STUDIES... ................................... 2
STUDY LOCATIONS.............. ..... .... ......... . ...... A
VARIABLES UNDER CONSIDERATION.. ....................... 6
INVESTIGATIVE PROCEDURES.. ............................ 7
SITE DESCRIPTIONS.................. ......... . ......... 9
CASE STUDY........... ...... . ...... . ................... 10
Case study locatio
Shoreline evolution
Discussion
Conclusion
BLUFF RETREAT - RATES, FACTORS AND MODES..... ......... 31
Comparison of average annual bluff erosion
rates prior to, and since, 1956
Selected variables investigated in relation
to bluff erosion rates
Modes of bluff slope retreat
Erosion potential as related to storms,
time of year, and high water levels
CONCLUSIONOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO ..... .0... 50
LIST OF REFERENCES.......... ......................... 52
APPENDIX ....... .. ........... . ................... ..... 5%

iv

LIST OF TABLES

Table Page

1. Average retreat of upper bluff-line along
the Lake Michigan shoreline property of
William J. Gibbs, Jr. during each of the
severe erosion periods since 1916......... ..... 13

2. Recorded measurements along the north and
south boundary lines of the William J. Gibbs,
Jr. property............. ............ .......... 15

3. Location of study sites and their total
bluff-line retreat since the original U.S.
land surveys were completed. ................... 32

A. Description of the study site variables. ...... . 33

LIST OF FIGURES

Figure Page
1. Location of study sites.... ..................... 5

2. The present shoreline bluff at the case
Study locatiOHOOOOO ....... 0...... ......... 0.0... 12

3. Good recovery of the beach and bluff after _
the erosion period of 1928 to 1931.... .......... 1o

H. Bluff badly eroded by the spring of 19H9
during the erosion period of 19H6 to 1953....... 17

5. View of the bluff on December 3, 1950,
following the devastating storms of late
November...... ...... . ............... . ......... .. 19

6. Demolished seawall and groins, built two
years earlier, following the intense storms
of late November, 1950...... .................... 2O

7. Reestablishment of vegetation on the bluff
sloge during the low water stage, 195% to
196 0.0000000000000000000000 OOOOOOOOOOOOOOOOOOOO 21

8. Sequence of nine photos showing the pro-
gressive damage done to the bluff in a
two year period as the Lake Michigan level
rose during the last erosion period ......... .... 22

vi

INTRODUCTION

The earth's landscape is continuously being modified
by natural forces. This fact is especially apparent to
many during periods of high water levels on the Great
Lakes. Storm systems over the Lakes, especially in fall
and spring, generate substantial waves with high energy
potential that may vigorously erode the shoreline. Beaches,
which are extensive during low water levels and which act
as energy absorbers to protect coastal uplands, may be
drastically narrowed or even disappear. As a result, waves
may be able to reach and erode the landface.

Since 1969 Lake Michigan water levels have been higher
than average. The 1972 annual level measured over one foot
above the mean while in 1973 the water height has been over
1.5 feet above the average (U. S. Department of Commerce,
1973). This is contrasted to the 1956 to 1968 period when
levels were below the long-term mean. For example, in 196%
the lake registered a level 1.82 feet below the norm.

Much of the land adjacent to the Lake Michigan coastline
consists of unconsolidated, easily erodible, glacial drift.
In many places along the coast are bluffs, 10 to 150 feet

in height, which may be experiencing rapid retreat due,

in part, to their glacial drift composition.1

INVESTIGATIVE OBJECTIVES

The objectives of this study are: (1) to assess rates
of bluff retreat at several selected sites along the south-
eastern shore of Lake Michigan, comparing their present
positions with their positions in 1956 and at the time the
original land surveys were conducted in the 1830's; (2) to
provide information from a case study which reveals three
severe erosion periods since 1916; (3) to determine relation—
ships of selected variables to rates of bluff retreat; and

(A) to ascertain and describe modes of bluff slope retreat.

PREVIOUS STUDIES

"There is an overall lack of systematic field data on
coastal erosion in the Great Lakes as a whole and Lake
Michigan in particular" (Davis, 1972, p. 1). This short-
coming is especially acute for the shore zone along the
eastern part of the lake. The earliest recorded measure-
ments of erosion along this portion were cited by Leverett
(1899, p. H58) for eight points along the shoreline of

Berrien County, Michigan. He published results of

 

1This is not to say, however, that all bluffs of
glacial drift are experiencing substantial erosion.

measurements made by a surveyor named Galvin which indicated
an average annual bluff loss of 3.30 feet during the #1 to
57 years following 1828 (Powers, 1958, p. 87). The Corps
of Engineers (U. S. House of Representatives, 1958, p. 17)
estimated that the average rate of bluff erosion along the
shoreline segment between St. Joseph and the village of
Shoreham in Berrien County was 2.1 feet per year prior to
1958.2
Recently the Corps of Engineers (U.S. Department of

the Army, 1971a) has published a report which included an
appraisal as to the erosional problems existing along the
eastern shoreline of Lake Michigan. No quantitative
measurements, however, were included. More recently Davis
(1972) has completed a two year beach profiling project at
17 nearly equally spaced sites along the eastern coast.
Erosional rates varied from zero to 25 feet in any one year
at a specific location.

Of all locations, 12 exhibited at

least some erosion . . .however,

only half of these (6) show some

erosion during both years. . . .

A wide variety of combinations exist

between location, year and amount of

erosion. Only two locations . . .

experienced approximately the same

amount of erosion both years (Davis

and Fingleton, 1972, p. 33).

The most extensive study known is that done by Powers

(1958). During two field seasons he mapped in detail the

geomorphology of the Lake Michigan shoreline. At 13%

 

2It was not indicated how many years ”prior" meant.

L,

section—line sites, to include 56 along the eastern shore of
the lake, he ascertained rates of bluff retreat by comparing
original land survey measurements taken in the early 1800‘s
with his own measurements taken in 1956. In addition, he
divided the Lake Michigan shore zone into 609 segments and
indicated for each the landform type, the height and slope
of the inland landscape, and the width and material compo—

sition of the backshore and foreshore.

STUDY LOCATIONS

Eight bluff sites located along the southeastern Lake
Michigan shore from west of Montague in Muskegon County,
south through Ottawa, Allegan and Van Buren Counties to the
village of Shoreham, immediately south of St. Joseph in
Berrien County, were selected for this study (Figure 1).
Erosional rates had been determined for each of these sites
by Powers (1958) in 1956 and conditions allowed for accurate
re-survey measurements to again be taken during the course of
this study at seven of the eight locations. The shoreline of
the sites tends to be gently curved with the northern sector
oriented to the north-northwest with a gradational change in
direction to the southern part which extends in a south-
westerly direction. The landface consists mainly of sand
dunes or high and low bluffs composed of till, sand and

gravel, silt and clay or an assortment of these sediments.

 

85
44

 

LAKE

MICHIGAN

II": IN.

FRAmflflHDRT

0cm Co

WHITEHALL
MWduoaICh
MUSKEGON

ISRAJHJIHANHN

Ottawa Co

HOLUMND

“bean Co

$OUWH HAVEN

. 0° ,
186°
' MI.

--—--—--L

«"0

IO 0 IO 20

 

Figure 1:

Location of

study sites.

 

These sediments are unconsolidated and are from Pleistocene
to Recent in age. No pre—Pleistocene bedrock formations are
known to be exposed. The Corps of Engineers (U. S. Depart-
ment of the Army, 1971a, pp. 86 and 9%) has designated
approximately 48 percent of this shore segment as areas of
critical erosion.

In addition to these eight sites is one associated
with a case study to be discussed later. The location of
this site is in the village of Shoreham and is adjacent to

the most southern of the eight sites mentioned above.

VARIABLES UNDER CONSIDERATION

In addition to obtaining the annual rate of bluff
retreat between the 1830's and 1956 and ascertaining the
yearly rate of bluff erosion between 1956 and 1973, eight
factors were considered in relation to bluff erosion rates
at each of the study sites. These variables are: (1) the
bluff composition, (2) the bluff height, (3) the bluff slope,
(4) the vegetative cover on the bluff slope, (5) any
hydrological phenomena within the bluff, (6) the shoreline
orientation, (7) the existence or non-existence of beach
frontage, and (8) any visible effect of man-made shore

zone structures on the bluff.

INVESTIGATIVE PROCEDURES

Following an initial reconnaissance in July, 1973, of
20 section—line sites included in the original study by
Powers, eight bluff sites were selected for re—examination
and assessment in August, 1973. These eight sites were
selected because they provided variations in the factors
under consideration,especially those of bluff composition and
shoreline orientation. In this study each of the eight loca-
tions is given a site number which duplicates that of
Powers' station number for that particular locality.

Powers had measured the distance from a section corner,
located within one mile of the lake shore, to the upper
bluff edge at each of the eight sites. He also obtained
from copies of the original U. S. land survey notes of the
1830's the chained distances from these same section corners
to the lake's ”meander line.” Assuming that in most cases
the ”meander line” represented the bluff line, he was there-
fore able to calculate an average annual rate of bluff
retreat at each of the study sites.

In this study the original government notes were re-
examined and the distances between bluff sites and section
corners recorded. Extrapolation from Powers' report
provided figures as to what these distances were in 1956.

On August 7-8, 1973, seven of the eight section-line dis—

tances were again chained, thereby enabling yearly erosion

rates to be computed between 1973 and both 1956 and the
1830's.3

With one exception, all distances in this present
study were chained with a 100 foot steel tape. A 300 foot
segment along the site 88 transit line was calculated by
trigonometric means due to the slope involved.

At each bluff site evidence pertaining to the eight
variables under consideration was recorded. At some sites
mass-wasted overburden covered part of the in-situ sediments.
In these cases the nature of the sediment was determined by
examining exposures generally within 30 feet either side of
the site line. Calculations from topographic maps were
used to determine bluff height at several localities because
dense or tall vegetation on the slope precluded accurate
trigonometric measurements. In places where pedestrian
traffic had notched sags in the bluff's upper edge and had
modified the natural slope, measurements as to bluff height
and slope angle were taken immediately north or south of

the site line.

The water of Lake Michigan generally rises to the
highest level in its annual cycle in the latter part of the
summer. A change in lake level affects a change in beach

width. Since each site was observed in both early July and

 

. 3N0 survey measurement was taken at site 112. The sec-
tion corner could not be adequately located in the field. In
addition, after examining the original land survey notes, the
15-minute, 1927, topographic map and the 7%-minute, 1970,
quadrangle, it became obvious that Powers' bluff retreat
figure in the 1958 report was substantially in error.

early August, two measurements of the beach width fronting

each bluff location were recorded.

SITE DESCRIPTIONS

The eight main study sites vary significantly. Three
bluffs (sites 89, 101 and 115) were primarily composed of
sand and gravel, three (sites 106, 109 and 111) largely of
glacial till, one (site 88) of dune sand, and one (site 112)
of assorted glacial drift. Bluff heights ranged from 30
feet (sites 88 and 101) to almost 110 feet (site 112) whereas
bluff slopes were generally greater than 30 degrees and at
several sites were vertical in at least a portion of their
slope. The shoreline trended from N279W (site 89) through
north-south (site 101) to N36OE (site 112). Vegetation, in
some degree, existed on the slopes at four sites (89, 106,
111 and 112). Groundwater percolated out from those bluffs
(sites 106, 109, 111 and 112) which contained till in some
portion of their profile. Beaches fronted four bluffs
(sites 89, 106, 111 and 112) during at least one of the
visits to each location. Man-made structures were present
at the water's edge at four sites (88, 89, 109 and 115).

All but three bluffs (sites 89, 106 and 111) presently
appeared to be experiencing substantial erosion.

A detailed descriptive account of each bluff site is

included as an appendix to this paper.

10

CASE STUDY

The purpose of this section is to describe how a
shoreline may evolve through time at a single bluff location.
A series of photographsl‘L is included to illustrate the
dynamic nature of a shore zone. Much of the information
presented in this section has been provided through the
courtesy of Mr. William J. Gibbs, Jr., whose coastal property
is the subject of this case study. Due to interest and as
proof for tax credit, he has kept records as to bluff erosion
along his property. These records include certified and
personal survey measurements, photographs, newspaper articles

and personal observations and notations.

Case study location

The shore zone bluff under consideration is located
along the Lake Michigan boundary of the village of Shoreham,
immediately south of St. Joseph in Berrien County (Figure 1).
The bluff segment is approximately four miles south of the
jetties at the entrance of the St. Joseph harbor. Mr. Gibbs'
property width extends 240 feet southward from a point 399.

feet south of the northeast corner of section 9, T5N, R19W.

 

UAll photographs with the exception of Figure 2 were
provided by Mr. William J. Gibbs, Jr. In no cases were
black and white negatives available. Due to this fact and
the age of the original photos, the quality of the reprints
may not be as good as would be desired. The sequence of
nine photos in Figure 8 were copied from color slides.
Black and white prints of those were unobtainable.

11

His property length measures from South Lake Shore Drive
westward to the water line.

The bluff along this shoreline segment is composed
largely of water-laid fine to medium sand with scattered
pebbles. The trend of the shoreline is north 27 degrees
east and in the past the bluff line was relatively straight.
Presently, however, the bluff line along a 425 yard stretch,
to include Gibbs' lot (in addition to site 115 in the main
study), is concave inland as a result of accelerated
erosion along that portion of the bluff (Figure 2).

Lake Michigan is approximately 70 miles in width at
this point. Of winds which generate waves affecting the
area, those from the north have the greatest fetch, 225
miles; those from the southwest have a fetch of 50 miles.
"During severe storms with a frequency of about once a
year, waves may range up to 11 feet in height in deep water,
but ordinarily waves of this height break before reaching

the shore structure" (U. S. House of Representatives, 1958,

p. H).

Shoreline evolution

Mr. Gibbs has recognized three periods of severe
erosion along the bluff since his family purchased the shore—
line property in 1916 (Table 1). He reports that at that
time abundant vegetation including large trees grew on top

of, and on the slope of, the bluff along the entire area from

12

 

 

 

Figure 2: The present shoreline bluff at the case study lo-

cation. Note its concave nature where accelerated
erosion has taken place.

13

Table 1: Average* retreat of upper bluff-
line along the Lake Michigan shore-
line property of William J. Gibbs,
Jr. during each of the severe ero-
sion periods since 1916. All mea-
surements were provided by Mr.

 

 

Gibbs.
EROSION AVERAGE LOSS AVERAGE LOSS
PERIOD DURING PERIOD PER YEAR
1928 - 1931 19.9' 6.60'
19H6 " 1953 5501' 7087'
1969 — 1973 63.0' 15.75'
(Jul 12)

 

 

 

 

 

* The average is computed by
dividing the sum of the bluff
retreat figure at the north
and south property lines by

wo.

1%

the St. Joseph harbor south to the southern limit of the
Grand Mere area. There included some 75 to 100 foot high
white pines along old wagon trails which led down to the
beach.

According to Gibbs, little erosion was evident along
the shoreline between 1916 and 1928.5 In the latter year,
however, the first of the three most recent severe erosion
periods began. From a 19HO certified survey it was estimated
that between 1928 and 1931 the bluff had retreated 19.9 feet,
for an average of 6.6 feet per year (Tables 1 and 2). A
picture taken in the summer of 1937 reveals a good recovery
of the beach and bluff (Figure 3).

The bluff was essentially stable from 1931 to 1945
after which time the second period of extensive severe
erosion began. Bluff loss totaled 55.1 feet when the
erosion period ended in 1953 (Tables 1 and 2). A photo
(Figure 4) taken on March 6, 19h9, depicts an eroded bluff
and according to Mr. Gibbs, by the spring of 1952 no growth
of any kind remained on the bluff.

In September, 19H8, three 50 foot groins and a 230

foot seawall were constructed because of occurring and

 

5Gibbs is also convinced after tracing the previous
ownership and legal description of his property back to
1872, from his own recollection of the shoreline bluff (and
expecially the nature of the vegetation on the slope) as a
youngster and after examining old family photos that no
serious bluff erosion occurred for many years prior to 1916.
He believes that the bluff line was essentially the same in
1872 as it was in 1916. This assertion could not be
presently confirmed by the investigator. '

15

.me.me u meme-meme
.um.: u memereame
.m:.m u meme-eeme

 

 

m<mw mam mmog mu<mm>¢

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o.mme o.me o.mee o.mm o.mem o.:oe o.:m o.emoe meme
.m Am:
0p
o.mm o.o 0.0m o.o o.o_oe 0.0m o.o o.omee meme
.mF ease
ea
o.mn m_.mm 0.0m F.om o.oeo_ 0.05 m.oe o.omee mmme
.eeo
oe
m.me m.me m.mm m.mm F.oeoe m.m m.m m.ompe came
am 008
op
..... - ------ ------- ---- o.omoe ------ ---- o.oome era?
.me..peo
mmoq ems QOHmma mmoq nae QOHmmm mmog nae QOHmmm
-aq52booa oszso -agbzpuoa oszpm mqu .«qb2booa eszbn mzHu mean
ammm 2H mmou zamz Emma 2H mmoq mamoz Heme 2H mmoq macaw
\(

 

 

 

.mnnflw .92 up woufl>09m one: mpnoaonzmmoa HH<

.mwdap cmMH£OHz oxmq exp we now can on o>Hpo onoam oxmq .m mo Hopmoo map mo puma
pmow mm omvmg w 809% Coxwp omonp one mooqumHU one .hpnogonn .Hh .mpnflc .h

amflflawz amp mo monHH hnmccdon SpSOm cam Samoa esp macaw upcoaohzmwoa counooom “N magma

 

16

 

 

 

Figure 3: Good recovery of the beach and bluff after the ero-
sion period of 1928 to 1931. The photo was taken
during the summer of 1937.

17

 

Figure h: Bluff badly eroded b the spring of 19%9 during the
erosion period of 19 6 to 1953.

18

anticipated erosion along the shoreline property during this
second erosion period. Sand had been removed from the lake
bottom to almost cover these structures. Two years later
during a storm, reported by Gibbs to have had a duration of
23 hours with winds up to 70 mph, on NOvember 18-19, 1950,
and another major storm a week later on November 25th, the
seawall and groins were demolished and the bluff slope
devastated (Figures 5 and 6). Mr. Gibbs indicated that

by the fall of 1951 the remaining structures appeared to

be ineffective and by 1953 they were almost totally gone.

A new era of bluff stability occurred between 195% and
1968. In 1961 the vegetative growth on the bluff was
substantial (Figure 7) and by 1968, as reported by Gibbs,
the bluff was covered with many young trees and other growth
(Figure 8-B).

Gibbs stated that in the summer of 1968 a large beach
fronted, and a heavy growth covered, the slope of the bluff
(Figures 8-A and 8-C). This scene changed early in 1969 when
the third and present period of severe erosion began.

Figure 8 dramatically illustrates the progressive damage done
to the bluff in a two year period. A survey on May 3, 1973,
revealed that during this latest attack on the bluff, average
loss from the top of the bluff had reached an alarming rate
of 15.75 feet per year; 63.0 feet had been lost since 1969
(Table 1 and 2).

19

 

 

Figure 5: View of the bluff on December 3, 1950 following the
devastating storms of late November.

20

 

Figure 6: Demolished seawall and groins, built two years ear-
lier, following the intense storms of late Novem-
ber, 1950. The photo was taken on December 3, 1950.

21

 

 

Figure 7: Reestablishment of vegetation on the bluff slope

during the low water stage, 195% to 1968. This
photo was taken on July , 1961. '

2?

Figures 8-A to 8-I: Sequence of nine photos showing the pro-
gressive damage done to the bluff in a
two year period as the Lake Michigan
level rose during the last erosion per-
iod.

 

 

Figure 8-A: March, 1968: Shows large beach and heavy growth
on bluff; breakwaters are neighbor's next door.

4. I . ‘
no). ‘
o

u ..

. .1.
.

..

I
...¥.T
... \\

69.

23

 

 

 

bluff before summer foliage is out.

 

Figure 8-B: March, 1968: Shows dense tree plantings on the

 

1968: Shows large beach and heavy

bluff growth before storm damage early in 19

Figure 8-C: October 31,

 

 

Figure 8-D: March 27, 1969: Shows severe damage to bluff and
loss of trees and path damage from storm of
MarCh 25-26, 19690

 

 

 

Figure 8-E: July 29, 1969: Shows more bluff and tree loss
from the storm of July 27-28, 1969.

 

 

25

3 .

._ " ‘

I. " . '8 n O
' .. .4'4 I ~1 \ ‘
I '~ I - l '
c. .. v . .. Q.” ~. - .nv
. - o,‘ v°~~\ .ffl’q.‘ - . . . .
1“ .. I , - -’ 1'01!“-
. — ‘ _ -

- V.“ ,9 5

 

Figure 8-F: November 22, 1969: Shows the bluff after the storm

 

of November 19-20, 1969.

 

 

 

Figure 8-G: November 22, 1969: Shows view of bluff and beach,

logking north after the storm of November 19-20,
19 9.

26

 

 

Figure 8-H: December 5, 1969: Looking north showing loss from
tog of bluff caused by the storm of November 19-20,
19 9.

 

 

 

 

 

Figure 8-I: December 5, 1969: Same view as photo 8-A; looking
south showing bare bluff and loss from top caused
by the storm of November 19-20, 1969.

27

Discussion

Since 1916 Gibbs' Lake Michigan shoreline bluff had
become stabilized during each period of low water heights
while during high levels loss of bluff material had been
considerable. The water level began to rise substantially
in 1928, although it still was below the long-term average.
The level was above the norm in 1929 and 1930 before
drOpping again in 1931. During the 1928 to 1931 erosion
period bluff recession was three times greater along the
northern property line than along the southern one (Table
2). It appears that at the southern extent remnants of
several groins and a breakwater, built in 1906, had somewhat
reduced the rate of erosion.

The lake rose slightly above the mean in 19H3, dipped
below normal between 19A8 and 1950 before rising above
average during the years 1951 to 1955. The severe erosion
occurring during November, 1950, can be at least partially
explained by the fact that it was associated with a storm
with wind gusts up to 70 mph and a duration of 23 hours.

The Lake Michigan level rose above the statistical
norm again in 1969, the same year severe erosion began once
more on the bluff. High lake levels and extensive bluff
retreat are continuing today.

During the present erosion period the northern bluff
crest has been eroded almost three times as fast as that at the
southern limit. This situation appears to be attributed to

the position of the remains of a breakwater, built in 1952,

28

adjacent to and south of Mr. Gibbs' property. Located a
short distance offshore, it prevents the full impact of the
waves from reaching the landface. This decaying structure
was supplemented by a concrete seawall constructed along
this same shore section last year. Similar structures do
not exist along or near the northern margin of the shoreline
property.

Several questions relating to the shoreline description
and erosion are not as easily answered. What factor(s) has
caused the present bluff line to be curved inward along a
#25 yard stretch while in the past it had been relatively
straight? Why has the annual average bluff erosion rate
been progressively higher during each of the three erosion
periods? If it is assumed that Gibbs was correct, why was
there apparently no serious erosion on the bluff between
1872 and 1916?6

Evidence is lacking but the jetties at the entrance of
the St. Joseph harbor may very well have been and are im—
portant factors in the shoreline evolution in a downdrift
direction which would include the vicinity of the case study
site. It is cautiously suggested that they may be related
to the answers to the questions raised above.

In 1836 the first 1,023 feet of the north pier at St.
Joseph was built, this being the first lake structure con-

structed in the area. By 1880 another 804 feet had been

 

6The reader is advised that this conclusion has not
been verified in this study.

29

added, this segment on a new alignment, thus forming an angle
point in the pier. By 1903 the remaining 1,002 feet of the
2,829 foot north pier had been built in addition to the con-
struction of a south pier (U. S. House of Representatives,
1958, p. 18).

Examination of the 1927 topographic map reveals a sub-
stantial build-up of sand behind and north of the jetties.
This accumulation extends further lakeward than the angle
bend in the north pier, thus implying that a sand beach at
least 1,023 feet in width had formed immediately adjacent
to the pier since it was first constructed. More recent
aerial photographs and the 1970 topographic map reveal that
the sand extends approximately 150 feet further waterward
than the angle bend.

Prior to the construction of the north pier the mouth
of the St. Joseph River had turned to the south (U. S.

House of Representatives, 1958, p. 18). This deflection
suggests that dominant movement of significant amounts of
sediment is southward along this part of the shoreline.
Thus it appears that the construction of the piers dis-
rupted the natural littoral transport of beach building
material to the south of the harbor entrance and changed the
regimen of the beach zone. The Corps of Engineers confirms
these conclusions (U. 8. House of Representatives, 1958,
pp. A and 22).

Although a beach may temporarily be

eroded by storm waves and later restored
by swells, and erosion and accretion

3o

pattern may occur seasonally, the long-
range condition of the beach - whether
eroding, stable or accreting - depends
on the rates of supply and loss of
littoral material. Erosion or re-
cession of the shore occurs when the
rate of loss exceeds the rate of supply
(U. s. De artment of the Army, 1971b,

pp. 21-23 0
Furthermore, the Corps of Engineers (U. S. House of Repre-
sentatives, 1958, p. 23) state that
groin construction alone is not a
suitable means of stabilizing the
bluff in the portion of Berrien
County immediately south of St.
Joseph Harbor, as natural sources
of supply are insufficient to supply
material at a rate required for im-
pounding a suitable protective beach.

The foregoing statements seem to lead to a possible con-
clusion that ever since the construction of the jetties at
the St. Joseph harbor entrance a naturally occurring pro-
tective beach is no longer able to maintain itself to the
south. Once the original beach was carried away through
wave action, and with no possibility for re-nourishment,
accelerated erosion of the backing bluffs was inevitable,

especially during substantial storms and periods of high

lake level.

Conclusion

Several relationships have been illustrated by this
case study. Unconsolidated glacial drift is easily sus-
ceptible to erosion by waves. Bluff erosion is much more

likely during periods of high lake levels than during low

31

water periods when slopes tend to become stable. Storms

can have drastic consequences on the landface, even during
times of low lake level. Man-made structures may possibly
either reduce or accelerate bluff erosion but no structure

seems to fully prevent it.

BLUFF RETREAT - RATES, FACTORS AND MODES

Comparison of average annual bluff ero-
sion rates prior to, and since, 1956

On the basis of the current investigation and Powers'
study (1958) a comparison can be made of bluff erosion rates
between the period of the 1830's to 1956 and the period 1957
to 1973 at seven,7 of the eight study sites (Table 3). Aver-
age annual erosion rates were not always similar for the two

periods under consideration at a given site (Table 4).

Sites which recently experienced in—
significant amounts of bluff erosion

The crest-line at bluff sites 89, 106 and 111 exper-
ienced no or an insignificant amount of retreat between 1956
and 1973 (Tables 3 and H). In contrast, the calculated annual
rate of bluff erosion between the 1830's and 1956 at these

3. 09 we"
locations revealed losses averaging from 1.03 tO‘Tfkr feet
C 60‘?ch

per year (Table A).

 

7Bluff retreat figures were not established at site 112.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ o.mmm _ o.am .m:_ _ o.emm mmma maem.mmm\m eeexaoe 3m nemaaam mew
.................... _ 0.:Fm.e omme swam.mmm\m eeaxaoe mm camaaem New
_ 0.0mm _ o.o .omm _ o.omm.: omwe swam.mmm\_m eam\aoe mm nemaaam Fee
.::.mmm _::.o: .wam .am.mam emme meem.mem\m_ eam\aoe 32 steam aa> ace
.mm.mme .mm.m .ome .mm.emo.e mee meem.2mm\om eea\eaoa «z aamemma eoe
_ o.mm _ o.m .ee _ 0.:mm mmme meem.zmm\ee eea\eeoa «m asapeo For
My
_ 0.06? . o.o .mee _ o.wm:.: umme meem.zeem\_m eaa\aoe mz somemaaz mm
mama _|¢t¢m+ .nmhmu .mme _ u.wmo.m mmwe swam.2mem\mm eem\aoe 32 aowammaz mm
‘lw...39 \®0§W\ swswm
A3
meme om emme msz
mammmmm mama-emae om mmmzmmz: mean .02
mmbgm 85mg mm B<mgmm O .H. nggm ZOH H<OOQ MHZDOO mHH m
mmmom mmmmm mmmmm mozmmmmm mazHUHmo
.umpoamaoo mmoz mmo>m3m acma .m.D
Hmcmmmno emu mocmm pmompom ommanmmsan HmpOp mmosp cum mopflm budmm mo mzompmooq "m magma

 

33

Table A: Description of study site variables.

 

 

 

 

 

 

 

 

 

 

 

 

 

Small Pebbles

 

 

30-35

 

a a e ,.
213% fig S 3 53
E5; m H (D EHO)
ta [nos E4 9 Eih
am Em .. a ..
fi m3 m7 mg mg Egg EEG
aa aa; sat Ea a; BBB eta
$3 E58 Ear m0 mm mmv mew
88 1.07' 2.17' Dune Sand 30' 313-37 N1MW
89 1.h1' 0.00l Med. Sand With 45' 30—37 N27W
Scattered Pebbles
101 .53' .53' 2- ' Eolian Sand 30' 30-3H N-S
2 ' Med.-Cr. Sand
106 1.03' .17' 5' Sand & Gravel 55' 35-49 N6E
26' Till
Top 14' To N
Efi' Sand 90 N13E
I I ' I
109 2.74 2.38 31' gipg & Gravel H1 Remain. To S
35-H0 S28W
Sa. & Gr.
6-2' Sand & Gr. 55-90
2—3' Sand Nip
3o-us
12-1 ' Till ,
85-9g' Various Beds 100
112 ---------- of 8a., Sa. to! 30-90 N36E
& Gr., Silt 110
&fllay
TOP 3'
water-Laid Sands ' 90
I l '
115 1.11 4.76 hath Scattered 6O Remain. N27E

 

 

Table 4: continued

34

 

 

 

 

 

 

 

 

 

 

é :
z D Em
C) HI a: 2;
H U [Hg 0
H OU) O E‘IH
< am me 29 E
B 00 m HO mH
hi me o 0: mm mg
0 as: a a a as a
E Ea. m = a 830 as
Breakwater to
North Severe
None None None Groins to Erosion 88
North & South
Grass 3 Jul
Bushe: 10X Stable
Young None Au . Groins 2' Nip in 89
Trees E: Base of Slope
Severe
None None None None Erosion 101
July Rel. Stable
Grasses Ground 10' 3' Fig N
Small water Aug. None T6; Whig; ear 106
Trees Seepage None Slumped
Ground Concrete
Water Slabs Adje
None Seepage None & South Severe 109
Pipe Mile North
Ju1y Rel. Stable
Ground 39: Minor Slump
W00ded Water Aug. None Near Top, None 112
Seepage 25: At Base
Ground Unstable
Sparse water July & Slumping
to Seepage Aug. None Gullying 112
None Drain 20-25' 10' Nip At
Pipe Base
Breakwater -
Groin Complex; S ver
None None None Adj. & South Eiosign 115

 

 

 

Jetties Four

 

Mileg North

 

 

 

35

The bluff at site 89 is composed predominantly of water-
laid sand whereas at sites 106 and 111 exposures reveal till.
Remnants of a low foredune, probably developed during low
water stage, fronted the bluff at site 111. All three slopes
had an abundant cover of vegetation with the most dense
growth at site 111. Furthermore, these three bluffs were
the only ones which had been fronted by a beach during at
least one of the visits to the sites8 during the summer of
1973.

Man-made shore zone structures (in this case, permeable
groins) exist only at site 89. One-hundred and forty yards
south of this site erosion has been quite severe in a seg—
ment of the sand and gravel bluff immediately adjacent and
north of an impermeable groin.

Slumping and block gliding of undetermined age have
occurred on the slope of both till sites (106 and 111). At
each of these locations sandy beds overlie the till. Ground-
water was observed seeping out from the bluff slope at the

base of this sandy layer at both locations.

Sites which recently experienced
accelerated erosion rates

The sand dune bluff at site 88 has eroded back at an
average annual rate twice as fast (2.17 feet per year average)

as it did prior to 1956 whereas the rate at site 115 is four

 

8This does not include site 112 for which no rates of
retreat were determined.

36

times greater (4.76 feet per year average) than before 1956.
The bluff at this latter site consists largely of water
deposited sand. Both bluff slopes are presently free of
vegetation and each is currently experiencing severe erosion.
Breakwaters and groins, constructed within the last
three years, are adjacent to the resurveyed section line at
both sites and site 115 is located approximately four miles
south of the jetties at the St. Joseph harbor entrance.
(For a detailed discussion relating to these jetties see

page 28).

Sites which recently experienced erosion
rates similar to those prior to 1956

Two bluff segments (sites 101 and 109) are presently
retreating at average annual rates similar to those which

T50. prPJaamuh'G [.17‘/“ff “fl- 9,)“: (c7 LS
they were experiencing prior to 1956. Beth—bluffe—are

p

currently retreating at a rate five times greatefit(5:38 feet eififi’

per year average) than that of the sand bluff’kTable 4).

Each of their bluff slopes is presently void of vegetation.
Protective barriers do not exist in the area adjacent

to the base of the sand bluff (site 101). However, at the

till bluff (site 109) concrete slabs have been placed at

the base of the slope immediately south of the site. The

slope behind these slabs is partially covered by clumps of

grasses and is less steep than at the site line.

At site 109 groundwater may seep out from the bluff

face at the base of the permeable sand and gravel layer

37

which overlies the till and the mid-slope is covered by

a mixture of till and sand and gravel which has slid down
from further upslope. No groundwater percolation is visible
at site 101.

The bluff at site 109 is situated one mile south of
the South Haven harbor. From a point one-third of a mile
south of this survey point the shoreline to the north curves
noticeably inland all the way to the harbor jetties. Ero-
sion along this stretch is severe (U.S. Department of the
Army, 1971a, p. 86).

A major area of sand dunes exists along the shoreline
three miles north of site 101. Examination of the topo-
graphic map and air photos suggest that the dominant lit-
toral drift is mainly to the south along here. No major
structures extend into the lake between the dunal area and
the bluff at site 101, which since 1832, has been retreat-

ing at an average annual rate of .53 feet per year.

Selected variables investigated in
relation to bluff erosion rates

At each bluff site eight variables9

were observed to
determine if they might have a clear relationship to recent
erosion rates. Only two factors appeared to have a definite

relationship to rates of bluff recession and a third factor

 

9(1) Bluff composition, (2) bluff height, (3) bluff
slope, (4) vegetative cover, (5) hydrological phenomena,
(6) shoreline orientation, (7) beach frontage, and (8) man-
made structures.

38

may be related in a meaningful way. Three variables, al-
though not related to actual rates of erosion, were related
to modes of bluff retreat. Two other factors revealed no

apparent relationship to rates of erosion.

Beach frontage and vegetative cover

”A wide beach is the best protection the upland shore
can have from wave attack" (U.S. Department of the Army,
1971a, p. 13). It acts as an energy absorber as incoming
waves over-steepen and collapse along the shoreline. During
periods of above average lake levels, however, beaches may
become narrowed or submerged. In these cases wave action
may then more easily reach the base of, and become quite
destructive on, the slope of the landface.

The preceding relationships were reaffirmed during this
investigation. The three bluff sites (89, 106 and 111)
which had insignificant rates of bluff retreat since 1956
had been fronted by a beach during either or both July
and August of 1973 (Table 4). (Beaches existed at no other
sites which were resurveyed and at all of these sites the
bluffs had experienced recent erosion and their slopes were
presently void of all vegetation.)

Even though waves have eroded a two-to-three foot nip
in the sand (site 89), till (site 106) or low foredune (site
111) at the base of the bluffs, an extensive cover of
vegetation exists on the slope. However, in the cases of

sites 89 and 106, an especially critical stage seems to

39

have been reached. It appears that if the water level should
remain high above average for an extended period, erosion
and removal of vegetation by storm waves at the base of
these bluffs could become severe. This situation is pos-
sible, but not quite as likely, at site 111 for two reasons.
Here the beach is still relatively wide (25 feet) and the
bluff itself is fronted by the remnants of a low foredune.
Foredunes may act as effective secondary defenses
should beaches become diminished. It is apparent that they
act much like breakwaters in preventing wave attack along
the base of the bluffs. Exactly what accounts for their
formation is not known but investigation into this matter
could be promising. If the mechanics and nature of their
formation could be determined it may become possible to
artificially induce them to form along the shoreline at
times of low or average water levels. This in turn may
help reduce or prevent bluff retreat during times of high

water.

Shore zone man-made structures

The construction of man-made structures along the Lake
Michigan shore zone has increased in number since 1956,
yet limited evidence indicates that they may be of little
or no help at given locations (Table 4). In fact, in the
case of groins and jetties, they may accelerate erosion, if
not at their own sites, at adjacent localities and especially

in the average downdrift direction.

1+0

Brunn (1968, p. 69) states:

It is not a matter of coincidence
that most beach erosion in Florida
occurs where the most concentrated de-
velopment along the shores has taken
place. Man's interference with natural
shore processes is the major cause of
the increased erosion in some of the
highly populated areas on the lower East
Coast and the lower Gulf Coast of Flor-
ida. Groins have caused severe leeside
erosion problrps in many areas along
the Florida shore, and have, unfortun-
ately, often done more harm than good,
although this conclusion shall not be
generalized. In numerous cases, sea-
walls have protected valuable property
or dune faces, but when built as verti-
cal bulkheads, they have, in fully as
many cases, caused a lowering and in—
creased erosion of the beach itself,
giving rise to heavy water turbulence
by waves colliding with the wall at
high tides.

Various types of structures have been built in the
vicinity at four of the five locations (to include the case
study site) which are presently undergoing substantial ero-
sion (Tables 1, 2 and 4). As previously mentioned, there
is reason to believe that current erosion at sites 109 and
115, in addition to at the case study location, is possi-
bly related to jetties projecting into the lake at several
harbor entrances. Breakwaters and groins built within the
last three years at sites 88 and 115 and at the case study
site are not preventing erosion from occuring on the bluff.
One-hundred and forty yards south of site 89 two impermeable
wooden groins, 25 yards apart, extend 50 feet into the lake.
No beach zone exists and bluff erosion is severe for about

40 yards to the north of the northern groin. This erosion

41

appears to be precipitated by refraction of incoming waves
off of the groin and into the bluff. Between the two groins
a beach approximately 25 feet wide, is elevated two to three
feet above the base of the eroded bluff at the north.

There is a real need to evaluate exactly what effect
these so-called protective structures are having on shore-
line evolution and bluff erosion along the western shore of
Michigan. It may very well be that the shore zone acts
much like a river system. River systems seek a state of
dynamic equilibrium. If any portion of the system is
disturbed or changed in some way, adjustment must occur in
other portions of the river system. This principle of
teleconnection may apply in the case of shoreline struc-
tures. For instance, the construction of a groin or jetty
at one location along the shore zone may disrupt or change
the entire regime along another segment of the coastline
because of its ability to prevent or reduce littoral drift

and to affect localized currents.

Bluff height and shoreline orientation

The variables of shoreline orientation and bluff height
appeared to have no bearing on rates of bluff recession nor
do they contribute directly to processes of slope retreat,
at least at the locations in this study (Table 4). It seems
that even though bluffs of increasing heights provide in-
creased volumes of sediments, waves are nevertheless able

to effectively remove the material resulting in a situation

42

where average bluff recession rates appear not to be

related to bluff heights.

Bluff composition, slope and
hydrological phenomena

Results were inconclusive in an attempt to relate reces—
sion rates with bluff composition, slope and hydrological
conditions (Table 4). Nevertheless, these three variables
appear to perform critical roles in the varying processes of
bluff slope retreat. Therefore, discussion of these factors

is deferred until modes of slope retreat are treated.

Modes of bluff slope retreat

10 studied were under-

The bluffs at the nine locations
going different modes of recession. Variations depended
largely on bluff composition, slope, hydrological situations
and vegetation. Sandy (to include sand and gravel) bluffs

tend to retreat primarily by the process of debris sliding,11

 

0The eight sites and the case study location.
11
Debris sliding: the rapid rolling or sliding of un-

consolidated earth debris without backward rotation of the
mass; these slides are characterized by the tendency of the
slide material to behave as a more or less cohesionless mass,
suffering considerable distortion during movement. They gen-
erally occur on fairly steep slopes, typically between about
15 and 40 degrees, and are frequently fairly rapid. The depth
of the movement and the degree of distortion involved is
influenced largely by the cohesion of the slide debris. At
one extreme, heavily weathered clay debris may approach the
nature of a ”slab slide;" at the other is the "sand run" in
slopes of dry cohesionless material, in which the movement

involves only the grains in a thin surface layer (Hutchinson,
1968, p. 691 .

Lt3

‘2 block gliding'3

till slopes by the processes of slumping,
and fluvial activity, and bluffs of assorted glacial drift

by a mixture of the aforementioned processes.

Retreat of sandy bluffs

Disregarding for present the factor of vegetative cover,
sediment movement on sandy bluffs tends to be frequent and
the slopes less steep and more uniform than on till bluffs.

Except at the top of the profile where there is usually a

 

12Slumping (rotational slip): downward slipping of one
or several units of rock debris usually with a backward
rotation with respect to the slope over which movement takes
place. They occur principally in slopes largely formed of,
or underlain by, a fairly thick and relatively homogeneous
deposit of clay or shale. Failure takes place, usually fairly
rapidly, by shearing on a well-defined, somewhat curved, slip
surface. This, being concave upward, imparts a degree of
backward rotation to the slipping mass which produces sinking
at the rear, heaving at the base and back-tilting of the
slipped strata. Elongated pools commonly collect in the
depression formed behind the slipped mass (Hutchinson, 1968,

p. 695).

13Block glides: a type of slide in which the moving
material moves out or downward without the backward tilting
characteristic of a slump (Selby, 1970, p. 32). A block
glide is one type of a translational slide. Landslides of
this type are usually fairly rapid and involve shear failure
on a fairly plane surface running roughly parallel to the
general slope of the ground. These slides are widespread in
cohesive soils in which the failure surface is predetermined
by a marked heterogeneity, such as a sharp transition from
soft to hard material with depth, or the presence of an
adversely located weak layer within the slope (Hutchinson,
1968, p. 690).

as

near vertical section three to six feet thick,1U the sedi-
ments tend to settle at their angle of repose (approximately
32 degrees) although in actuality the slopes may vary from
30 to 37 degrees depending on the cementation properties and
cohesiveness within individual sand layers. Movement down-
slope is by individual grain particles. Any slope distur—
bance tends to result in a downward adjustment of material
on the slope. These adjustments result mainly from waves
that attack and remove material at the bluff's base, result—
ing in debris slides of varying magnitude and duration.

It is obvious that vegetation on sandy slopes does not
prevent erosion through wave action at the base of the bluff,
but it does help to prevent mass-wasting on the slope itself.
However, during storms waves may erode and wash away
flora rooted in the unconsolidated material at the base of
the slope. Under repeated wave action the lower portion of the
bluff may erode back and the slope become over-steepened,
thus inducing debris slides further upslope. Through
this process additional vegetation cover may be removed.
Eventually the entire slope may become bare of vegetation

and thus even more susceptible to accelerated mass—wasting.

 

1LIThe reasons that the slopes are near vertical at
the upper edge of the sandy bluffs was not investigated in
detail in this study but contributing factors may involve
rooting of vegetation on the top of the bluff and an upper
clay-silt enrichment zone.

45
Erosion of till bluffs

Bluff slopes in till are generally steeper and vary in
form to a greater extent than those developed on sandy
bluffs. Slopes are inclined from 30 degrees to near vertical
(Table 4) irrespective of position on the slope face. Mass-
wasting on the till slopes is discontinuous in time and
distance and when it does occur, is mainly by downward move—
ment of blocks of till rather than by way of individual
clastic particles.

In order to understand the nature of slope retreat on
the till bluffs investigated it may be worthwhile to review
several attributes of clay, soil cohesion, porewater
pressure and landslides.

Clay is an important constituent of the tills exposed
at the study sites. The fine texture of clays accounts
for their ability to be both highly porous and highly
impermeable. High porosity accounts for their high water
holding capacity but their microscopic pores impede an
effective flow of water through the clays.

The cohesion of particles depends
largely upon the water between them...
In unsaturated soils attractive forces
between soil particles exist because of
the capillary tension of adsorbed water.
The actual strength of the capillary
tension is proportional to the radius
of the curvature of the water surface
between the soil particles. The smaller
the radius the greater the capillary
tension and hence the more cohesive the
soil. Hence,...high moisture contents
...contribute to low cohesion. In a

saturated soil surface tension is com—
pletely eliminated (Selby, 1970, pp. 36-38).

46

If water enters a soil it tends to fill up the inter-
granular spaces causingra rise in porewater pressure, and
because of the loss of cohesion of the solid particles,
their weight also may be supported by the porewater. If
the porewater pressure becomes great (for instance, as
water enters the soil from a rainfall) overlying soil and
water may "float" on the porewater beneath and the porewater
may burst out of the voids. This may cause the soil to
move (Selby, 1970, p. 41).

Slumping and block gliding are types of landslides.

The condition of movement of any
particular landslide can be extremely
complex. Simply stated, however, a
slide is a shear failure which will
be set in motion when the stress along
the potential surface of rupture ex—
ceeds the resistence to shear along
that surface. This unbalanced condi-
tion may be brought about by (a) an
increase in the weight of the over-
lying material due to the absorption
of water; (b) a decrease in the re-
sisting mass due, for example, to
undercutting a slope; (c) a decrease
in shear strength of the material
itself, which may be caused by the
absorption of moisture (Leopold
welman and Miller 1964, p. 3403.

All till bluff sites revealed evidence of slumping
or block gliding on the slope, although it may not have
occurred within the last several years. At each of these
sites evidence of groundwater percolation was noticed on
the slope face. At several sites portions of the clay till
appeared saturated with water. Since the role of water in
the reduction of shearing strength is a major one (Savage,

1968, p. 697) and since clay and silt naturally have low

47

bearing strengths (Savage, 1968, p. 697), it appears that the
moisture content within a bluff plays an important role in
bluff recession.

The base of a till bluff may be eroded back to a cer—
tain threshold without causing downward movement of till
blocks from upslope. The critical situation, at which time
movement does take place, may occur once the slope attains
a critical angle and the groundwater, reinforced by excessive
rainwater, fully saturates a portion of the till profile,
thereby adding mass, eliminating particle surface tension
(cohesiveness), and inducing high porewater pressure within
the clayey till.

Since clay and silt are important constituents of most
tills, the tills tend to be rather impervious to a very effec-
tive downward movement of water. Any channelization of the
water, which in effect increases the water's potential
erosional power, may initiate extensive gullying into the
bluff slope. Site 112 illustrates a bluff segment which has,
and is, undergoing severe gullying. In at least a portion of
the bluff two drain pipes protrude from the slope face
near the crest line. Water flowing from these pipes has
carved gullies downslope into the in-situ till section or
the thick slumped or glided till blocks covering the bluff
face.

Since till is able to stand at angles up to 90 degrees,
and since slump or glide blocks frequently do not travel

extended distances downslope, the profile of a till slope may

48

change over a period of time. This may occur without any
corresponding change in the position of the crest line; or,
the top edge of the bluff may recede very slowly or not

at all while the base of the bluff may recede rather rapidly,
at least until some critical stage is reached. At most

till bluff study sites it appeared that slump or glide blocks

of several ages were associated with the slope.

Erosion of bluffs of assorted
glacial drift

In many cases along the southeastern Lake Michigan
shoreline groundwater seepage appears detrimental to the
maintenance of bluff slope but in one case such water may
aid in the prevention of accelerated erosion. This situation
may take place if a sandy material comprises the lower por-
tion of a bluff whereas the upper part consists of two layers
of impermeable material (till and clay, for instance) separ-
ated by a permeable sandy sediment. Groundwater may then
migrate along this upper permeable sediment bed, saturating
the top portion of the impermeable clay layer below While
at the same time continuing to remove small quantities of
the sand above as it seeps out from the slope face. Even-
tually support for the till will fail causing the till to
slump or glide downward. If the till block is large enough
it may cover the sandy portion of the slope below, thus
acting as a protective cover against the frequent mass-

wasting and removal of the sandy material at the base of the

49

bluff. A situation such as this exists at site 112.

Much the same condition could exist if a bluff of sand
and gravel, overlain by till, was eroded by waves at its
base. Initially sand and gravel would be removed with the sub-

sequent collapse of the till over the sand and gravel.

Erosiongpotential as related to storms,
time of yearLAand high water levels

High water levels in themselves do not necessarily
result in eroded shoreline bluffs. However, in combination
with storms, they may act as highly efficient transport
mediums allowing incoming waves to reach the landface.
Cyclonic disturbances occur in association with, and are
steered by, upper atmospheric wind flow in the middle
latitudes. Major disturbances appear to be more common dur-
ing the transitional seasons and are especially frequent in
November and March. During these seasons baroclinity15 in
the westerlies is frequently enhanced, resulting in stronger
cyclonic development. More frequent winds with greater speed
induce larger waves over the Lake Michigan fetch and when
these over-steepen and collapse along the shoreline they may
possess great erosional ability. In terms of volume, during
most years, it appears that relatively little material is

lost during the high or low sun periods.

 

Baroclinity refers to the meridional pressure gradient.

50

CONCLUSION

Current rates of bluff retreat were determined at seven
of eight study sites along the southeastern shore of Lake
Michigan. These rates were compared to those calculated
at the same sites for the period from the time of the land
surveys of the 1830's to 1956. Bluff erosion rates are not
uniform along the shoreline nor can recession rates be
predicted to be similar at very similar sites during two or
more distinct time periods. Severe erosion is most likely
during major cyclonic disturbances. High water levels, al—
though not necessarily a prerequisite for erosion, provide
a situation for incoming waves which especially favors erosion
of the shoreline bluffs. Of eight variables only that of
beach width appeared related clearly to erosional rates.
Bluffs fronted by beaches retained vegetation on their slopes
because the beaches prevented waves from reaching and dis-
rupting the environment necessary for plant growth. .Two
factors, bluff height and shoreline orientation, were seem-
ingly unrelated to rates of bluff recession. Investigation
as to the effect on shore zone erosion as related to man-made
structures ended in inconclusive results. It is believed,
although no clear proof is available, that these structures
do indeed affect shore zone processes, and therefore, shore—
line evolution and erosion. The remaining variables, bluff
composition, slope and hydrological condition, all were

involved in modes of bluff slope retreat.

51

It appears that better success in understanding the
causal factors in shoreline erosion may be met by extending
the investigation into the waters along the shore. Physical
limnoligical studies may be more productive than geomorphic
studies. Such factors as localized eddies and currents, off-
shore bars and littoral drift may prove to be primary factors
in the dynamic production of shoreline evolution. In addi-
tion, an urgent need seems to exist in determining exactly
what effect do man-made structures have on the natural

processes which act within the Lake Michigan shore zone.

L IST OF REFERENCES

LIST OF REFERENCES

Brunn, P. "Beach Erosion and Coastal Protection," Encyclope-

WW- Volume III, W-
morphology, ed. Rhodes W. Fairbridge. ew York: Rein-
hold Book Company, 1968.

 

Davis, Richard A., Jr. "Systematic Beach Profile Study of
Eastern Lake Michigan," Final Report, December 31, 1972,
of Contract DACW72-70-C-OO37 for the Department of the
Army, Coastal Engineering Research Center. Kalamazoo,
Michigan: Department of Geolog , western Michigan Uni-
versity, 1972. (Mimeographed.)

and W. George Fingleton. "Systematic Beach Profile
Study of Eastern Lake Michigan (1971—1972)," Annual
Report, August 20, 1972, of Contract DACW72-70-OO37 for
the Department of the Army, Coastal Engineering Re-
search Center. Kalamazoo, Michigan: Department of Geo-
logy, western Michigan University, 1972. (Mimeographed.)

Hutchinson, J.N. "Mass Movement," Encyclopedia of Earth
Sciences. Volume III, Encyclopedia of Geomorphology,
ed. Rhodes W. Fairbridge. New York: Reinhold Book
Company, 1968.

Leopold, Luna B., M. Gordon WClman, and John P. Miller. Flu-
vial Processes in Geomorphology. San Francisco: W.H.
Freeman and Company, 1964.

Leverett, Frank. The Illinois Glacial Lobe. United States
Geological Survey Monograph XXXVIII. washington, D.C.:
Government Printing Office, 1899.

Powers, William E. ”Geomorphology of the Lake Michigan Shore-
line,” Final Report of Project No. NR 387-015, Con-
tract No. Nonr-1228(O7) for the Geography Branch, Earth
Sciences Division, Office of Naval Research, Department
of the Navy. Evanston, Illinois: De artment of Geo-
graphy, Northwestern University, 195 . (Mimeographed.)

Savage, C.N. "Mass Wasting," Encyclopedia of Earth Sciences.

Volume III, Encyclopedia of Geomorphology, ed. Rhodes W.
Fairbridge. New York: Reinhold Book Company, 1968.

52

53

Selby, M.J. Slopes and Slope Processes. Publication No. 1,

waikato Branch of the New Zealand Geographical Society.
Hamilton, New Zealand: A.O. Rice, Ltd., 1970.

U.S. Department of the Army, Corps of Engineers. Great

Lakes Region Inventory Report. National Shoreline
Study. Chicago: North Central Division, Corps of
Engineers, 1971a.

U.S. Department of the Army, Corps of Engineers. Shore Pro—

tection Guidelines. National Shoreline Study. wash-
ington, D.C.: Government Printing Office, 1971b.

U.S. Department of the Army, Corps of Engineers, Coastal

Engineering Research Center. Land Against the Sea.
Miscellaneous Paper No. 4—6H. Washington, D.C.: Govern-
ment Printing Office, 196%.

U.S. Department of Commerce, National Oceanic and Atmospheric

Administration, National Ocean Survey. Monthly Bulletin
of Lake Levels. Detroit, Michigan: Lake Survey Center,
July, 1973.

U.S. House of Representatives. Beach Erosion Study Berrien

 

County, Michigany_Michigan-Indiana State Line to Benton
Harbor. House Document 336, 15th Congress, 2nd Session.
Washington, D.C.: Government Printing Office, 1958.

 

APPENDIX

APPENDIX

Site Descriptions

The description of each of the study sites follows the

outline listed below. A more detailed description may be

obtained from the author upon request.

1.

Location

a)
b)
c)

a)
b)
c)

d)

County

Topographic map, series, date of issue

Township and range of the point where the survey
measurement begins

Date of original land survey

Distance from the section corner to the "meander

line" at the time of the original survey

Bluff recession between the date of the original

survey and 1956

Average annual rate of bluff retreat between 1956
and the year of the original survey

Distance measured from the section corner to the
top edge of the bluff in 1973

Total bluff retreat between 1956 and 1973

Average annual rate of bluff retreat between 1956

and 1973

Bluff composition

Bluff height

Bluff slope

Vegetative cover on the bluff slope

Hydrological phenomena within the bluff

Shoreline orientation

Beach width

a)
b)

July
August

5%

55

11. Man-made "protective” structures
12. Present condition of the bluff

13. Additional comments

12.
13.

56

Site 88
a) Muskegon
b) Montague, 15', 1959
c) NW corner/sec. 23/T12N,R18W
a) 1837
b) 3,098.7 ft.
c) 128 ft.
d) 1.07 ft.
8.) 2993308 ft.
b) 36.9 ft.
c) 2.17 ft.
Dune sand
30 ft.

31%-37 degrees
None
None
N14W

a) None
b) None

Groins extend north and south along the base of the bluff.
A seawall is built along the property adjacent and north
of the site line. Powers indicated no structures here

in 1956.
Erosion is severe

Small debris slides or sand runs occur on the bluff
slope whenever anything disturbs the surface or base
of the slope. The dry, cohesionless, sand grains move
downslope in thin surface layers. There is no nip at
the base of the slope.

2.

12.
13.

57
Site 89

a) Muskegon

b) Montague, 15', 1959

c) The original chained distance was from the NE corner/
sec. 31/T11N,R17W: however, in Powers' notes, a
measurement is indicated for the distance to the
bluff's edge from the intersection of McMillan Road
and Scenic Drive.

a) 1837

b) 4,H88 ft.
0) 169 ft.
d) 1.41 ft.

a) 679 ft. - from the intersection of MCMillan Road
and Scenic Drive.
(In 1956 Powers had measured this distance as 675 ft.
Material definitely has not been added to the bluff.
The bluff slope is presently stable and appears to
have been so for some years. It is assumed that there
has been no loss of bluff since 1956.)

C) 0 ft.

d) 0 ft.

Medium sand with scattered pebbles
45 ft.
30—37 degrees

. Vegetative cover of small trees, bushes and grasses

None
N27W
a) 10 ft.
b) 1+ ft.

Permeable groins extend lakeward at the site and to the
north and south (these did not exist in 1956). From one-
hundred ards south of the site-line no groins occur for
another. O yards and erosion is severe here. This ero—
sion appears to be precipitated by refraction of incoming
waves off of an impermeable wooded groin at the southern
limit of this #0 yard shoreline stretch.

Stable. A two-foot nip is cut into the base of the slope.

Since wave swash may now reach the bluff's base during
periods of gusty winds, the bluff appears to be at a cri-
tical stage. During a storm waves could become very de-
structive on the bluff slope, removing much of the vege-
tation which is now acting as a sediment stabilizer.

12.

13.

58
Site 101

a) Ottawa
b) Holland, 15'. 1929
c) 8% post/sec. 16/T5N,R16W

a) 1832 ft.
b) 92% ft.

c) 66 ft.
d) 0.53 ft.

a) 8H9 ft.

b) 9 ft.

c) 0.53 ft.

2- ft. wind-blown sand

2 ft. medium to coarse sand; mostly horizontally bed-
ded, some crossbedding

30 ft.

30-34 degrees; step-like

None

None

N-S

a) None
b) None

None

Unstable and retreating. Sand runs (debris slides) occur
on the slope whenever anything disturbs the surface of
the bluff slope.

Any nip carved by waves in the base of the bluff is
quickly covered by sand sliding down from upSIOpe.

11.

12.

13.

59
Site 106

a) Allegan
b) Fennville, 15', 1928
c) Ni post/sec. 30/T2N,R16W

a) 1831

b) 1,051.38 ft.
c) 130 ft.

d) 1.03 ft.

a) 918.5 ft.
b) 2.88 ft.
c) 0.17 ft.

g ft. fine sand and gravel
O ft. gray till, clayey with pebbles and scattered
cobbles

55 ft.
35-49 degrees

Grass and small trees cover the slope with the excep-
tion of a 10-12 foot thick section six feet from the top.

water percolates out from the bluff slope at the till
and sand and gravel interface.

N15E

a.) 8‘10 ft.
b) None

None except for a steel retaining wall which is leaning
lakeward 150 feet north of the site line.

It is relatively stable along the upper edge of the bluff
but evidence of base erosion and slumping and block
gliding on the slope exists. Wave swash reaches the base
of the slope. A four-foot nip is cut into a till slump
or glide block at the base of the slope.

A local resident said that this particular site has been
relatively stable for the last 15-2O years, but back then,
however, a major collapse and down sliding occurred. The
slope form appears to reveal previous downward movement
of till blocks on the bluff slope.

10.

11.

12.

60
Site 109

a) Van Buren
b) South Haven, 15', 1928
c) NW corner/sec. 15/T1S,R17W

a) 1837

b) 995.94 ft.
c) 348 ft.

d) 2.7% ft.

a) 607.5 ft.
b) no.84 ft.
c) 2.38 ft.

5% ft. sand, no pebbles

1 ft. sand and gravel

3: ft. gray, massive clay till with pebbles; much of
this part of the slope is covered by a mixture
of till and sand and gravel which has slid down
from upslope

41 ft.

Near vertical from the top to 7% ft. into the till;
the remaining slope measures 35—40 degrees

None

water seepage along the interface of till and sand
and gravel

N13E toward the north
S28W toward the south

a) None
b) None

Concrete slabs are placed at the base of the bluff adja-
cent and south of the survey point. The slope here is
more gentle and portions are covered by vegetation (large—
ly grasses) but slumping and block gliding have occurred
near the top of the bluff. The only structure to the
north is a steel seawall 75 yards from the site.

Erosion is severe. Much of the lower two-thirds of the
slope is covered by material which has slid down from
above. A four-foot nip is cut into in-situ till at the
base of the slope. Erosion is more severe immediately
to the north of the site line and the volume of the
slope material is much less and the slope steeper than
at the site line.

61

Site 109 (continued)

13. The site is located approximately one mile south of the
jetties and waterworks intake at the South Haven harbor
entrance. From a point one-third of a mile south of
the survey point the shoreline to the north curves no-
ticeably inland until the harbor jetties are reached.
Erosion along this stretch is severe (U.S. Department
of the Army, 1971a, p. 86).

11.
12.

62
Site 111

a) Berrien

b) Benton Harbor, 15', 1927
Benton Heights, 75', 1970

0) SE corner/sec. 21/T3S,R18w

a) 3915.5 ft.
This figure is 15.5 feet greater than that measured
by Powers in 1956. The bluff is at least 70 feet
in height at this location. No aggradation has taken
place. Except for minor slumping the bluff appears
to be very stable as is evident by the dense vegeta-
tive cover and medium size trees on the slope. It
is not known whether an error was made by Powers or
this investigator but it will be assumed that there
has been no erosion from the upper edge of the bluff
since 1956.

b) 0 ft.

0) 0 ft.

6-7 ft. sand with widely scattered very small pebbles

60-62 ft. gray pebbly clay till

2-3 ft. sand nip three feet wide in remnants of a low
foredune

70 ft.

30-h5 degrees on the till
55 to near 90 degrees on the upper sand

wooded with relatively dense ground cover

water lay in linear depressions on the mid-slope where
old slump blocks cover the bluff face. The water may
have seeped out from the sand and till interface al—
though this was not evident on the days on which the
site was visited.

N35E

a) 39 ft.
b) 23-25 ft.

None

Stable. Minor mass-wasting appears to have taken place
sometime in the last several years near the top of the
bluff. An exposure three to five feet in vertical ex-
tent is visible straddling the contact between the upper
sand and the till.

63
Site 111 (continued)

13. Examination of the slope form and tree shapes below the
exposure mentioned in number 12 indicates that major
slumping took place an undetermined number of years
ago. Several shallow, vegetated, linear depression
scars lie parallel to the bluff's edge on the upper
third of the till profile. water saturated several of
the depressions. Many trees on this area have bends in
their trunks two-to-two and one-half feet from their
bases.

The small sand nip at the base of the slope is probably
the remnants of a low dune constructed during a low
water stage of Lake Michigan.

64

Site 112

1. a) Berrien
b) Benton Harbor, 15', 1927
Benton Heights, 7%' 1970
c) SE cor/sec 3/T3S,R18w

2. a) 1830
b) 191% ft.

According to Powers, bluff loss amounted to only
a total of 25 feet in the 126 years until 1956. This
investigator is convinced that Powers' survey
measurement was incorrect. At his site 113, approx-
imately one mile south of the present study site, the
1956 survey revealed a bluff loss of 35% feet since
the original land survey. Two and one-half miles
north of site 112 Powers calculated a bluff retreat
total of 390 feet at his site 111 for the period
1830 to 1956. The shoreline is virtually straight
between sites 113 and 111 - this would probably not
have not been the case if the bluff at site 112 had
only retreated 25 feet since 1830.

The 1927 to ographic map reveals a distance of about
1,450 to 1, 75 feet to the bluff's top edge and ap-
proximately 1,700 feet to the bluff's base from the
section corner. The 1970 topographic map of the area
reveals a distance of about 1,325 feet to the bluff's
upper edge whereas the distance was 1,550 feet to

its base.

On the 1927 topographic quadrangle the M-139 highway
marked both the eastern extent of sections 31 and 6.
The section 31 segment was offset to the east of the
section 6 segment. Since then, however, the section
31 highway segment has been shifted westward and a-

ligned with the section 6 road segment.

The old road coinciding with the eastern limit of
section 31 is evident on Powers' 1950 areal photo
but is not revealed presently in the field. Because
of this situation, a 1973 survey was not performed.

 

 

3. ---
4. 12415 ft. gray clayey till with pebbles
2 ft. alt. beds of fine sand and gravel
2 ft. fine to silty sand

 

6-7 ft. gray clay with small pebbles

12-15 ft. no exposure; till has slid over the bluff face
17 ft. silty to clayey sand, with scattered clay balls
20 ft. no exposure; slope is covered by till slide

‘2 ft. med. sand with numerous small pebbles

‘1-12 ft. no exposure; till slide covers slope

‘2 ft. slumped or glided till block covers med. sand

containing small pebbles

 

11.
12.

13.

65

Site 112 (continued)

Two-hundred yards north of the site is an exposure
at the base of the slope of bedded and crossbedded
sand with pebble inclusions and gravel layers. It
is approximately 30 feet thick where exposed.

100-110 ft.
35 to near 90 degrees

MUch of the slope is devoid of vegetation. Small bushes,
grasses and several trees are growing on a linear ridge
between gullies extending down the center of the slope
area.

water percolates out from the area between the two clay
layers. Two drain pipes extend out from the top of the
bluff face.

N36E

a) 25 ft.
b) 20 ft.

None
Unstable and eroding; see number 13 below

The study site is at the southern margin of a 225 foot
wide indentation into the bluff's upper edge. On either
side of the concave area the top bluff line extends lake-
ward another 30 feet. This difference does not exist

at the base of the slope.

The slope extending downward along the 225 foot wide con—
cave area is extensively gullied.

A local resident stated that the bluff edge and slope
has receded each year along the 225 foot section. A
staircase had to be continually repaired due to the un-
stable slope. There are no steps presently. This con-
cave stretch along the top of the bluff has been here
since at least 1956 when the resident moved to the pro-
perty. The 1927 topographic map shows a similar feature,
although its southern margin had not yet receded to the
section line and the present study site.

According to the property owner, the straight line por-
tion of the bluff 25 feet south of the study point had
been much steeper in 1960. Presently its slope is 39-40
degrees. She stated that in 1966 the water table in the
area had changed and water began to seep out from the
bluff. (This change was attributed to heavy building

66
Site 112 (continued)

construction taking place in the vicinity.)

A large till block had slumped down several feet from
the upper edge of the bluff 200 feet south of the study
site this past June. The owner of the lot indicated

that water frequently seeped out from the upper bluff
slope.

1.

A.

67

Site 115

a) Berrien

b) Benton Harbor, 15', 1927

c) Sw corner/sec. 3/T5S,R19w
A curve was added to the road at this point. At a
point in the center of the old abandoned road segment
a sunken metal post under a small metal cover marked
by a painted yellow X was found. It is believed that
this post indicated the location of the section corner.

The site is #00 feet north of the northern limit of
the case study bluff segment and much of the descrip-
tion presented here applies to that location.

a) 1829

b) 957 ft.
c) 1A2 ft.
d) 1.11 ft.

a) 734 ft.
b) 81 ft.
0) 8.76 ft.

6 ft. fine silty sand, pebbles strewn on top

38 ft. medium clean sand with a few scattered pebbles,
crossbedded in places but predominately hori-
zontally bedded

10 ft. fine sand and silt

6 ft. covered

This profile was abstracted from Powers' field notes
as the present condition of the bluff prohibited a safe
examination of the sediments.

60 ft.

30—35 degrees; near vertical in the upper 3-h ft.
None

None

N27E

a) None
b) None

A half star shaped low breakwater was built three years
ago at the water’s edge immediately adjacent and south
of the survey site line. Another breakwater is posi-
tioned about 75 yards to the north. A concrete barrier,
built last year, protects the bluff 850 feet south of
the study location. Lakeward of this barrier is the de-

12.

13.

68
Site 115 (continued)

caying remains of a wooden permeable breakwater built
in the 1950's.

Erosion is extremely severe. The slightest disturbance
on the slope will cause sand runs on the bluff face.

The study site lies in a concave inland segment of the
shoreline bluff extending about 425 yards in length.
This bluff segment lies approximately four miles south
of the jetties at the St. Joseph harbor. Littoral drift
is southerly but the jetties effectively prohibit move-
ment of sediment south of the harbor entrance (U.S.
House of Representatives, 1958, p. n).

VARIATIONS
IN
WATER LEVEL AND PRECIPITATION
IN THE
LAKE MICHIGAN BASIN

By
William R. Buckler

A.RESEARCH PAPER

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS
Department of Geography
1973

ACKNOWLEDGMENTS

The author wishes to express appreciation to Dr. Jay R.
Harman for his advice and guidance in the preparation of this
manuscript. Special thanks is extended to Mrs. Susan.Buckler
for typing the paper and Mr. David P. Lusch for preparing the

illustrations.

ii

TABLE OF CONTENTS

Page

LII'D‘T OF TA-BLES........................ 00000 ........... i'J
LIST OF FIGUR‘ES............. ..... ....... ........ ...... V
INTRODUCTION................. ......... .... ..... .. ..... 1
PIJRPOSE........................ ........... ... ......... 1
BASIC DATA................................. ..... ...... 2
LAKE MICHIGAN—HURON BASIN CHARACTSRISTICS... .......... 3
LAKE MICHIGAN-HURON WATER SUPPLIES. .................. . 5

Precipitation and Runoff

Evaporation

Diversions

Inflow

Outlet Conditions
Effects of Diversion, Inflows and Outflows
Groundwater Inflow and Seepage

SURFACE PRECIPITATION, LAKE LEVELS AND UPPER AIR FLOW.. 27
Mid-Latitude Upper Air Flow and Associated

Surface weather Conditions
Investigative Hypothesis

Data
Method
Discussion of the Findings
Conclusion
LIST OF REFERENCES ..... ................ ........ ....... 1+1

APPENDI:{.............................................. 1+6

Non-Hydrological Elements Affecting the Lake
Michigan-Huron water Level

iii

Table

LIST OF TABLES

Page

The mean annual water level and precipitation
amount and their departures from normal in the
Lake Michigan-Huron basin, l875-l972............. 7

The mean basin runoff in cubic feet per second
per square mile, 1935-1961+0000000000000.0.0000...16

The average monthly runoff in inches depth on
la1{e, 1935-1961}..................................16

Departures from normal of the annual water level

and precipitation amounts and the mean monthly

500 mb trough axis positions over the Lake Mich-
igan baSin, 1950-1971.0.00000000000000000.00.0000 3?

The mean monflily 500 mb trough axis positions

for the years when the annual precipitation

amounts were above and below normal over the

Lake Michigan-Huron basin, l950-l971............. 38

The mean monthly 500 mb trough axis positions

over the Lake Michigan-Huron basin for the ex-
tended periods of above normal water levels,
1950-195% and 1968-1971, and below normal water
levels, l955-l967................................ 38

iv

LIST OF FIGURES

Figure

1.
2.

The Great Lakes region..........................

The average annual Lake Michigan-Huron water
levels and basin precipitation amounts,
1875—1972..................................O..0.

The computed average annual evaporation for
Ialce Ificnigan-Huron.......O.....................

Mean 700 mb contours (tens of feet) for July,

1966............................................

A schematic diagram of a tropospheric wave and
aSSOCiated air flOWOOO..........................

The relationship between precipitation and mean
Wind flow Fit the 500 mb Surface.................

10

19

31

33

INTRODUCTION

Much concern in the last several years has been focused
on shoreline erosion in the Great Lakes. This attention has
been fostered to a large extent by man's intense development
of much of the shoreline. This development depends on the
stability of the lake system's levels. The conditions
giving rise to the present erosion and resulting economic
losses along the eastern shoreline of Lake Michigan are re-
lated to both the unconsolidated glacial sediments composing
the lake shore areas and the nonperiodic trends in the
regional climate. In the continuing natural processes which
act on the Lake Michigan shoreline, erosion occurs at all
stages of lake levels. During periods of high levels, how-
ever, the rate is accelerated and the extent is greatly

expanded.

PURPOSE
The purpose of this paper is to describe and account for
those natural and man-created factors which affect Lake
Michigan's fluctuating water level. Precipitation rates over
the Lake Michigan-Huron basin will be considered in detail
since these rates are probably most responsible for fluc-
tuations in the lake level. An attempt also will be made to

explain the cause for the long-term and annual changes in the

precipitation amounts over the basin.

“th- and upper-level tropospheric air flows are
generally responsible for mid-latitude weather phenomena:
Therefore, mean monthly 500 mb air flow patterns over the
Michigan-Huron basin will be examined for the period 1950
to 1971 in order to determine if the varying precipitation
rates can be at least partially explained due to changing

air flow patterns.

BASIC DATA

Basic data relating to the Great Lakes water levels are
routinely compiled by the National Ocean Survey, Lake Survey
Center, formerly under the U. S. Corps of Engineers but pre-
sently under the auspices of the National Oceanic and
Atmospheric Administration (NOAA) in the U. S. Department of
Commerce. Mean monthly and annual Lake Michigan-Huron water
levels and diversions into and out of the lakes by way of the
St. Marys and St. Clair Rivers, respectively, along with the
mean annual precipitation amounts for the basins were ob-
tained from the Lake Survey Center for the period of record.
Reliable basin precipitation data prior to 1900 but after
187%, not available from the Center, were obtained from Day
(1926). Dependable Lake Michigan-Huron precipitation infor-
mation before 187% is unobtainable. The U. S. Commerce
Department's monthly publication - Climatological Datai
National Summary - was utilized in order to record and plot
the mean monthly 500 mb surface wind flow directions and

trough axis positions for the period of record over the

3

Lake Michigan-Huron basin. Mid- and upper-tropospheric air

flow data have only been published since 1950.

LAKE MICHIGAN-HURON BASIN CHARACTERISTICS

Prior to the onset of late Cenozoic glaciation the
present Lake Michigan basin consisted of a series of sub-
sequent stream valleys which were related to the general
drainage of the mid-continent (Dorr and Eschman, 1970). The
four major ice advancements enlarged the existing lowlands
leaving behind, upon retreat of the ice masses, several
proglacial lakes. The present-day Lake Michigan lies in the
west-central portion of the-Great Lakes basin, south of Lake
Superior and west of Lake Huron (Figure 1).$ Hydrologically
Lakes Michigan and Huron are considered to be a single lake
due to their wide and deep connection at the Straits of
Mackinac. Their combined drainage area of 1%2,700 square
miles includes %5,300 square miles of water surface. The
average depth of Lake Huron is 195 feet, whereas the average
for Lake Michigan is 279 feet, with a maximum.depth of 923
feet. Their combined volume of water totals 2,029 cubic
miles (Corps of Engineers, 1971).?

Direction of currents in the Straits alternates from
east to west depending on barometric pressure and wind
direction; however, the net flow is eastward from Lake
Michigan to Lake Huron (Great Lakes Basin Commission, 1972).
Since 18%8 water has been diverted from Lake Michigan to the
Mississippi River drainage basin via the Chicago drainage

canal. The St. Marys River drains into Lake Huron at the

.QOHwoh mmxmq pmoau 639 "P madmflm

 

 

(21...... \. I 202......

       

JV .<z<u ...: 9.3
>552: 003:8

- l'!

axmgmpm.2.;ga.s_e. nu¢utus_:§§§x

20.3mm
mu¥<a .5380

.81 a
ZOO—22

.7
...: ...,
. .52.... .808
’3. . ./, ,

 

 

Vt

north from Lake Superior and the St. Clair River diverts
water out of the basin at the south into the lower Great

Lakes.

LAKE MICHIGAN-HURON WATER SUPPLIES
At a given time the water level in the Lake Michigan-
Huron basin depends primarily on whether the lakes are
receiving more or less water than they are giving up.‘ The
water supply consists of precipitation on the lake surface,
runoff from the lake's drainage area, inflow from a lake
above, diversion of water into the basin, and ground water
inflow. water is removed from the lake by evaporation,
diversion to another drainage basin, outflow from the lake
through its natural outlet, and ground water seepage.”
DeCooke (1968 a,b), Laidly (1962) and the Great Lakes
Basin Commission (1972), among others, have expressed the
interrelationships of these factors in the form of a
"hydrological balance equation":
AC=’(P+R+I+Di+Gi+Ic)-(E+O+Do+Gs+Ic)
where
AC = change in the amount of water stored in the lake basin
P = precipitation on the lake's surface
R = runoff from the drainage basin
I = inflow from a lake above
Di = diversion into the lake from another basin
Gi = inflow of ground water
E = evaporation from the lake surface

0 = outflow from the lake through its natural outlet

0

Do = diversion from the lake into another basin

G5 = groundwater seepage

I0 = ice blockage of outlet channel

In addition to these hydrological factors which con-
tribute to a specific lake level, several subordinate
non-hydrological factors affect the Lake Michigan—Huron level
although not in themselves changing the volume of water in
the basin. These factors include crustal movement, wind,
barometric pressure changes, and tides and are discussed in

the appendix.

Precipitation and Runoff

Approximately 70 percent of the contemporary variation
in the Lake Michigan-Huron level is related to basin pre-
cipitation (Muller et al, 1965; Brunk, 1960). The
relationship is not simple for the many variables in the
"hydrological equation" and the vastness of the Lake Michigan
basin play supporting roles. Nevertheless, over the long run,
the water level rises and falls in relation to rises and falls
in the precipitation rates over the drainage area (Table 1
and Figure 2), and it is these fluctuations which are the
subject of this paper.

The U. S. Lake Survey computes the monthly and annual
precipitation amounts over all of the Great Lakes from data
supplied by the U. S. National weather Bureau and the
Meteorological Branch of the Department of Environment,
Canada. The record extends back to 1900 for the Lake
Michigan-Huron basin.‘5Day (1926) made a comprehensive study

of the precipitation regimes in the Great Lakes area for the

7

Table 1: The mean annual water 1 ‘

. evel and prec1 it t'
amount and their departures from normgl Inltfie
Lake Michigan-Huron basin, 1875-1972.

 

 

 

Mean Annual Departure Mean Annual Departure
Year water Level From the Precipita— From the
(in feet) Normal tion Amounts Normal
(in feet) (in inches) (in inches)
1875 579.90 1.42
1876 580.94 2.46 37:88 6°92
1877 580.65 2.17 34.34 2'38
1878 580.40 1.92 37.13 5'17
1879 579.52 1.04 33.46 1'50
1880 579.96 1.21 39.21 7°25
1881 580.07 1.59 39.81“ 7'85
1882 580.42 1.94 38.40 6'44
1883 580.69 2.20 39.42 7°46
1884 580.90 2.42 37.67 5°71
1885 581.06 2.58 35.77 3°81
1886 581.28 2.80 33.61 1°65
1887 580.63 2.15 30.66 —1°30
1888 579.98 1.50 28.93 -3°03
1889 579.48 1.00 29.44 -2°52
1890 579.31 .83 33.47 1°51
1891 578.74 .26 30.88 -1°08
1892 578.58 .10 34.95 2°99
1893 578.76 .28 34.01 2°05
1894 578.96 .48 29.06 -2°90
1895 578.10 - .38 26.52 -5°44
1896 577.79 - .69 31.58 - .38
1897 578.40 — .08 32.87 °91
1898 578.55 .07 33.50 1°54
1899 578.56 .08 30.24 -1°72
1900 578.57 .09 32.31 °35
1901 578.83 .35 28.91 -3°05
1902 578.47 - .01 32.51 °55
1903 578.61 .13 33.24 1°28
1904 579.09 .61 31.04 — °92
1905 579.21 .73 33.75 1°79
1906 579.23 .75 31.90 - '06
1907 579.30 .82 29.79 -2'17
1908 579.21 .73 29.16 -2°80
1909 578.69 .21 31.35 - °61
1910 578. 7 - .11 28.09 -3°87
1911 577. 3 - .65 32.37 °41
1912 578.29 - .19 32.24 °28
1913 578.92 .44 29.02 -2°94
577.89 - .59 29.37 -2.59

 

 

 

 

 

 

 

 

Table 1: Continued
Departure Mean Annual Departure
Mean Annual From the Precipita- From the

Year water Level .

(in feet) Normal tion Amounts Normal

(in feet) (in inches) (in inches)

1916 578.60 .12 33.70 1.74
1917 579.34 .86 27.78 —4.18
1918 579.55 1.07 31.06 - .90
1919 579.07 .59 30.47 -1.49
1920 578.70 .22 28.43 -3.5
1921 578027 " 021 31072 " .2
1922 578013 " 035 31080 - .16
1923 577.52 - .96 28.24 -3.72
1924 577.27 -1.21 30.81 -1.15
1925 576.39 '2009 26021 -5075
1927 577.18 -1.30 31.40 - .56
1928 578.17 - .31 35.50 3.54
1929 579.65 1.17 31.53 - .43
1930 578.85 .37 24.37 -7.59
1931 577.07 -10 1 30068 '1028
1932 576.51 -1.97 31.77 - .19
193 576.2 -2.23 30.49 -1.47
193 575.9 -2.54 27.31 -4.65
1935 576.34 -2.14 28.76 -3.20
1936 576.30 -1.98 29.30 -2.66
1937 576. E -2.03 30.52 -1.44
1938 577.1 -1.34 33.75 1.79
1939 577.58 - .90 29.71 -2.25
1940 577.16 -1.32 33.53 1.57
1941 577.09 -1.39 33.03 1.07
1942 577.81 - .67 34.77 2.81
19% 578.77 .29 32.94 .98
194 578.66 .18 29.94 -2.02
1945 578.57 .09 36.18 4.22
1946 578.68 .20 28.61 -3.35
1947 578.39 .11 33.64 1.68
19"1'8 578. L." " 0011' 28.2” -3072
1949 577.40 -1.08 31.11 - .82
1950 577.58 - .90 3 .70 1.7
1951 579.14 .66 3 .02 6.06
1953 579.79 1.31 30.61 - .&6
195% 579.47 .99 37.44 5. 8

 

 

 

 

 

 

 

\O

 

 

 

 

 

 

 

Table 1: Continued
Departure Mean Annual Departure
Y ar aginrAfiggzi From the Precipita- From the
e (.e f t) Normal tion Amounts Normal
1n ee (in feet) (in inches) (in inches)

1956 578018 " .30 29019 “2077
1957 577.60 - .82 32.95 .99
1958 577.01 -1.%7 26.15 -5.81
1959 576.76 -1.72 37-55 5.59
1961 577097 " 051 30.97 " 099
1962 577.45 -1.03 27.79 -4.17
196 576.48 “2.00 26012 ‘508Ll’
196 575.66 -2.82 29.81 -2.15
1966 577.25 -1.23 29.88 -2.08
1967 577.69 - .79 3%.99 3.03
1968 578.18 — 30 34.11 2.15
1969 579.01 .33 32.27 .31
1970 578.93 . 5 33.90 1.9%
1971 579.36 .88 30.67 -1.29
1972 579.66 1.18 35.1% 3.18

 

 

1O

1mpflmaoepa

manna use

at

no:

a:

.mmmFImmww .mpfl5o8m GOHP
mHo>oH Hopmz dogfim1meH30Hz oxmq assume owmao>m 03H

#6

 

UILVIA

13A31

fine

:3. .84.

 

. . . a a a a _

 

 

 

 

4Aqfi1~fi_q.

 

 

 

 

d§2( 240!

Ohm!

5.1

I!!!

 

 

 

 

 

 

. . _ . . 11. . .

 

 

____

 

 

 

 

q

_

q A

 

 

 

 

 

“m mafiwfim
QA’K
....._....Q«
l
E ,. m
=3. :3. m
a
I. S
.28st 30..-
29.38.92: 1 an
.0350! 03¢
L400
0
‘
«01:92:53..
:33... 231
u
I.
D
d
in
m
a.

 

 

 

11

UILVM

13A“

I)! .83

I!!!

 

mscwpcoo .m e

...-...: so...
20.2.8.8!

SIHDNI

...-8: 83¢

NOINLHDIIH

admflm

12

years 1875 to 192%. He noted that when official daily
weather statistics were initially collected in the late
1870's more stations per unit area were established in the
Great Lakes region than in other parts of the country.

Even though he used a total of only 39 control stations for
the Lake Michigan—Huron basin, as compared to the present
total of approximately 200, his results are considered
representative (Brunk, 1962):] Indeed, a comparison for the
period 1900 to 192% between the Lake Survey figures and
Day's figures reveals only a 2.5 percent higher precipitation
value in Day's yearly averages. Day's computed annual pre-
cipitation rates for 1875 to 1899 are utilized in this study
to extend the Lake Survey Center's figures back to 1875.

The record for the Lake Michigan-Huron water level will
assume an initiation date of 1875 for this paper. This will
allow favorable comparison between the two sets of variables
(Table 1).

\'Between 1875 and 1972 the mean Lake Michigan-Huron
level was 578.%8 feet1 while the average annual precipitation
over the basin was 31.96 inches. This compares to the 1860
to 1972 figure of 578.69 for the lake level. The annual
average precipitation for the Lake Huron basin was 32.08
inches while that of the Lake Michigan basin was 31.8%
inches. Utilizing only the Lake Survey data, 1900 to 1972,

the Lake Michigan-Huron level averaged 578.08 feet whereas

 

1All water levels are referenced to the 1955 Inter-
national Great Lakes Datum.

13

the annual mean precipitation measured 31.29 inches.

For the period of record, monthly mean levels for Lake
NHchigan-Huron have varied from a low of 575.35 feet in
March, 196%, to a high of 581.9% feet in June,1886, a
difference of 6.59 feet. Both Freeman (1926) and Day (1926)
cite the Board of Engineers on Deep waterways report
(Secretary of war, 1900) referring to the early levels of the
Great Lakes. The highest authenticated annual stage of Lake
Michigan - 58%.3 feet - occurred in 1838 and was 5.61 feet
above normal, %.6% feet above the 1972 level.

Even though precipitation varies considerably from year
to year, yearly totals oscillate around a norm. Periods
of abundant precipitation are usually followed by others of
lower rainfall. “Since 1875 21 periods have occurred, from
one to nine years in length, in which precipitation amounts
were below average, separated by intervals of above normal
precipitation of from one to twelve years in duration. In
total, nine different periods of high water levels and eight
distinct intervals of below average Lake Michigan-Huron
levels have occurred since 1875 (Figure 2).“

° In the Great Lakes drainage system considerable lag
occurs between precipitation and fluctuations of lake level.
Much of this is probably due to the unconsolidated glacial
material which comprises much of the drainage basin (Freeman,
1926). The sand, gravel and till, when unfrozen, absorb and
retain much of the rainfall as it falls.” Percolation may

take several years as the water migrates to the various

14

\\
stream beds and into the lakes. As reported by Richards

(1969), Brunk (1960) suggests that precipitation of the pre-
sent and preceding year contributes nine percent and 37
percent, respectively, to Lake Michigan-Huron levels, while
precipitation of two and three years earlier contribute 3%
and 20 percent, respectivelyj°

The excess rainfall in 1928 over the basin increased
the Michigan-Huron level by 1.%8 feet. QIn 1930, however,
the greatest deficit in precipitation for the period of
record occurred and caused a lowering of the level by 1.78
feet.fi/The 1959 surplus initiated a rise of 1.52 feet in
1960 although the water level was still below normal. A
deficit of %.17 inches in the 1962 yearly mean precipitation
amount was followed by a lowering of the lake by about a foot
in 1963. A further deficit of 5.8% inches in 1963 led to an
additional drop in the following year of .82 foot. The
above normal precipitation in 1972 is expected to cause an
increase in the annual Michigan-Huron level by at least .5
foot.

Approximately two-thirds of the Lake Michigan-Huron
basin is covered by land area. Although the annual amount
of precipitation has a large bearing on the total runoff in
the basin, seasonal distribution of precipitation and the
north-south temperature gradient are equally important
(Pentland, 1968, p. 328):Q°Runoff is generally at its peak
during March, April and May as the winter snow accumulation

melts and while evaporation rates from the land surface are

15

low. Runoff is least in August and September due to the high
evapotranspiration losses. Increasing runoff in October and
November occurs because of the sharp drop in evapotranspira-
tion (Tables 2 and 3):?

Above average rainfall can occur in the summer season
without adversely affecting runoff and lake levels if it is
evenly distributed. The abundant rainfall would go mainly

into vegetative growth.<°Lesser amounts of rainfall, however,
may produce much greater runoff if it falls on saturated or
frozen ground. /A situation in 195% is considered unique.
Heavy October rains resulted in immediate runoff which
caused an October maximum in the Lake Michigan-Huron level,
then held the next winter's water level up, but did not pro-
duce a rise in 1955 (Eichmeier, 196%, p. 20).

Q Since 30 percent of the Michigan-Huron surface is water,
precipitation falling directly on the lake affects the lake
level immediately.'°Since the late 1920's, however, there
has been disagreement as to whether the rate of precipitation
falling on the water surface is approximately equal to, less
than, or greater than that falling on the remaining land
surface of the drainage area.

Freeman (1926) believed that there was no difference in
precipitation rates over the land and water surfaces. Using
island and land station data in Lake Michigan for the years
1871 to 1927, Horton and Grumsley (1927) concluded that on
Lake Michigan winter precipitation averaged 93 percent of

that at shore stations and summer precipitation 9% percent

16

Table 2: The mean basin runoff in cubic feet per second per
square mile, 1935-196%.

 

 

 

Month Lake Michigan Basin Lake Huron Basin
January 0.75 0.62
February 0.72 0.67
March 1.16 1.%3
April 1.72 2.6%
May 1.17 1.70
June 0.80 0.9%
July 0.37 0.65
August 0. 9 O.%%
September 0.5% O.%6
October 0.60 0.61
November 0.69 0.86
December 0.60 0.8;
ANNUAL 9.81 11.

 

 

 

Source: Pentland, R.L. 1968. Runoff characteristics in the

Great Lakes basin. Proc., 11th Conf., Great
Lakes Res., Intern. Assoc. Great Lakes Res.,

P-331-

 

Table 3: The average monthly runoff in inches depth on lake,

 

 

Month Lake Michigan Basin Lake Huron Basin
January 1.77 1.61
February 1.5% 1.58
March 2.73 3.71
April 3.92 6.63
May 2.76 %.%1
June 1.83 2.36
July 1.3% 1.69
August 1.16 1.1%
September 1.23 1.16
October 1.41 1.58
November 1.37 2.16
December 1. 1 2.16
ANNUAL 22.67 30.19

 

 

 

 

Source: Great Lakes Basin Commission. 1972 (Final Draft).

Levels and flows. Appendix No.11, Great Lakes

Baéén Framework Study. Ann Arbor, Michigan,
p. .

 

 

 

17

of the shore station averages. They credited this reduction
to the lower thunderstorm frequencies and intensities over
the lake. Day (1926) found that less precipitation was
registered in gages on islands in northern Lake Michigan than
on the adjacent mainland. Nevertheless, he attributed the
difference to increased wind velocities over the lake
materially lessening the catch of the gages. He therefore
assumed the precipitation over the land and the water was
approximately equal. As reported by Changnon (1976), Kohler
(1959) and Blust and DeCooke (1960) have made similar state-
ments more recently. Brunk (1962) noted that the yearly
amounts of precipitation over the Michigan-Huron basin land
and water surface were of little difference.

In a comprehensive Illinois study Changnon (1967, 1968)
estimated that the average precipitation over Lake Michigan
is six percent less than that over the land portion of the
basin. He believes that ”the results of this study are more
accurate than the earlier findings because they are based on
more detailed information from recent lake precipitation
studies, a greater knowledge of the effects of the lake on
the atmosphere and a greater volume of weather data than
. existed when the previous studies were made" ( 1968, P-1).

The significance of the accuracy in cor—
rectly estimating or calculating lake
precipitation can be realized by comparing the
basin water yield based on the assumption that
there was no lake difference with that based on
the six percent difference. If the average value
is 31.12 inches, then the average precipitation

over the entire basin produces 37,009 billion
gallons of water annually. If the average basin

18

precipitation is compiled using the lake value

that is six percent lower than the land value,

the water produced annually by precipitation is

36,309 billion galllons, a difference of 700

billion gallons annually. This difference be—

tween the two values represents 2,970 cfs (p. %%).
Evaporation

Although precipitation plays the major role in modifying
contemporary lake levels, evaporation also plays a significant
role. "The importance of considering evaporation becomes
plain when one realizes that there are long periods in each
year during which more water goes up into the air from the
Great Lakes by evaporation than the amount which flows down
the Niagara or St. Lawrence Rivers" (Freeman, 1926, p.XVII).

Conditions which affect evaporation from the Great
Lakes are very different from those which control evaporation
from ordinary lakes and reservoirs because of their vast size
and heat storage capacity (Freeman, 1926).ԤIn winter, when
the lake surfaces are relatively warmer than the enveloping
air masses, evaporation occurs at a much more rapid rate than
from the surface of ice or snow. Just the reverse is true
in early summer. During this period the water of the Great
Lakes is usually colder than the surrounding air. This
condition greatly retards evaporation.’ Freeman pointed out
that the greatest evaporation takes place in the fall or
winter and the smallest in the late spring and early summer
over the Great Lakes district (Figure 3):(

Because of its vast size and storage capacity and lack

of over-water data, truly accurate evaporation rates have

19

 

 

 

 

 

 

 

 

 

 

I
*Irill
I I I I I
3 I I I I I
U)
1
LZD2IIIII
F4 I I 71 I I
:2
H
5
£4 L l I | I I I I
:31! IIIIITII1
C)
a.
<r:
>
[:3
I I I I I I I I I
0II I IIIIIO
-1IIIIIIIIII_1
NIH I—Jt/I {>154 ma:
M525 Mama.
EDS<€ F3 55.1.3 ....
2.1%“ aé‘iofifl‘é’i
Pam CLOCK!
In Ed 20
U)
Source: Great Lakes Basin Commission. 1972 (Final Draft).
Levels and flows. Appendix No. 11, Great Lakes
Basin Framework Study. Ann Arbor, Michigan, p.80.

 

 

Figure 3: The computed average annual evaporation for Lake
Michigan-Huron.

 

20

not been determined for Lake Michigan-Huron. By subtracting
the measured inflow from the measured outflow, Freeman

(1926, pp. 1%2-l%3) estimated an average annual evaporation
rate of 25.%7 inches for the water surface of Lake Michigan-
Huron. By another computation based on the vapor pressure
differences of air and water and on the velocity of the wind
he verified the previous "water budget" estimate with a "mass
transfer" figure of 29.12 inches. ”Energy budget" estimates
and evaporation pan observations have been utilized but have
yet to be proven effective in Lake Michigan-Huron evaporation
studies (Richards, 1969, pp. 62-6%; Bruce and Rodgers, 1962,
pp. 5%-65). Investigators have most often used the water
budget and mass transfer methods. The Great Lakes Basin
Commission (1972, p. 89) noted that the more recent inves-
tigators on Lake Michigan-Huron have corroborated Freeman's

estimate of approximately 26 inches.

Diversions

Diversion of water to and from one drainage system into
another has an effect on the water flows and the lake levels
within the second system. A diversion to or from a single
lake within a drainage basin has a consequence on those lakes
upstream and downstream within the same system.

If water is continuously diverted into a drainage sys-
tem from another, the receiving system will eventually
convert this additional water into higher lake levels and

greater outflow. The reverse is true if water is continuously

21

diverted from a drainage system. A diversion into a par-
ticular lake within a chain of lakes, such as the Great
Lakes, will cause an increase in outflow and water levels
downstream while at the same time may cause increased water
levels upstream. Both outflow and water levels upstream and
downstream will decrease if water is diverted out of a lake
within a particular drainage system.

Lake Michigan-Huron is affected by two primary diversions.
water is diverted out of Lake Michigan into the Mississippi
River drainage system at Chicago while water is diverted in-
to Lake Superior which is further upstream in the Great Lakes
system.4

A portion of the area originally draining into Hudson
Bay by way of the Albany River now drains into northern Lake
Superior (Figure 1). This was precipitated by the Long Lake
and 0goki projects. In 1939 water was diverted from the
Kenogami River through Long Lake to help flush pulp logs
through Long Lake and the Aguasabon River to Lake Superior.

A hydroelectric power plant was constructed in 19%8 on the
Aguasabon River to utilize the increased flow. The 0goki
power project diverts water, initially draining to the north,
through the Albany River to the south via the Little Jackfish
River which flows into Lake Nipigon and thence into Lake
Superior.

Since 19%0 the water supply to Long Lake has averaged
approximately 1,700 cfs. Of this, 1,%50 cfs has been diverted
to Lake Superior (Great Lakes Basin Commission, 1972, p. 93).

22

The 0goki Reservoir has had an average inflow of 5,000 cfs
since 19%3. Approximately %,000 cfs of this is diverted to
Lake Superior (Great Lakes Basin Commission, 1972, p. 96).
Kirshner (1968, p. 296) has estimated that the 0goki-Long
Lake diversions into Lake Superior have effected an increase
downstream in Lake Michigan-Huron water levels of about .37
foot in spite of the regulatory works in the St. Marys
River.2

The diversion of water into the Lake Michigan—Huron basin
via the 0goki-Long Lake projects is partially compensated for
by diversion out of its basin by way of the Chicago Sanitary
and Ship Canal. This canal was completed in 1900 and by 1910
had forced the abandonment of the Illinois and Michigan Canal,
completed in 18%8, which had first diverted water from the
Lake Michigan drainage basin at Chicago. The present diver-
sion includes water diverted directly from Lake Michigan in
addition to water from the drainage basin in the Chicago
area which normally would flow into the lake. As initially
conceived, the diversion was for the purpose of dilution and
subsequent flushing of sewage into the Des Plaines River
which connected to the Mississippi River system. The water
now also serves as a navigation route to the Illinois River
from Lake Michigan.

Prior to 1938, up to 10,000 cfs of Lake Michigan water
flowed through the canal but in 1938 the U. S. Supreme Court

 

2Recently the 0goki diversion project has been shut
down.

23

limited the amount to 3,100 cfs. In 1967 this was revised
upward to 3,200 cfs, 1,700 cfs of which is pumped for
domestic and industrial use while the remainder includes the
surface runoff which originally flowed into Lake Michigan and
the water diverted directly from Lake Michigan for navi-
gational purposes.

The effect of this diversion at Chicago on Lake Michigan-
Huron is not an instantaneous one. Due to the great storage
capacity of the basin considerable time is necessary for the
readjustment of lake level because of the diversion. Ramey
(1952 b, p. %) estimated that it took five years for Lake
Michigan-Huron to reach 9% percent of its ultimate lowering
due to the diversion. Freeman (1926, p. 218) had estimated
91 percent of maximal lowering after five years.9 Presently,
outflow from Lake Michigan-Huron at Chicago has lowered the
water level approximately .23 foot (Kirshner, 1968, p. 296;
Great Lakes Basin Commission, 1972, p. 98). With a greater
diversion out of the drainage system in the early part of the
century Freeman (1926, p. 218) estimated that the total
lowering had been about .%2 foot by 1926. The maximum lower-
ing of Lake Michigan-Huron due to the Chicago diversion was
attained in 1926 when the water level was lowered by .5 foot
(Ramey, 1952 b, p. %).

As was implied previously, water diverted into one lake,
in a chain of lakes, affects all the water levels. This
teleconnection is evident with respect to Lake Michigan—Huron

and the welland Canal. water is diverted through this canal

91+

L.

from Lake Ontario, two lakes downstream from the Lake Michigan-
Huron basin. According to figures cited by Kirshner (1968,

p. 296),this diversion ultimately has lowered the level of

Lake Michigan-Huron by about .10 foot.

Inflow

Discharge from Lake Superior flows through the St. Marys
River to Lake Huron, whose surface is 22 feet below that of
the upper Great Lake. The rate of outflow from Lake Superior
has averaged approximately 75,000 cfs. Since 1922 this dis-
charge has been regulated by means of power and navigation
canals and a grated dam at Sault Ste. Marie. In accord with
international regulations the outflow from Lake Superior into
Lake Huron is kept as close to natural conditions as possible,
although the stage of Lake Superior is to be held as nearly
as possible between elevation 600.5 and 602.0 feet. In
addition, the discharge is to be so controlled at Sault Ste.
Marie that the river level below the lake does not exceed
elevation 582.9 feet (United States Senate, 1957, p. 15).
Because of the rapids at Sault Ste. Marie, with a drop of
about 20 feet in a mile-long reach, changes in the levels
of Lake Michigan-Huron can have no effect on the outflow
from Lake Superior or on its level (United State Senate,

1957: Po 25)-

Outlet Conditions
The St. Clair River drains the water of the Lake

Michigan-Huron basin into the lower Great Lakes. Increases

25

in its cross-dimensional area would increase its carrying
capacity and would therefore allow greater discharge from the
basin than would occur otherwise. Natural erosion on the St.
Clair River gravel and sand bed through the Port Huron rapids
prior to 1900 was sufficient to cause a lowering of the Lake
Huron level by .3% foot. "Such erosion was probably promoted
by the dredging in Lake Huron, above, and in the St. Clair
Flats, below. Shoaling caused by the wrecks of two schooners
in the rapids in 1900 reduced this lowering to .2% foot"
(Ramey, 1952 a, p. 28). Between 1908 and 1925 commercial
dredging of gravel in the river enlarged the cross-dimensional
area enough to lower the Lake Michigan-Huron level by an
additional .29 foot (Great Lakes Basin Commission, 1972,

p. 106; Ramey, 1952 a, p. 28).

\IThe Joint Board of Engineers on the St. Lawrence water-
way Project (1926) estimated that the levels of Lake Michigan-
Huron have been permanently lowered by approximately .6 foot
since neither of the two influences have been compensated
for by any type of river structure./I

Navigational improvements in the St. Clair River com-
pleted in 1933 and 1962 further enlarged its discharge
capacity. Korkigian (1963, p. 6) cites a figure of .13 foot
at mean flow for the lowering of Lake Michigan-Huron water
due to the 25-foot, 1933, channel improvement project. The
Corps of Engineers (Great Lakes Basin Commission, 1972,

p. 106) calculated that the construction of the navigational

channels completed in 1933 and 1962 caused a total lowering

26

of .59 foot in the levels of Lake Michigan-Huron.

In two papers Brunk (1963, 1961), a member of the U.S.
weather Bureau, noted that peaks were discernible in 10-year
running averages of mean annual levels of the Great Lakes and
Lake St. Clair in the 1880's and around 1950. The 1950
maximum was more than 1.5 feet lower than that of the 1880's.
From studying Lake Erie water balance data he concluded that
this drop in water level was attributed to downcutting of the
channel connecting Lake Huron and Lake Erie. water supply
factors of precipitation, inflow, evaporation and diversion
were not pertinent in this drop of water level. Lawhead
(1961), of the Corps of Engineers, in a discussion of Brunk's
1961 article, claimed that the channel changes caused a
lowering of the Lake Michigan—Huron level of only .73 foot
between 1887 and 1955.

Combining the Corps of Engineers' estimate of the
effects of dredging the 25-foot and the 27-foot channel in
the 1930's and 1960's and the St. Lawrence waterway report
(Joint Board of Engineers, 1926) calculation prior to 1925,
it can be inferred that the Michigan-Huron water level has
been lowered approximately 1.19 feet from what it would be
under "natural" conditions. This figure agrees somewhat
better with Lawhead's figure than it does with Brunk's cal-
culation. Perhaps Brunk's outcome was somewhat biased by the
long-term above-average precipitation which played such a

key factor in the high levels of the 1880's.

Effects of Diversions, Inflows and Outflows

The Long Lake-0goki diversion into Lake Superior and
ultimately into Lake Michigan-Huron tends to cancel the out-
flow diversions at Chicago and through the welland Canal.
As of 196% the effect that regulating Lake Superior by way
of the St. Marys River system had on Lake Michigan-Huron
levels was negligible (Great Lakes Basin Commission, 1972,
p. 132). As previously mentioned, the St. Clair River
channel modifications lowered the Michigan-Huron level by
more than one foot, possibly one reason why the high levels
of the 1950's did not approach the high levels recorded in
the 1880's. One must remember, though, that the accumulated
effect of all those years prior to the 1880's, when annual
precipitation rates were always above average, may be a prime

factor involved.

Groupdwaterlnflow and Seepage

The contribution of ground water to the water budget of
Lake Michigan-Huron has not been determined but it is be-
lieved to be of insignificant proportion. Small amounts of
water flow directly into the lake by subterranean movement,
in addition to that which seeps into the stream channels and
is included in the runoff.

SURFACE PRECIPITATION, LAKE LEVELS
UPPER 32g FLOW
The primary factor affecting the Lake Michigan-Huron

water level is precipitation falling on its basin. During

28

periods extending several years when precipitation rates are
below normal the lake level will appreciably fall. The re-
verse will usually occur if the rates are above the mean
during the extended period. An explanation of these long-
term shifts in the precipitation pattern is complicated but
would contribute to an understanding of natural fluctuations
in lake levels. It is the purpose of this section to demon-
strate that shifts in mean planetary wave positions occur
between periods of contrasting precipitation rates; these
shifts are apparently related to long-term variations in the
level of Lake Michigan.
Mid-Latitude Upper Air_Flow and Associated Surface weather
Conditigng

The Great Lakes district lies in the westerly wind belt
of the mid-latitude Northern Hemisphere. Here air masses
migrating from contrasting thermal regions create a zone of
baroclinity. The earth's rotation and air movement res-
ponding to this baroclinity generate a middle atmospheric
west-to-east circumpolar wind flow. These upper level
westerlies are rarely zonal, however, but assume a wave-like
or sinuous plan (Figure %). Two factors are responsible for
this pattern (O'Connor, 1963; Harman, 1971). Vertical
obstructions, such as mountain ranges, and land-sea thermal
contrasts, particularly during the winter season, distort
zonal flow and "force" an initial wave. This wave produces
others (teleconnection) due to the tendency of the air to

conserve its absolute vorticity (or spin); that is, "the

 

29

 

      
 
 

 

 

 

  
  
 

25. g: lift/{€35 r
\ .... ._ “I q
qu'fk‘ “‘3‘"

 

    

M.) 4‘
.'::_._ '

 

 

Figure %: Mean 700 mb contours (tens of feet) for July,
1966, illustrating a typical mid-tropospheric
summertime flow pattern. Note the trough a-
long the U.S. west coast and a ridge over the
northern Atlantic Ocean (adapted from Harman,
J.R., 1971, p.21).

30

vorticity of the air stream, composed of the spin of the
earth plus the curvature and the shear of the air motion,
if any, tends to remain constant" (O'Connor, 1963, p. 1007).

These distortions in the upper level westerlies are
generally referred to as long waves or Rossby waves. Within
the air stream regions of anticyclonic curvature are re-
ferred to as ridges while those of cyclonic curvature are
called troughs (Figure 5). 0n the long-term, some waves
show geographic preference from month to month, but, at any
particular time, ranging from a day to a decade or more,
other positions may be observed.

Mid— and upper-tropospheric wind flow patterns are
linked with surface weather conditions in the middle latitudes.
”Although atmospheric interactions are very complex, circula-
tion at these levels is generally responsible for guiding
and generating the surface features of the daily weather map
that produce our day to day weather" (Harman, 1971, p. 1).

In fact, nearly every surface pressure center must start with
upper level support.

Surface regions east of a trough axis, the region of
positive vorticity advection, are characterized by surface
air convergence, vertical uplift, and upper level divergence
(Figure 5). The probability is high for precipitation
associated with this area of uplift, and southwesterly winds
tend to advect warmer air masses into the region. Clear
surface weather predominates under the negative vorticity

advection area to the rear of the trough axis. Here, upper

31

.sofipoe
capesgmoSpm Hecappe> cepmflOOmmm cam e>m3 capesamomonp w mo seawmflc OHpmSenom < um mndwflm

 

 

 

 

Bahama?“ :3 mmbmmmma mafia
8“,th Ease amamn 5.5a Bahama/Ho Sam... masoa
232% gnome a! I II moammbo 030%
Room oooamai Boom 89w:
moaambm m2 oom I II II Baambm m: com
moa,._,m.e..E.Hm amamfi mmaae. mozmemmEoo SE... mam:
.\&Sfl€

 

      

A10;\

I

 

HDGIH

 

 

32

level convergence, subsidence, and surface divergence occur,
thereby fostering a high pressure region (Figure 5). Cooler
air from higher latitudes tends to be adveeted into the area.
Cyclonic tracks (Figure 6) show close spatial proximity to
the former area while anticyclonic tracks display a positive
spatial agreement with the latter region (Harman, 1971, p.
17; Klein, 1958, p. 102).

O'Connor (1963, p. 1007) suggests that ”the abnormal
longitudinal positioning of planetary waves (long waves) is
one of the basic factors producing abnormal weather over
large areas and long periods of time." Namias (1966) found
that the 1962-1965 drought which plagued the Northeast
(while the Southwest and Northern Plains simultaneously had
a superabundance of precipitation) resulted from abnormal
hemispheric wind patterns in layers of the atmosphere ex-
tending from the surface to the mid-troposphere, and
probably much higher. Later, Namias (1969) concluded that
abnormally high sea-surface temperature in the north-central
Pacific Ocean between 1961 and 1967 was a primary factor in
explaining a shifting of the upper waves from their modal
(and mean) positions. This warm ocean temperature forced a
southward winter-time displacement of the Aleutian Low.

The developing and occluding cyclones associated with this
system transferred their vorticity alof , thus producing a
change in the mean mid-tropospheric wind pattern. By tele—
connection a strong mean ridge over the western United States

accordingly was forced, while a strong mean trough resulted

33

.Aoa.a .auma “.m.h .aeaaom aoau cop

ummomv mflxm ammonp map mo pmmo esp ou seam maneumeznudom exp Spas

mewfloqfloo moapmuflmfloeng hbmen esp wasp epoz .oOHnem mafia Hmeflpenp

nomha m quHSU mepepm cepflcb map ne>o 30H“ mo mecflaameapm agape

unmam map powgew mecHH UHHOm saga map mmeamgz mxowpp oflnoaoho umea

pcemeaaen meGHH cashew h>wo£ esp mamamxe mfisp SH .eowmhsm as com
esp pm 30am baa: news out coapmpflgfloeng :eeBpon QHQmGOHumHeH one no madmflm

 

 

 

 

 

 

3%

over the eastern United States. Colder than normal tempera—
tures occurred in the East and warmer than average
temperatures persisted in the west. In addition, departures

of precipitation from normal were widespread.

Investigative Hypothesig

Changes in the mid-latitude upper air flow should cause
changes in the surface precipitation regime over the Lake
Michigan-Huron basin. For example, the mean monthly trough
axis should be positioned more frequently to the west than
to the east of the Lake Michigan-Huron drainage area during
extended periods of abnormally high precipitation and lake
levels. Periods of low precipitatibn amounts and water
levels should be associated with a mean trough axis east of
the basin, in contrast. This general relationship between
mean trough position and lake level variations formed the

investigative hypothesis of this paper.

Day;

Mean monthly charts of average 500 mb heights,
temperatures and resultant winds were obtained from
Climatological Data - National Summary for the period 1950
to 1971. These 26% charts were reproduced and trough axes
were plotted on each according to the average height con-
tours. Only those troughs whose associated wind flow
crossed at least part of the Lake Michigan-Huron basin were
considered.

Many important synoptic episodes of relatively short

35

duration no doubt were undetected in this study because mean
monthly wave positions often mask important intra-monthly
variations. Nevertheless, mean monthly charts were used
because of their availability and because it was assumed that

important planetary wave patterns would still be evident.

Method

For each of the years, 1950 to 1971, each month was
placed into one of three categories depending on whether the
mean monthly trough was east (category "one"), west (category
"two"), or directly over the study area (category "three").
Mean troughs west of the region typically produce south-
westerly flow and abnormally frequent precipitation over the
study area; dry subsident conditions prevail under a mean
northwesterly flow when the trough axis is east of the region.
Of the 26% months, eight displayed split flow patterns when
part of the basin was under a mean southwesterly upper air
flow while a mean northwesterly flow overlay the other portion
of the drainage area. These eight months were placed in
category "two."

Next, each of the 22 years was grouped into one of three
time periods corresponding to episodes of above or below nor-
mal water level in Lake Michigan-Huron. High lake levels
occurred between 1951 and 1955 and from 1969 to the present.
Below normal water levels prevailed from 1956 until 1968
(Table 1). Due to the lag between basin precipitation and

lake level fluctuations, these three study periods were

36

offset one year in relation to the true water level periods.
The 22 years were thus grouped into time spans to include
1950 to 195%, 1955 to 1967 and 1968 to 1971 (Table %).

For 11 of the 22 years during the study period pre-
cipitation in the Lake Michigan-Huron drainage area was
above normal. For the total of these years the average annual
number of months in each of the circulation categories was
determined. The same procedure was performed on the remaining
11 years during which time the annual precipitation amounts

were below the mean (Table 5).

Discussion of the Findings

During the 11 years when basin precipitation totals
were above normal the mean 500 mb trough axis was positioned
west of the drainage area 1.55 months longer than during the
11 years when mean basin precipitation was below normal.
(Table 5). This finding implies that southwesterly flow and
conditions favorable to general precipitation were more fre-
quent than during the "dry" years and would seem to explain
the long-term changes in precipitation. Hence, a correlation
between long-wave positions and lake level fluctuations is
evident even when monthly data are used.

The average number of months per year when the 500 mb
trough axis overlay part of the basin was similar for both
categories of high precipitation years and low precipitation
Years (1.91 and 1.82 respectively).

The mean monthly trough axis was located east of the

Great Lakes basin an average of at least seven months each

37

 

 

 

 

 

m m 5 mm. mm.e- Fame
m s m m . se.e came
o m or m . em. mom,
m a 1m om. - Ima.m Imema
m m m on. u mowm meme
: a a mm.eu mo.m- meme
a a or mo.m- m.: meme
m m a mm.m- e.m- some
a 0 ea oo.m- aw.m- meme
0 0 ma mo.eu ma.:- meme
a m w am. - mm. - Fee?
a a oe om. - ea.e come
a m e ma.e- mm.m mmme
a a or aa.e- am.m- mmma
a m e mm. - mm. amme
m o or om. - ma.m- emme
a a or we. mm.m- mmee
m a a me. w:.m amme
: m e am.P om.a- mmee
m o or mw.a mm.e- mmma
m m m eo. eo.o amma
m e om. - aa.a omea
samcm camam one samem one
one go pace uo emu: cap uo em<m one Apcoa sac Anemone nae
maamm>o mama op was case on mmz mega Haaaoz Heaaoz
amnepa ecu Amneea may nwzoha ecu 509m ehdpnmmem Seam weapaemem anew
genz mnpcoz cmgz mnpcoz nesz mango: He>eq hope: coepmpfiqfloenm
mo Henasz mo penapz mo 909852

 

 

 

 

 

 

 

.Faea-omee .saaae aoaam
lemmHQOHz oxen esp ae>o mGOHpflmog mean nwsoep as com hanpcoa ewes map can
meadoSm coepmpfiaeoeaa use He>ma mean: Hedger exp mo Hmanoc Bonn meHSuawmeQ «: magma

38

 

 

 

 

Table 5: The mean monthly 500 mb trough axis positions
for the years when the annual precipitation
amounts were above and below normal over the
Lake Michigan—Huron basin, 1950—1971.

Average Annual Average Annual Average Annual
Number of Number of Number of
Months When Months When Menths When
the Trough the Trough the Trough
Axis Was to Axis Was to Axis OVERLAY
the EAST of the WEST of Part of the
the Basin the Basin Basin

gears When the‘

ake Michigan—

Huron Precipi-

tation Amounts 7'27 2‘82 1'91

were ABOVE the

ormal

ears When the

ke Michigan-

ur°n Freelpl‘ 8.91 1.27 1.82

ation Amounts
vere BELOW the
.ormal

 

 

 

 

Table 6: The mean monthly 500 mb trough axis positions
over the Lake Michigan-Huron basin for the ex-
tended periods of above normal water levels,
1950-195% and 1968-1971, and below normal water
levels, 1955-1967.

 

verage Annual Average Annual

Average Annual

 

 

 

 

 

 

umber of Number of Number of
Months When ionths When Months When
the Trough the Trough file Trougl
Axis was to is was to Axis OVERLAY
the EAST of the WEST of Part of the
the Basin the Basin Basin

1959wgt§99+ 7.20 1.80 3.00

If
19’Idéy3967 8.8% 1.77 1.38
1 68 - 1 '1
9 (wet 97 5-75 3.25 2.00

 

 

 

 

39

year for each of the two groups of years under consideration.
The locational pattern resulting in these high yearly
averages is typical of the mean North American flow, since a
ridge is frequently "forced" over the Rocky Mountains, re-
sulting in a mean trough over the eastern United States.

It should be expected that this relationship would hold
true between extended, continuous periods of high water levels
(and therefore high drainage area precipitation) and extended
periods of lower lake levels (and therefore low basin pre-
cipitation). During the period 1955 to 1967 when the Lake
Michigan-Huron level and the mean precipitation receipts
were both low, the mean monthly trough axis was positioned
east of the basin for over 1.5 months longer than during
either period, 1950 to 195% and 1968 to 1971, when water
levels and precipitation rates were high. In the latter case,
the mean annual difference was over two months (Table 6).

The trough axis was positioned to the west of the Lake
Michigan-Huron basin for a yearly average of 1.8 months during
the high water period 1950 to 195%. During the other extended
period of high water, 1968 to 1971, the average was higher,
3.25 months per annum. This apparent discrepancy can be
misleading. During the 1950-195% period the trough axis had
overlain the basin region for a month longer each year than
during the 1968-1971 period, but was positioned west of the
basin less frequently than during the latter. This fact
suggests that the mean flow pattern was oriented in a posi-
tion to increase the precipitation probability during both

periods.

%O

The mean 500 mb trough axis was positioned over the
drainage area for an average of only 1.38 months per year
during the low water period, 1955 to 1967. This mean was
.%% months less per year than that calculated for the total
years when the precipitation amounts were below normal
(Table 5). Nevertheless, during the 1955 to 1967 interval
the trough axis was found to be positioned to the west of the
basin for an average of .5 months longer per year than during
the 11 years when the precipitation amounts were below the
norm (Table 5). Therefore, it seems that any effect which
these extended position locations may have caused would have

been negated by each other on an annual basis.

Conclusion

From this cursory examination there appears to be a
definite contrast between upper wind flow patterns character-
istic of periods of low and high lake levels in the Lake
Michigan-Huron basin. During periods of high water levels
the Lake Michigan-Huron basin appears to be under the influ-
ence of a mean southwesterly planetary wind flow considerably
longer each year than during periods of low water levels.
This mean southwesterly flow suggests a higher daily fre-
quency of southwesterly flow patterns and attendant unstable
weather. During abnormally low Michigan-Huron water levels
northwesterly flow is more prevalent over the basin, this

implying more frequent stable, dry anticyclonic conditions.

LIST OF REFERENCES

LIST OF REFERENCES

Blust, F. and B.G. DeCooke. 1960. Comparison of precipitation
on islands of Lake Michigan with precipitation on the
erimeter of the lake. Journal of Geophysical Research,

p
65(5): 1565-1572.

Brater, E.F. 1950. Beach erosion in Michigan. Research Publi—
cation No. 2, Lake Hydraulic Laboratory, Engineering
Research Institute, University of Michigan. Ann Arbor,
Michigan, 39pp.

Bruce, J.P. and G.K. Rodgers. 1962. water balance of the
Great Lakes system, %1-70. In H.J. Pincus (editor),
Great Lakes Basin. Publication No. 71, American Associa-
tion for the Advancement of Science. washington, D.C.

Brunk, I.w. 1968. Evaluation of channel changes in St. Clair
and Detroit Rivers. water Resources Research, %(6):

1335-13%6.

. 1963. Additional evidence of lowering of Lake
Michigan-Huron levels. Proceedings, Sixth Conference
on Great Lakes Research, Great Lakes Research Division,
University of Michigan. Ann Arbor, Michigan, 191-203.

 

. 1962. Precipitation estimates in the Great Lakes
drainage basin. Monthly Weather Review, 90(3): 79-82.

 

. 1961. Changes in the levels of Lake Michigan and
Hurgn. Journal of Geophysical Research, 66(10): 3329-
333 .

 

. 1960. Precipitation and the levels of Lake Mich-
igan and Huron. Proceedings, Third Conference on Great
Lakes Research, Great Lakes Research Division, Univer-
sity of Michigan. Ann Arbor, Michigan, 1%5-150.

 

Changnon, S.A.,Jr. 1968. Precipitation climatology of Lake
Michigan. Bulletin No. 52, Illinois State water Survey.
Urbana, Illinois, 88pp.

. 1967. Average precipitation on Late Michigan.
Proceedings, Tenth Conference on Great Lakes Research,
Great Lakes Research Division, University of Michigan.
Ann Arbor, Michigan, 171-185.

 

1+1

%2

Corps of Engineers. 1971. Great Lakes region inventory re-
port, national shoreline study. North Central Division.
Chicago, Illinois, 221pp.

Day, P.C. 1926. Precipitation in the drainage area of the
Great Lakes, 1875—192%. Monthly Weather Review, 5%(3):
85-1060

DeCooke, B.G. 1968a. Great Lakes regulations. Proceedings,
Eleventh Conferencecn Great Lakes Research, International
Association for Great Lakes Research, 627-639.

. 1968b. The Great Lakes water levels problem.
Limnos, 1(1): 22-26.

 

Dorr, J.A. and D.F. Eschman. 1970. Geology of Michigan. Uni-
versity of Michigan Press. Ann Arbor, Michigan, %76pp.

Eichmeier, A.H. 196%. Precipitation and upper Great Lakes
levels. Proceedings, Great Lakes Levels Conference, The.
Kellogg Center for Continuing Education, Michigan State
University. East Lansing, Michigan, 16-22.

Ewing, M., F. Press and w.L. Donn. 195%. An explanation of
Egg égke Michigan wave of 26 June 195%. Science, 120:
" 60

Fox, W.T. and R.A. Davis, Jr. 1970. Profile of a storm-wind,
waves and erosion on the southeastern shore of Lake Mich-
igan. Proceedings, Thirteenth Conference on Great Lakes
Research, International Association for Great Lakes Re-
search, 233-2%1.

Freeman, J.R. 1926. Regulation of elevation and discharge of
the Great Lakes. Sanitary District of Chicago, 5%8pp.

Great Lakes Basin Commission. 1972 (Final Draft). Levels and
flows. Appendix No. 11, Great Lakes Basin Framework
Study. Ann Arbor, Michigan.

Harman, J.R. 1971. Tropospheric waves, jet streams and United
States weather patterns. Association of American Geo-
graphers, Commission on College Geography, Resource Paper
No. 11. Washington, D.C., 37pp.

Harris, D.L. 1957. The effect of a moving pressure disturbance
on the water level in a lake. Meteorological Monograph,

2(10): %6-57.

. 1953. Wind tide and seiches in the Great Lakes.
Proceedings, Fourth Conference on Coastal Engineering,

25-51.

 

43

Horton, R.E. and C.E. Grunsky. 1927. Hydrology of the Great
Lakes. Sanitary District of Chicago, %32 pp.

Hough, J.L. 1958. Geology of the Great lakes. University of
Illinois Press. Urbana, Illinois, 313 pp.

Hughes, L.A. 1965. The prediction of surges in the southern
basin of Lake Michigan, part III: the operational basis
for prediction. Monthly hbather Review, 93(5): 292-296.

Irish, S.M. 1965. The prediction of surges in the southern
basin of Lake Michigan, part II: a case study of the
sgrge of August 3, 1960. Monthly weather Review, 93(5):
2 2.291.

Joint Board of Engineers. 1926. St. Lawrence waterway report.
Appointed by file governments of the United States and
Canada, 13.

Kirshner, L.D. 1968. Effects of diversions on the Great Lakes.
Miscellaneous Paper 68-7, U.S. Lake Survey, Corps of
Engineers, Department of the Army. Detroit, Nachigan,
277-3100

Klein, W.H. 1958. The frequency of cyclones and anticyclones
in relation to the mean circulation. Journal of Meteoro-

Kohler, M.A. 1959. Discussion of evaporation of Lake Ontario
by I.A. Hunt. Journal of Hydraulics, A.S.C.E., 85, HY9:
109‘1120

Korkigian, I.M. 1963. Channel changes in the St. Clair River
since 1933. Proceedings of the American Society of
Civil Engineers, J0urnal of the Waterways and Harbor
Division, 89(WW2): 1-1%.

Laidly, W.T. 1962. Regimen of the Great lakes and fluctuations
of lake levels, 91-106. In H.J. Pincus (editor), Great
Lakes Basin. Publication No. 71, American Association
for the Advancement of Science. washington, D.C.

Lawhead, H.F. 1961. Discussion of paper by Ivan w. Brunk,
'changes in the levels of Lake Michigan and Huron.' Jour-
nal of Geophysical Research, 66(12): n32n-n329.

Mansfield, J.B. 1899. History of the Great Lakes, volume one.
J.H. Beers and Company. Chicago, Illinois, %2-%%.

Muller, F.B., J.C. Gervais and R.W. Shaw. 1965. The effect of
precipitation on the levels of Lake Michigan-Huron.
Canadian Meteorological Branch, Canadian Department of
Environment, CIR-%26%, TEC-576.

LIL.

Namias, J. 1969. Seasonal interactions between the north Pa-
cific Ocean and the atmosphere during the 1960's. Monthly
weather Review, 97: 173-192.

. 1966. Relation between fluctuations in United
States climatic patterns and 1962—65 drought. Journal
of American waterworks Association, 58: 1528-15%8.

 

O'Connor, J.F. 1963. Extended and long-range weather fore-
casting. Journal of American waterworks Association,
55: 1006-1018.

Pentland, R.L. 1968. Runoff characteristics in the Great
Lakes basin. Proceedings, Eleventh Conference on Great
Lakes Research, International Association for Great
Lakes Research, 326-359.

Platzman, G.w. 1965. The prediction of surges in the southern
basin of Lake Michigan, part I: the dynamic basis for
prediction. Monthly weather Review, 93(5):275-281.

Ramey, H.F. 1952a. Great Lakes levels and their changes, part
I. Midwest Engineer, %(6): 5—30.

. 1952b. Great Lakes levels and their changes, part
II. Midwest Engineer, %(7): 3-2%.

 

Richards, T.L. 1969. The Great Lakes and the weather. In D.V.
Anderson (editor), The Great Lakes As an Environment.
Great Lakes Institute, Report PR 39, University of Tor-
onto. Toronto, Ontario, 51-72.

Saville, T. Jr. 1953a. Hindcast wave statistics for the Great
Lakes. Proceedings, Fourth Conference on Coastal Engin-
eering, 1-12.

. 1953b. wave and lake level statistics for Lake
Michigan. U.S. Corps of Engineers, Beach Erosion Board
Technical Mimeograph No.36.

 

Secretary of War. 1900. Report of the board of engineers on
deep waterways. 56th Congress, 2nd Session, Document
No. 1%9, Part I, 37.

Strong, A.E. 1968. weather and waves. Limnos, 1(2): 22-26.

and F.R. Bellaire. 1965. The effect of air stabil-
ity on wind and waves. Proceedings, Eighth Conference on
Great Lakes Research, Great Lakes Research Division,
University of Michigan. Ann Arbor, Michigan, 283-289.

 

%5

United States Senate. 1957. Effects of an additional diver-
sion of water from Lake Michigan at Chicago. Senate
Document 28, 85th Congress, 1st Session. United States
Government Printing Office. washington, D.C., 7%pp.

APPENDIX

APPENDIX

NON—HYDROLOGICAL ELEMENTS
AFFECTING THE LAKE MICHIGAN-
HURON WATER LEVEL

In addition to long-term variations, seasonal and
hourly variations in lake level occur within the Michigan-
Huron basin. Seasonal and long-term variations result from
change in the volume of water in the lake whereas daily and
hourly variations arise from temporary shifts of the water
within the lake basin, and are therefore non-hydrological in
nature.

Daily and hourly changes in water level are local in
nature and differ from place to place around the lake (Great
Lakes Basin Commission, 1972, p. 99). Brief variations are
caused by winds, shifts in barometric pressure, ice-blocked
outlets and tides. hands are by far the most important
factor.

Crustal rebound, although long-term by its nature, is
also a contributing non-hydrological factor affecting the
Lake Michigan-Huron water level. It can hardly be thought
of as a present-day threat to lake level stability, but in

centuries to come, it could pose quite a problem.

%6

%7

Wind

Winds can produce temporary changes in lake levels.
Frictional drag by the wind on the water surface can pile up
water on the downwind side of the lake basin. This behavior
of the water due to the wind is of utmost importance when a
lake with a large surface area also possesses a shoreline
zone which consists of unconsolidated and easily erodible
material such as exists along the Lake Michigan basin
boundary.

wave heights are directly related to velocity, duration
of the wind and the fetch (Brater, 1950, p. 13). The fetch
refers to that open water distance over which the wind flows.
For a given wind velocity the wave height tends to increase
in the downwind direction until a maximum is reached. Due
to its vast surface area, Lake Michigan provides a well-
defined fetch for the westerly winds to travel over. Increase
in wind velocity increases the potential wave height.

”The energy of a wave varies as the square of its
height. Consequently, in a general way, it may be said that
the power of a wave becomes four times as great when the wave
height is doubled. It may be seen, therefore, that a know-
ledge of the frequency of occurence at various wave heights
is essential in studying the erosion problem at any location"
(Brater, 1950, pp. 12-13). Saville (1953 a) has utilized
synoptic weather charts to predict past wave characteristics
off of Milwaukee, Wisconsin, for a three year period. He

also authored a Corps of Engineers Beach Erosion Board

%8

Technical Mimeograph (1953 b) portraying wave characteristics
along most portions of Lake Michigan.

Greater wave generation generally occurs on Lake Michigan
during the fall and early winter. At this time the water
temperature is characteristically warmer than the overlaying
air masses. This situation creates an unstable air mass con—
dition and tends to transfer momentum from the upper level {8 ”MT
gradient winds to the water surface (Strong and Bellaire, I
1965; Strong, 1968). Spring conditions generally generate

stable air conditions because the air masses are usually

 

warmer than the water body. b-~—--—---»"’
Ramey (1952, p. 2%) cites several examples. In October,

1929, a continuous strong north wind raised the Lake Michigan

level 1.95 feet at Chicago in 37 hours. In November, 1951,

an east wind, shifting to a strong %2 mph northerly wind

raised the water level at Chicago by 1.70 feet. As this re.

wind subsided over the next 2% hours to nine mph the level

dropped 1.10 feet. As it shifted to a southerly wind and

increased its velocity to 23 mph, the Lake Michigan level

dropped an additional .8 foot.

Barometric Pressure Changes

Harris (1953, p. 26) noted that short period disturbances
in the Great Lakes tended "to occur in zones of disturbed
weather, that is, near fronts, squall—lines, thunderstorms,
etc . . ." In June, 195%, an abrupt 10 foot surge in the

Lake Michigan water level inundated the Chicago shoreline

49

shortly after a storm system had passed. In an article in
Science Ewing, Press and Donn (195M) noted that a few hours
earlier a severe squall line with winds up to 60 mph had
arrived from the northwest and had crossed southern Lake
Michigan. Associated with this squall was a very rapid and
strong pressure jump. They suggested that this disturbance
could be explained "on the basis on resonant transfer of
energy from the traveling pressure-jump and its associated
high winds in the air to a gravity wave traveling with equal
velocity in the lake. Only for equal velocities can a large
wave be generated" (p. 685).

Harris (1957) expounded on this theory in relation to
the 1954 Chicago surge and arrived at the conclusion that
"in determining which pressure jump will be accompanied by
important water level disturbances in Chicago, it is likely
that the orientation of the pressure jump will be equally or
more important than the speed of the disturbance. This is
because Shoaling, reflection, and convergence due to the con—
tours of the shore must all be considered to recount for a
disturbance of the observed magnitude" (p. 56).

Minor changes in water levels are probably commoa to
most all migrating pressure cells but significant surges are
relatively uncommon. When they occur their consequences can
be dramatic. The Menthly weather Review, in a three part
study (Platzman, 1965; Irish, 1965; Hughes, 1965), presented
papers concerned with the type of atmosphere-lake interaction

which occurred at Chicago. Hughes (1965) describes other

50

water level surges which have taken place on Lake Michigan.

Differences in atmospheric pressure at opposite ends of
a lake can have much the same effect on water levels as that
of a steady wind blowing from one end to the other. If a
high pressure cell overlies one end of the lake for a period
of time, while a low pressure system overlies the other end,
the high pressure cell has a depressing effect on the water.
This must be compensated for by an increase in water level
elsewhere where atmospheric pressure is less. Consequently,
discrepancies between levels at one end of the lake and the
other may be several feet.

The combined effect of wind and barometric pressure
changes which occur in certain storm systems has a magnifying
quality. Incoming high pressure cells tend to depress the
water level at one end of the lake while downwind the water
level increases beneath the low pressure cell. This increase
in water level is amplified by the fact that the wind is also
increasing the level in the downwind direction. In addition
to this, the large waves generated by the wind during the
storm have the capability of surging still further up the
beach zone. It can be seen, then, that the elevation of a
steady-state water level can be temporarily increased to a
great extent by combined changes in barometric pressure and
wind flow. If this condition occurs during periods of above
average water levels severe consequences may result along the
erodible shoreline. In one study (Fox and Davis, 1970),

analysis of 17 environmental parameters recorded during a

51

storm revealed positive relationships between shoreline
erosion and barometric pressure, wind, breaker height,
breaker angle and longshore current along the southeastern

shoreline of Lake Michigan.

Seiches

Short-time standing wave oscillations in the surface of
the lake may result from changes in atmospheric pressure and
wind direction. Such oscillatory waves are called seiches
and on Lake Michigan are reported to have caused local
changes in water levels as great as five feet (Dorr and
Eschman, 1970, p. 22H). "The initial impulse which starts a
seiche in the Great Lakes is probably muCh more frequently due
to the wind than to barometric gradient" (Hough, 1958, pp.
44-45).

It is interesting to note that seiches were recognized
to have occurred "due to large and rapid atmospheric changes
noticeable upon the barograph preceding or during thunder-
storms" as early as 1899 (Mansfield, 1899, p. #2). Mansfield
describes a particularly violent one on Lake Michigan which
occurred in 1893 at Chicago. water suddenly had risen four

to six feet in a series of heavy waves.

Ice Retardation

Ice jamming of the St. Clair River during the winter
reduces outflow from the Lake Michigan-Huron basin. The re-
sult is a lowering of the Lake St. Clair and Lake Erie water

levels while the level of Lake Michigan-Huron is increased

52

because of the restricted flow. In some winters the dis-
charge of the St. Clair and Detroit Rivers has been reduced
by as much as one half for a monthly period of ice obstruc-
tion; in other winters there has been little retention of
outflow by ice (Freeman, 1926, p. 207). The Great Lakes
Basin Commission's draft report (1972, p. 8%) cites a 1965
Corps of Engineers study which estimated that the level of
Lake Michigan-Huron is about .H foot higher than it would be

without ice retardation.

Elias;

Both lunar and solar tides exist on Lake Michigan but
their effect is so small as to be inconsequential. The
National Ocean Survey, Department of Commerce (formerly the
U. 3. Coast and Geodetic Survey) has determined that the
spring tide (combined lunar and solar tides) is less than
two inches (Ramey, 1952, p. 2%). This is masked by the
greater fluctuation in levels produced by wind and barometric

conditions.

Crustal Rebound

In the 1890's G. K. Gilbert astutely perceived that
late- and post-glacial isostatic adjustment of the earth's
crust following glacial unloading had occurred and was
presently taking place in the Great Lakes region. The earth's
surface, previously depressed under the ice of continental
glaciation, had begun to warp slowly upward once the ice

thinned and retreated. This crustal uplift is not uniform

53

in all areas. In the Lake Michigan district a hinge line ex-
tends in a northwest-southeast direction between the south
tip of Green Bay to approximately the latitude of Manistee,
Michigan. North and east of this line the crust is still
rebounding.

At one time the upper Great Lakes drained northward
through northern Ontario, but, due to the isostatic adjustment
of the earth's crust, the drainage is now south and east
through the St. Lawrence system.

time passes the differential crustal uplift of Lake
Michigan's northeastern rim may tilt the water enough rela-
tive to its southwestern rim so that the lake may naturally
drain across the divide at Chicago into the Mississippi
basin. Freeman (1926) suggested that this may become a
serious problem in a thousand years.

As should be evident, then, water levels along the shore
at localities north and east of the hinge line are receding
with respect to the water level at the Lake HUron outlet and

at Chicago.

   

"IT'ElllllllfillllilITS