(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Studies of local winds and alongshore currents : December, 1967"

, BENTON HARBOR POWER PLANT LIMNOLOGICAL STUDIES. 
PART II. STUDIES OF LOCAL WINDS AND ALONGSHORE CURRENTS, 

December, 1967 

By: John C. Ayers 
Alan E. Strong 
Charles F. Powers 
Ronald Rossmann 



,Under Contract With: 

American Electric Power Service Corporation 



INTRODUCTION 

During late spring, summer, and early fall of 1967 the 
directions and velocities of near-water winds and of the along- 
shore water currents were recorded simultaneously in front of 
plant site #5. 

It was necessary to study the alongshore water currents because 
they will entrain and move the plume of warmed effluent water from 
the plant. The very local near-water wind conditions had to be 
investigated at the same time as the water currents because the 
area of interest is close inshore (1000 feet and less from the 
beach) , and because the very local winds which might be expected 
to at least modify the alongshore currents are subject to severe 
topographic control by the bluff faces of the sand dunes along 
the shore. Topographic control of the local winds was adjudged 
to be so severe that wind records from the meteorological tower 
atop the sand dunes would probably not be applicable in the 
question of the control of the local alongshore water currents. 

At the plant site the trend of the shoreline is north- 
northeast to south-southwest. About 200 feet from water line the 
bluff of sand dunes arises from a grassy beach berm about three 
feet high; the beach rises about a foot from water line to the 
base of the berm. The dunes protect the local inshore waters from 
winds from north-northeast, around through east and south, to 
south-southwest. The NNE-SSW orientation of the bluff of dune 
faces led us to suspect that there would be enforced channelling 
of winds from northerly and southerly directions into the NNE-SSW 



direction. If this were true there probably would be a pre- 
dominance of currents parallel to shore but under the influence 
of local surface- winds different from those recorded by the 
meteorological tower on top of the dunes. 

For these reasons, and also because of the spring and early 
summer blanketing effect of the stable spring air column (warm 
air over cold water) in keeping wind away from the water surface, 
we felt obliged to record the very local winds near to the water 
and separately from the meteorological tower. 

METHODS 

Alongshore Current Data ; The regimen of alongshore currents 
in front of the plant site was studied by means of a tripod- 
supported pendulum current meter. Because the area of interest 
was the region between the two sand bars near shore and was subject 
to wave action, the entire meter installation had to be heavily 
built, and some sacrifice of delicacy of measurement had to be 
accepted, to obtain ruggedness sufficient to withstand the environ- 
ment. 

Built of heavy-duty 2-inch pipe braced with angle iron, the 
tripod-supported pendulum meter stood five feet high above an iron 
foot-plate under each foot. From the top of each leg a pipe 
member four feet long reached horizontally inward to a center iron 
plate beneath which a sealed iron-plate pendulum was suspended by 
gimbals- The pendulum contained a multiple series of modified 
mercury switches capable of indicating eight current directions 



and four rough current speeds. The meter was secured by three 
earth-auger anchors screwed into the bottom, one just inside each 
leg. From the anchors vertical hold-down wires ran. up to the 
outer end of the horizontal pipe members. 

The meter was installed in 15 feet of water between the 
inner and outer sand bars in front of the cottage at the north 
side of the plant site. The output of the meter was transmitted 
by multiconductor cable to the cottage and recorded on an Esterline- 
Angus Events Recorder. The free-swinging pendulum of the meter 
could sense both wave action and current, and current direction 
was determined from the record by the dominance of numbers of 
north and west swings over south and east swings (or vice versa) . 

The pendulum meter was calibrated in the University of 
Michigan Department of Naval Architecture towing tank. Deflection 
of 5° (0.98 foot per second) was necessary to activate the direction- 
switches. Further current velocities needed to activate the sub- 
sequent velocity switches were: 2.62 ft/sec; 3.44 ft/sec; and 
4.92 ft/sec. It was verified in the field that currents less than 
0.98 ft/sec were all northward during the period of investigation 
(discussed later) . 

The meter withstood the wave action of winds over 30 mph, but 
on 11 October it was hit by a two-bushel mass of nylon fish net 
with incorporated driftwood moving under 35 mph winds and recorded 
water current of 4.92 feet per second. The impact broke a poorly 
welded joint where one of the -legs joined its horizontal member 
and that side of the tripod collapsed, laying the pendulum on 



bottom and stopping the recording of current. The meter was clean 
and still capable of operating when it was recovered on 10 November. 

Wind Data : Inshore over-water winds that might be expected 
to be the driving or modifying force for the alongshore currents 
between the sand bars (at about 500 and 1000 feet from shore re- 
spectively) were recorded at 15 feet above the water by a Weather 
Bureau Standard Cup Anemometer and Wind Vane (sold by Science 
Associates, No. 440). This installation was about 50 feet out from 
the face of the bluff and on the grass-covered beach berm at the 
foot of the bluff. On a heavy wooden base and stoutly guyed, this 
installation performed well throughout the period of record. 

Outputs from the anemometer and wind vane were led by multi- 
conductor cable to the Esterline-Angus recorder in the cottage, 
and provided a simultaneous wind record to go with the current 
recordings . 

ACKNOWLEDGEMENTS 

We are indebted to several members of the Indiana and Michigan 
Electric Company for help and advice at various times. Particularly 
we are grateful to Mr. John Banyon for unfailing help at all times. 
Especial thanks go to Mr. Roy Whitehead who was in residence at 
the cottage during part of the recording period and who tended the 
recorder. After he moved from the cottage his tending of the 
recorder involved special travel and inconvenience. Not knowing 
who was responsible, we extend thanks to Mr. Banyon for Indiana 
and Michigan's installation of rows of closely spaced light-pole 



stubs at both edges of the plant site property. They materially 
decreased the chances of vandalism to our gear by denying the 
beach to local vehicles. 

RESULTS 

The current meter and wind instruments were installed on 11 
May 1967. Recordings began at 12 noon EST on that day. Current 
recording continued until the current meter was disabled at 6:30 
PM EST on 11 October 1967. The recordings of wind continued 
until 1:00 PM on 14 November. 

The period of recordings covers part of the last month of 
spring (May) , all the summer (June, July, and August) , and most 
of the fall (September, October, and half of November) , extending 
from cold-water conditions in late spring to cold-water conditions 
in late fall. It covers the summer period when conditions of warm 
water and the presence of summer residents are important. 

The results consist of simultaneous records of wind direction 
and speed and current direction and speed. 

Table 1 gives the relations of wind directions to the trend 
of the shoreline; it is these relationships that probably make 
the very local over-lake winds different from those recorded on 
the meteorological tower. This table deals only with the thirteen 
directions from which our wind instruments were not sheltered by 
the bluff of sand dunes; the seven-month mean percentages of winds 
from these directions are also given, as a means of indicating the 
effects of the shore. Calms and variable winds are not given. 



Table 1, 



Relation of thirteen directions of wind to the plant site shoreline. 

Wind Seven-month mean % 

from Orientation of wind to plant site shoreline frequency of wind 

NE from the land . 04 

NNE.... .parallel to the shore 14.2 

N. obliquely onto shore at about 22° .- . 4.5 

Nlsrw\ .. .obliquely onto shore at about 45° 10.2 

NW obliquely onto shore at about 67° 1.4 

WNW....onto shore at about 90° 2.6 

W obliquely onto shore at about 67° 1.0 

WSW. .. .obliquely onto shore at about 45° 3.5 

SW obliquely onto shore at about 22° 1.0 

SSW. .. .parallel to the shore 22.6 

S ..... .. from the land 2.5 

SSE .... from the land 2.4 

SE.....from the land 



The peak mean wind frequencies at NNE and SSW are considered 
to be shore-caused abnormalities. This suspicion is supported by 
the low percentages of the winds on either side of the peak fre- 
quencies. We regard the drop from 14.2% to 4.5% between NNE and 
N and that from 22.6% to 1.0% between SSW and SW as indicating sub- 
stantial redirection of north and southwest winds as they come into 
contact with the dune faces. The percentages of N and SW winds 
also appear to be suspiciously low and indicative of wind channelling 
when compared to the percentages of NNW and WSW winds. Independent 
support for these beliefs is afforded by some comparisons made 
between our wind records and those from the 200-foot level of the 
meteorology tower. 

During the first two weeks of exposure of our wind instruments 
they were compared under different wind directions and velocities 
to the 200-foot-level wind instruments on the meteorological tower. 



The comparisoi was made under the assumption that the 200-foot 
level on the meteorological tower atop the dunes was relatively 
free of shore effects and could provide a measure of the shore 
effect on the surface winds driving the very local currents. 
These comparisons are presented in Table 2. 

Table 2 . 



Comparisons of winds at 16 feet on the beach to those at 200 feet 
on the meteorological tower. 

Date Time Beach wind at 16 feet Tower wind at 200 feet 



18 May 0600 from SSW (204^) 10.5 mph from 204° 22 mph 

1200 from SSW (204°) 15.5 mph from 210° 30 mph 

1800 from SSW (204°) 17.5 mph from 200° 30 mph 

2400 from SSW (204°) 15.5 mph from 235° 32 mph 

19 May 600 from WNW (294°) 6.2 mph from 290° 30 mph 

20 May 0000 from NNE ( 23°) 13.0 mph from 0° 16 mph 

Among these comparisons the diversion of an unobstructed 
upper level wind from 210° to a beach wind from 204° (1200 on 18 
May) , the diversion of an unobstructed upper level wind from 2 35° 
to a beach wind from 204° (2400 on 18 May) , and the diversion of 
an unobstructed upper level wind from 0° to a beach wind from 23° 
(0000 on 20 May) are evidently redirections of lower-level wind 
due to its contact with the bluff of dunes. In a different sense, 
the redirection of a wind from 200° at upper level to a beach wind 
from 204° (1800 on 18 May) might be due to the wind dropping into 
the lee of the dune bluffs and turning to run up the beach in the 
lee of the dunes . 

The simultaneous wind velocities in Table 2 further indicate 
that the very local currents are influenced by local surface winds 



8 



of considerably less speed than those at the 200-foot level on 
the tower. 

As a check on the validity of our results we. have computed 
the monthly ratios of mean wind velocity /mean current velocity. 
The norm to which these ratios are compared is Olsen's finding in 
Lake Erie that the surface current is "about 2%"* of the surface 
wind velocity. In the 15 feet of water where the current meter 
was installed the surface current should reach to bottom. This 
computation of ratios is shown in Table 3. 

Table 3. 



Ratios of monthly mean surface wind velocities to monthly mean 
surface current velocities. 

Month Wind, mph Current, mph Ratio: Current/Wind, in percent 



May 


8.0 


0.13 


June 


5.6 


0.07 


July 


6.9 


0.10 


August 


7.1 


0.13 


September 


6.4 


0.12 


October 


9.3 


0.15 



1, 


.6% 


1, 


.3% 


1 


.4% 


1 


.8% 


1 


.9% 


1 


.6% 



Grand Mean 1 . 5% 

Within the limits of the norm, our results appear to be valid. 
The fact that none of the ratios attains to 2% is undoubtedly an 
effect of bottom friction, for the pendulum of the current meter 
cleared the bottom by only about eight inches. 



*See: Hutchinson, G. E. A Treatise on Limnology . Volume I, 
John Wiley & Sons, New York, 1957, page 291. 



The Local Surface Winds : From our beach installation the 
distributions of percent of hours that the wind blew from the 13 
usable -directions (also calms and variable winds) was as shown 
in Table 4. 

Calms and High Winds : The distribution of mph wind speeds 
in percent of the hours, and monthly weighted-mean wind speeds are 
given in Table 5. 

Table 5. 

Percent distribution of wind speeds, mph, and monthly mean speeds. 
Month Calm (0-3mph) 4-12mph 13-24mph 2 5-38mph >38mph Monthly Mean 



May 


27.0% 


55.9% 


17.1% 


0% 


0% 


8 . mph 


June 


41.5 


55.6 


2.9 








5.6 


July 


33.4 


56.1 


10.5 








6.9 


August 


37.9 


48.0 


13.6 


0.5 





7.1 


September 


47.2 


39.4 


13.2 


0.2 





6.4 


October 


24.2 


50.3 


23.8 


1.7 





9.3 


November 


23.2 


45.0 


29.7 


2.1 





10.1 



This tabulation follows Resolution 9 of the International 
Meteorological Committee, Paris, 1946. The allocation of "Calm" 
to all winds of 3 mph or less is also a recognition that our 
anemometer and wind vane required about 4 mph of wind to operate 
accurately . 

Table 5 shows that the highest percents of calm hours occurred 
during June and September. Calm conditions may be expected to 
provide minimal dispersion of the heated plant effluent; wave 
action would be absent, and the effluent would drift with the 
residual current remaining from the previous wind. While the 



10 



•H 
U 

> 






O 



O 



in 
o 



CN 

o 



o 



rd 
U 



CN 



iH 






cr> 



ro 






CN 
CN 



CN 



ro 

CN 



LO 



ro 
ro 



O 
U 
M-l 






H 
CO 

cn 



w 



CO 
CO 



r^ 


KD 


(T\ 


vO 


vX) 


o 


ro 


^ 


CN 


ro 


CNJ 


ro 


iH 


CN 


O 


CN 


KD 


^ 


LO 


CO 


LO 


^ 


CN 


LO 

• 


iH 


CN 


CN 


H 


O 


ro 


LO 


<N 


00 


r^ 


o 

• 


• 


00 

• 


• 


CN 

« 


• 



CN 



O^ ro [^ ^ CT> O cvj 

CN (N nH pH CN ro CN 



Q) 

rd 

EH 



fd 

o 
o 

CD 
tjn 
rd 
-P 

cu 
u 
U 

CD 



CO 



CO 

15 



CM O 

ro rH 



CN CN 
00 



00 



LO 


^ 


ro 


rH 


00 


ro 


00 


O 


o 


o 


nH 


o 


o 


iH 


CM 


rH 


• 


^ 


• 


r- 


00 

• 


• 


(Ts 


LO 



CTi LO 

^ 00 
ro 



^ V^ 00 CJ^ ^ «H '^ 

rH rH rH 



ro 
O 

rH 



rH 


.H 


LO 


lO 


CM 


cr> 


00 


• 


CN 


O 


CN 


.H 


CN 


^ 


^ 


CSJ 


ro 


^ 


O 


ro 


LO 


VD 


^ 


^ 


O 


O 


ro 


O 


O 


CN 


CN 


.H 


CN 


rH 


ro 

• 


0^ 




• 


CN 

• 


CN 



O 


CN 


ro 


ro 


r-\ 


00 


O 


LO 


^ 


LO 




CN 


LO 


CN 


r-\ 


^ 





r- 


o 


ro 


LO 


00 


00 


CN 


CM 


!?1 


• 


(Ts 


00 


rH 


r^ 


kD 


kD 


^ 


53 


CN 

ro 






rvj 


i-H 


' 




iH 
O 


w 


• 


O 


o 


o 


o 


o 


O 


• 


^ 


o 














O 


5 




CD 


>1 




-P 


. 


^ 




r; 


>i 


C 


nH 


tP 


Ph 


-p 


> 


fd 


o 


fd 


:3 


Hi 


::5 


CD 


o 





CD 


S 


S 


^D 


l-D 


< 


CO 


o 


:^ 


^ 



11 



percents of calm hours are high in June and September, their mere 
numbers do not indicate the numbers of consecutive calm hours 
during which residual currents might die away. This condition is 
examined in Table 6. 

Table 6. 



Daily maximum consecutive hours of calm. Night calms extending 
past midnight are added to the following day. 



Date 

1 

2 

3 

4 

5 

6 

7 

8 

9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
29 
30 
31 



Ma^^ 






12 

14 
6 

12 
4 




10 






5 



12 

1 



June 

2 

5 


24 
18 
11 

8 

1 



1 

5 

2 

4 

2 

1 

2 

2 

3 
14 

4 
21 

9 
11 

5 

4 

3 
14 

6 

2 





July Aug. 



2 
1 
3 
2 
7 

13 
5 
9 
1 

24 
6 
3 



12 

13 
2 
2 
3 




1 
5 
5 
8 
6 
4 
9 

11 



15 

9 

13 



16 

14 



16 





6 

16 

16 

9 

8 

9 



2 

4 

4 



1 

16 

15 

14 

12 



7 

4 



3 



Sept, 

14 
16 
17 
16 
15 
15 
19 
22 

2 
16 
17 
14 





19 

17 

11 

10 



5 

10 

2 

8 





1 

9 





Oct, 

20 



3 
2 

13 



2 
7 



15 

8 


7 


15 
2 



7 


27 
8 

11 



Nov. 

18 
1 
3 




14 

17 


20 







It is evident from the table that the great majority of calms 
during the summer months are of a few hours duration during which 



12 



it would be unlikely that the alongshore current would die away. 
During the first three weeks of September long periods of calm 
were dominant; under these conditions removal of warmed effluent 
by currents would be poor . 

Highest velocities of surface wind occurred sporadically in 
August (Table 5) and, to judge from the reduced hours of calm 
beginning 21 September (Table 6) , entered the picture again in 
the last week of September. High winds increased in frequency 
during October and November. The greater wave action and current 
movement accompanying such winds would provide improved dispersal 
of the effluent. 

The Very-Local Currents : The percentage distribution of 
alongshore water current directions and of currents too slow to 
activate the meter's direction-switches are given in Table 7. 

Table 7 . 

Percent of hours the current flowed toward: 



Month 


NE 
0% 


N 


NW 
62 . 3% 


W 
3.0% 


SW 
0.3% 


S 

0% 


SE 
0% 


E 
0% 


T< 
reco; 

«o 


DO low to 

rd direction 
.98 ft/sec) 


May- 


9.1% 


2 5.3% 


June 





2.0 


29.9 


4.8 
















63.3 


July 





2.9 


51.3 


2.9 
















42.9 


Aug. 





4.1 


50.2 


1.8 


2.3 


1.8 










39.8 


Sept. 


2.2 


6.5 


29.2 


2.2 


2.2 


3.0 


0.8 







53.9 


Oct. 


0.3 


12.0 


45.1 


4.4 


3.6 


5.7 










28.9 



The two outstanding features of this table are the predominance 
of recorded currents toward the northwest and of currents too low 
to activate the direction-switches of the meter. 



13 



Except for one hour in May, the current between the sand 
bars flowed toward north, northwest or west throughout May, June, 
and July. During these months the current flowed to these 
directions regardless of wind direction or wind velocity. 

On August 9th the first southward current since the one hour 
in May was recorded. After this, current to southward directions 
became progressively more frequent. 

Independent determinations of current direction at the meter 
site were made by float runs at several times during the summer. 
On all. these occasions the current was moving northward; on some 
of these occasions the meter was recording northward current, on 
others the current was too low to record direction. On 7 August 
our divers inspected the meter; it was clean and operating 
properly; the meter was recording northward current and the 
divers measured northward current beside it. 

The August southward currents were recorded after brisk winds 
from northerly directions had blown for some hours. As the fall 
progressed, less velocity and duration of these winds were needed 
to produce southward current. In a few cases the recorder was 
not operating when southward current began, but in every case 
where the beginning of southward current was recorded it began 
abruptly after current to the northward had been being recorded. 

The evidence from our float tests and other field measurements 
and from the recorded beginnings of south current is that the 
current during the too-low- to-record condition was always to the 
northward directions . 



14 



This being the case, the too-low- to-record-direction column 
of Table 7 should be added to the northward (NE, N, and NW) 
currents. If this is done, and SW, S, and SE currents are combined, 
Table 7 recombines as in Table 8. 

Table 8. 

Percent of hours current flowed northward, westward, and southward. 
Month Current Northward Current Westward Current Southward 



May 


96.7% 


3 . 0% 


0.3% 


June . 


95.2 


4.8 





July- 


97.1 


2.9 





August 


94.1 


1.8 


4.1 


September 


91.8 


2.2 


6.0 


October 


86.3 


4.4 


9.3 



From this we conclude that, if 1967 was reasonably normal, 
and if the warmed effluent from the plant is discharged between 
the sand bars, the effluent will in the vast majority of the cases 
during the critical warm months travel northward or westward away 
from the near-by water intakes of Bridgman and Orchard Beach. Our 
evidence still is that the plume of warmed water would not reach 
the St. Joseph water intake nine miles to the north. 

The distribution of weighted-mean water current velocities 
during the months of recording are given in Table 9. 

This table shows maximum frequencies of lowest current speeds 
in June and in September, lowest monthly mean current velocities 
in June and July with third-lowest in September, and a steady 
increase in current velocities in the 3.44 to 4.92 ft/sec range 



15 



from June through October. In most respects it is very similar 
to the behavior of the causative winds as shown in Table 5. 
Perhaps' its most important aspect is that the percentage of 
very low currents in September is not so low (by nearly 10%) as 
in June, implying that the conclusions drawn from Table 6 may not 
be so unfavorable as percentages of calm winds alone would indicate, 

Table 9. 

Percent distribution of water current speed in ft/sec., by hours. 
Month ' 0-0.98 0.98-2.62 2.62-3.44 3.44-4.92 >4.92 Monthly Mean 



May- 




2 5.3% 


47.0% 


15.6% 


12.1% 


0% 


, ly9 5 fps 


June 




63.3 


32.7 


3.0 


1.0 





. 1X04 


July 




42.9 


38.3 


11.6 


7.2 





,ly55 


Aug. 




39.8 


25.0 


17.7 


17.5 





.1»92 


Sept. 




53.9 


12.6 


5.6 


27.4 


0.5 


. 1X84 


Oct. 




28.9 


27.9 


14.6 


28.6 





'2/29 




Thi 


e Clima 


itological 


Representativ 


eness of 


1967: As 


a means 



of climatological comparison, the winds and air temperatures at 
Muskegon during the months of May through October were compared to 
the 30-year averages for these months. The departures of these 
parameters from the 1931-1960 averages are summarized in Table 10. 

Table 10. 

1967 departures from the 30-year normals. 

Month Air Temperature Wind Velocity 

May 

June 

July 

August 

September 

October 



-3.3°F 


+0 . 5 mph 


+2.7 


-0.1 


-1.5 


+0.6 


-3.8 


+0.1 


-1.9 


-1.5 


- -1.6 


+0.5 



16 



Except for the month of June, the summer was one of the 
coolest on record in the Midwest. Windwise, the months of June, 
.arid August were about normal; May, .July and October had somewhat, 
more wind than normal; and the winds of September were substantially 
below normal . 

The cooler air of May established over the lake a weaker than 
usual thermal inversion (warm air over cold water) through which 
the above-average May winds could more easily break to impart 
momentum to the water. Therefore, the observed current velocities 
of May must be considered higher than normal. 

The above-normal winds of July and October, to which consider- 
ations of stability do not apply, also would produce greater cur- 
rent speeds than normal . 

The decrease in mean current speed observed in September re- 
sulted from subnormal winds and an anomalous frequency of calms. 
September currents must be regarded as slower than usual. 

In the typical year, wind velocities become progressively 
greater from June through November and should be reflected in the 
mean current speeds . 

Considering all the data, it appears that in the normal year 
there should be no material danger that the alongshore current 
would die out for more than very short periods of time. 

DISCUSSION 

The current-direction results presented above are in accord 
with our best recent information about the underlying causes of 
alongshore currents. They are also compatible with our older 



17 



ideas about basic circulation patterns in the lake. 

Spring warming of the lake begins in the shallow water along 
shore/ while the main body of the lake is still at temperature of 
maximum density (4°C) or lov;er (continued wind-mixing in winter 
drives down and mixes in surface water cooled below 4°C by contact 
with colder winter air) . In spring the warmed shallow water along 
shore is expanded and less dense; it creates, between the cold 
main body of the lake and the warm shore, an alongshore sloped 
water surface that tilts upward toward the land. On this sloped 
water surface there will be, as a resultant of the earth's rotation 
(Coriolis force) and gravity, a tendency for establishment of a 
geostrophic circulation in which water current runs in such direc- 
tion that the elevated part of the water surface is on the current's 
right (observer's right when he looks downcurrent) . At the plant 
site the warmed inshore water makes a tilted water surface on 
which a current runs toward the north. In all seasons the pre- 
vailing westerly wind pushes surface water of the main body of the 
lake toward the east shore; in spring this push (wind set-up) 
narrows the alongshore belt of tilted water surface, increases the 
tilt of the tilted water surface, and strengthens the tendency 
toward northward- flowing current. 

By late spring or early summer the inshore warming has extended 
to greater depths and to a greater distance from shore, reducing 
the area of very cold surface water in the main body of the lake. 
In the alongshore areas where the warmed water has reached com- 
pletely to bottom, the basic local summer pattern of current eddies 



18 



begins to establish itself. Thus by June the elongate narrow in- 
shore counterclockwise eddy lying between Michigan City and Benton 
Harbor (identified by "x" in Figure 4 of Part I of this report) 
could have been established, though a reduced area of surface water 
of 4'^C or less was still present in the main body of the lake. The 
surface slope due to warmest water against shore, aided by wind 
setup narrowing and steepening the slope, favors the northward geo- 
strophic current on the slope and reinforces the inshore northward- 
moving half of the Michigan City-Benton Harbor eddy. The offshore 
southward-moving half of the eddy would be opposed by the geo- 
strophic tendency and probably break down into small mixing vortices 
contributing to the warming of the deep-basin water. 

As the summer progresses continued warming of the lake from 
shore toward the center eventually establishes a thermocline across 
the lake but, through July, surface temperatures of water in mid- 
lake are less than those along shore. The temperature-created 
tendency for a sloped surface tilting upward toward land is less 
as midlake-to-shore temperatures become more nearly alike, but 
wind setup still pushes surface water against shore to strengthen 
and maintain the slope upward toward land. Geostrophic northward 
current along shore should then still be the rule, but the off- 
shore south-moving half of the Michigan City-Benton Harbor eddy 
is probably established and flowing. 

By August the thermocline has been established and pushed to 
full depth across the whole of the lake, though slightly cooler 
surface temperatures are present in midlake than at the shores. 
The temperature induced slope upward toward land is now at its 



19 



weakest and is maintained largely by wind setup pushing warm 
surface water against shore. Northerly winds can now begin to * 
overpower the geostrophic tendency to north current and can drive 
the Michigan City-Benton Harbor eddy south of the plant site or 
wipe it out temporarily. Southward currents at the plant site 
appear when the eddy is out of the region. 

In September autumnal cooling has begun and the shallow 
inshore waters become cooler than the warm midbody of the lake. 
Cooler, contracted, more dense water alongshore now tends to 
produce a sloped water surface tilted upward toward midlake; this 
tendency to tilt is, however, opposed by wind setup pushing warm 
midlake surface water against shore. Northerly winds, temporarily 
replacing the prevailing westerlies, can now eliminate the wind 
setup and more easily and frequently than before produce southward 
flowing current at the plant site, by pushing the eddy south or 
wiping it out. Southward currents should now occur more frequently 
at the plant site, as is observed. 

Continued autumnal cooling, progressing from shore lakeward, 
produces increased tendency for a water surface tilted upward 
toward midlake, for the shallow water cools faster than the deep. 
Prevailing westerly winds still overpower this tendency by pushing 
warm surface water from midlake against the shore, but now wind 
setup is working against a strengthening geostrophic tendency for 
clockwise circulation around the lake (southward currents at the 
plant site) . Cessation of the prevailing wind is now more quickly 
and more frequently followed by southward current at the plant site. 



20 



as is observed in October. In all probability the Michigan City- 
Benton Harbor eddy is now being wiped out more frequently, but 
evidence is lacking. 

Currents Inshore of the Inner Bar : We have a limited number 
of observations of current direction in water immediately against 
the beach inshore of the inner sand bar. On every occasion 
current here was moving in the downwind direction. 

Apparently the water between the sand bars partakes in the 
most inshore part of the main-lake circulation, and the dominant 
north-south movements there may well be important factors in the 
maintenance of the two sand bars. 

Water inshore of the inner sand bar appears to be water which 
has spilled over the inner bar during surf action and which is cut 
off from the main-lake circulation by the inner bar. Cut off from 
the main-lake circulation, this water is subject only to the along- 
shore component of the wind blowing at the moment and its movement 
is downwind. If the plant effluent is discharged between the sand 
bars, it may be expected that under winds from north quarters some 
of the effluent will work across the inner bar to join the south- 
ward flow along the beach. Under winds from south quarters the 
spillage over the inner bar would move northward along the beach. 

The undesirable southward flow inside the inner bar under 
northerly winds plus the dominant northward flow between the bars 
under all winds are both reasons that the plant effluent should 
be discharged between the sand bars, as recommended by letter to 
Joel Gingold on 7 September 1967.